repository_name
stringclasses
316 values
func_path_in_repository
stringlengths
6
223
func_name
stringlengths
1
134
language
stringclasses
1 value
func_code_string
stringlengths
57
65.5k
func_documentation_string
stringlengths
1
46.3k
split_name
stringclasses
1 value
func_code_url
stringlengths
91
315
called_functions
listlengths
1
156
enclosing_scope
stringlengths
2
1.48M
hyde/fswrap
fswrap.py
Folder.child
python
def child(self, fragment): return os.path.join(self.path, FS(fragment).path)
Returns a path of a child item represented by `fragment`.
train
https://github.com/hyde/fswrap/blob/41e4ad6f7e9ba73eabe61bd97847cd284e3edbd2/fswrap.py#L543-L547
null
class Folder(FS): """ Represents a directory. """ def __init__(self, path): super(Folder, self).__init__(path) def child_folder(self, fragment): """ Returns a folder object by combining the fragment to this folder's path """ return Folder(os.path.join(self.path, Folder(fragment).path)) def child_file(self, fragment): """ Returns a `File` object representing the `fragment`. """ return File(self.child(fragment)) def make(self): """ Creates this directory and any of the missing directories in the path. Any errors that may occur are eaten. """ try: if not self.exists: logger.info("Creating %s" % self.path) os.makedirs(self.path) except os.error: pass return self def zip(self, target=None, basepath=None): """ Zips the contents of this folder. If `target` is not provided, <name>.zip is used instead. `basepath` is used to specify the base path for files in the archive. The path stored along with the files in the archive will be relative to the `basepath`. """ target = self.parent.child(target or self.name + '.zip') basepath = basepath or self.path from zipfile import ZipFile with ZipFile(target, 'w') as zip: with self.walker as walker: @walker.file_visitor def add_file(f): zip.write(f.path, f.get_relative_path(basepath)) def delete(self): """ Deletes the directory if it exists. """ if self.exists: logger.info("Deleting %s" % self.path) shutil.rmtree(self.path) def copy_to(self, destination): """ Copies this directory to the given destination. Returns a Folder object that represents the moved directory. """ target = self.__get_destination__(destination) logger.info("Copying %s to %s" % (self, target)) shutil.copytree(self.path, unicode(target)) return target def move_to(self, destination): """ Moves this directory to the given destination. Returns a Folder object that represents the moved directory. """ target = self.__get_destination__(destination) logger.info("Move %s to %s" % (self, target)) shutil.move(self.path, unicode(target)) return target def rename_to(self, destination_name): """ Moves this directory to the given destination. Returns a Folder object that represents the moved directory. """ target = self.parent.child_folder(destination_name) logger.info("Rename %s to %s" % (self, target)) shutil.move(self.path, unicode(target)) return target def _create_target_tree(self, target): """ There is a bug in dir_util that makes `copy_tree` crash if a folder in the tree has been deleted before and readded now. To workaround the bug, we first walk the tree and create directories that are needed. """ source = self with source.walker as walker: @walker.folder_visitor def visit_folder(folder): """ Create the mirror directory """ if folder != source: Folder(folder.get_mirror(target, source)).make() def copy_contents_to(self, destination): """ Copies the contents of this directory to the given destination. Returns a Folder object that represents the moved directory. """ logger.info("Copying contents of %s to %s" % (self, destination)) target = Folder(destination) target.make() self._create_target_tree(target) dir_util.copy_tree(self.path, unicode(target)) return target def get_walker(self, pattern=None): """ Return a `FolderWalker` object with a set pattern. """ return FolderWalker(self, pattern) @property def walker(self): """ Return a `FolderWalker` object """ return FolderWalker(self) def get_lister(self, pattern=None): """ Return a `FolderLister` object with a set pattern. """ return FolderLister(self, pattern) @property def lister(self): """ Return a `FolderLister` object """ return FolderLister(self)
hyde/fswrap
fswrap.py
Folder.make
python
def make(self): try: if not self.exists: logger.info("Creating %s" % self.path) os.makedirs(self.path) except os.error: pass return self
Creates this directory and any of the missing directories in the path. Any errors that may occur are eaten.
train
https://github.com/hyde/fswrap/blob/41e4ad6f7e9ba73eabe61bd97847cd284e3edbd2/fswrap.py#L549-L560
null
class Folder(FS): """ Represents a directory. """ def __init__(self, path): super(Folder, self).__init__(path) def child_folder(self, fragment): """ Returns a folder object by combining the fragment to this folder's path """ return Folder(os.path.join(self.path, Folder(fragment).path)) def child_file(self, fragment): """ Returns a `File` object representing the `fragment`. """ return File(self.child(fragment)) def child(self, fragment): """ Returns a path of a child item represented by `fragment`. """ return os.path.join(self.path, FS(fragment).path) def zip(self, target=None, basepath=None): """ Zips the contents of this folder. If `target` is not provided, <name>.zip is used instead. `basepath` is used to specify the base path for files in the archive. The path stored along with the files in the archive will be relative to the `basepath`. """ target = self.parent.child(target or self.name + '.zip') basepath = basepath or self.path from zipfile import ZipFile with ZipFile(target, 'w') as zip: with self.walker as walker: @walker.file_visitor def add_file(f): zip.write(f.path, f.get_relative_path(basepath)) def delete(self): """ Deletes the directory if it exists. """ if self.exists: logger.info("Deleting %s" % self.path) shutil.rmtree(self.path) def copy_to(self, destination): """ Copies this directory to the given destination. Returns a Folder object that represents the moved directory. """ target = self.__get_destination__(destination) logger.info("Copying %s to %s" % (self, target)) shutil.copytree(self.path, unicode(target)) return target def move_to(self, destination): """ Moves this directory to the given destination. Returns a Folder object that represents the moved directory. """ target = self.__get_destination__(destination) logger.info("Move %s to %s" % (self, target)) shutil.move(self.path, unicode(target)) return target def rename_to(self, destination_name): """ Moves this directory to the given destination. Returns a Folder object that represents the moved directory. """ target = self.parent.child_folder(destination_name) logger.info("Rename %s to %s" % (self, target)) shutil.move(self.path, unicode(target)) return target def _create_target_tree(self, target): """ There is a bug in dir_util that makes `copy_tree` crash if a folder in the tree has been deleted before and readded now. To workaround the bug, we first walk the tree and create directories that are needed. """ source = self with source.walker as walker: @walker.folder_visitor def visit_folder(folder): """ Create the mirror directory """ if folder != source: Folder(folder.get_mirror(target, source)).make() def copy_contents_to(self, destination): """ Copies the contents of this directory to the given destination. Returns a Folder object that represents the moved directory. """ logger.info("Copying contents of %s to %s" % (self, destination)) target = Folder(destination) target.make() self._create_target_tree(target) dir_util.copy_tree(self.path, unicode(target)) return target def get_walker(self, pattern=None): """ Return a `FolderWalker` object with a set pattern. """ return FolderWalker(self, pattern) @property def walker(self): """ Return a `FolderWalker` object """ return FolderWalker(self) def get_lister(self, pattern=None): """ Return a `FolderLister` object with a set pattern. """ return FolderLister(self, pattern) @property def lister(self): """ Return a `FolderLister` object """ return FolderLister(self)
hyde/fswrap
fswrap.py
Folder.zip
python
def zip(self, target=None, basepath=None): target = self.parent.child(target or self.name + '.zip') basepath = basepath or self.path from zipfile import ZipFile with ZipFile(target, 'w') as zip: with self.walker as walker: @walker.file_visitor def add_file(f): zip.write(f.path, f.get_relative_path(basepath))
Zips the contents of this folder. If `target` is not provided, <name>.zip is used instead. `basepath` is used to specify the base path for files in the archive. The path stored along with the files in the archive will be relative to the `basepath`.
train
https://github.com/hyde/fswrap/blob/41e4ad6f7e9ba73eabe61bd97847cd284e3edbd2/fswrap.py#L562-L576
null
class Folder(FS): """ Represents a directory. """ def __init__(self, path): super(Folder, self).__init__(path) def child_folder(self, fragment): """ Returns a folder object by combining the fragment to this folder's path """ return Folder(os.path.join(self.path, Folder(fragment).path)) def child_file(self, fragment): """ Returns a `File` object representing the `fragment`. """ return File(self.child(fragment)) def child(self, fragment): """ Returns a path of a child item represented by `fragment`. """ return os.path.join(self.path, FS(fragment).path) def make(self): """ Creates this directory and any of the missing directories in the path. Any errors that may occur are eaten. """ try: if not self.exists: logger.info("Creating %s" % self.path) os.makedirs(self.path) except os.error: pass return self def delete(self): """ Deletes the directory if it exists. """ if self.exists: logger.info("Deleting %s" % self.path) shutil.rmtree(self.path) def copy_to(self, destination): """ Copies this directory to the given destination. Returns a Folder object that represents the moved directory. """ target = self.__get_destination__(destination) logger.info("Copying %s to %s" % (self, target)) shutil.copytree(self.path, unicode(target)) return target def move_to(self, destination): """ Moves this directory to the given destination. Returns a Folder object that represents the moved directory. """ target = self.__get_destination__(destination) logger.info("Move %s to %s" % (self, target)) shutil.move(self.path, unicode(target)) return target def rename_to(self, destination_name): """ Moves this directory to the given destination. Returns a Folder object that represents the moved directory. """ target = self.parent.child_folder(destination_name) logger.info("Rename %s to %s" % (self, target)) shutil.move(self.path, unicode(target)) return target def _create_target_tree(self, target): """ There is a bug in dir_util that makes `copy_tree` crash if a folder in the tree has been deleted before and readded now. To workaround the bug, we first walk the tree and create directories that are needed. """ source = self with source.walker as walker: @walker.folder_visitor def visit_folder(folder): """ Create the mirror directory """ if folder != source: Folder(folder.get_mirror(target, source)).make() def copy_contents_to(self, destination): """ Copies the contents of this directory to the given destination. Returns a Folder object that represents the moved directory. """ logger.info("Copying contents of %s to %s" % (self, destination)) target = Folder(destination) target.make() self._create_target_tree(target) dir_util.copy_tree(self.path, unicode(target)) return target def get_walker(self, pattern=None): """ Return a `FolderWalker` object with a set pattern. """ return FolderWalker(self, pattern) @property def walker(self): """ Return a `FolderWalker` object """ return FolderWalker(self) def get_lister(self, pattern=None): """ Return a `FolderLister` object with a set pattern. """ return FolderLister(self, pattern) @property def lister(self): """ Return a `FolderLister` object """ return FolderLister(self)
hyde/fswrap
fswrap.py
Folder.delete
python
def delete(self): if self.exists: logger.info("Deleting %s" % self.path) shutil.rmtree(self.path)
Deletes the directory if it exists.
train
https://github.com/hyde/fswrap/blob/41e4ad6f7e9ba73eabe61bd97847cd284e3edbd2/fswrap.py#L578-L584
null
class Folder(FS): """ Represents a directory. """ def __init__(self, path): super(Folder, self).__init__(path) def child_folder(self, fragment): """ Returns a folder object by combining the fragment to this folder's path """ return Folder(os.path.join(self.path, Folder(fragment).path)) def child_file(self, fragment): """ Returns a `File` object representing the `fragment`. """ return File(self.child(fragment)) def child(self, fragment): """ Returns a path of a child item represented by `fragment`. """ return os.path.join(self.path, FS(fragment).path) def make(self): """ Creates this directory and any of the missing directories in the path. Any errors that may occur are eaten. """ try: if not self.exists: logger.info("Creating %s" % self.path) os.makedirs(self.path) except os.error: pass return self def zip(self, target=None, basepath=None): """ Zips the contents of this folder. If `target` is not provided, <name>.zip is used instead. `basepath` is used to specify the base path for files in the archive. The path stored along with the files in the archive will be relative to the `basepath`. """ target = self.parent.child(target or self.name + '.zip') basepath = basepath or self.path from zipfile import ZipFile with ZipFile(target, 'w') as zip: with self.walker as walker: @walker.file_visitor def add_file(f): zip.write(f.path, f.get_relative_path(basepath)) def copy_to(self, destination): """ Copies this directory to the given destination. Returns a Folder object that represents the moved directory. """ target = self.__get_destination__(destination) logger.info("Copying %s to %s" % (self, target)) shutil.copytree(self.path, unicode(target)) return target def move_to(self, destination): """ Moves this directory to the given destination. Returns a Folder object that represents the moved directory. """ target = self.__get_destination__(destination) logger.info("Move %s to %s" % (self, target)) shutil.move(self.path, unicode(target)) return target def rename_to(self, destination_name): """ Moves this directory to the given destination. Returns a Folder object that represents the moved directory. """ target = self.parent.child_folder(destination_name) logger.info("Rename %s to %s" % (self, target)) shutil.move(self.path, unicode(target)) return target def _create_target_tree(self, target): """ There is a bug in dir_util that makes `copy_tree` crash if a folder in the tree has been deleted before and readded now. To workaround the bug, we first walk the tree and create directories that are needed. """ source = self with source.walker as walker: @walker.folder_visitor def visit_folder(folder): """ Create the mirror directory """ if folder != source: Folder(folder.get_mirror(target, source)).make() def copy_contents_to(self, destination): """ Copies the contents of this directory to the given destination. Returns a Folder object that represents the moved directory. """ logger.info("Copying contents of %s to %s" % (self, destination)) target = Folder(destination) target.make() self._create_target_tree(target) dir_util.copy_tree(self.path, unicode(target)) return target def get_walker(self, pattern=None): """ Return a `FolderWalker` object with a set pattern. """ return FolderWalker(self, pattern) @property def walker(self): """ Return a `FolderWalker` object """ return FolderWalker(self) def get_lister(self, pattern=None): """ Return a `FolderLister` object with a set pattern. """ return FolderLister(self, pattern) @property def lister(self): """ Return a `FolderLister` object """ return FolderLister(self)
hyde/fswrap
fswrap.py
Folder.move_to
python
def move_to(self, destination): target = self.__get_destination__(destination) logger.info("Move %s to %s" % (self, target)) shutil.move(self.path, unicode(target)) return target
Moves this directory to the given destination. Returns a Folder object that represents the moved directory.
train
https://github.com/hyde/fswrap/blob/41e4ad6f7e9ba73eabe61bd97847cd284e3edbd2/fswrap.py#L596-L604
[ "def __get_destination__(self, destination):\n \"\"\"\n Returns a File or Folder object that would represent this entity\n if it were copied or moved to `destination`.\n \"\"\"\n if isinstance(destination,\n File) or os.path.isfile(unicode(destination)):\n return destination\n...
class Folder(FS): """ Represents a directory. """ def __init__(self, path): super(Folder, self).__init__(path) def child_folder(self, fragment): """ Returns a folder object by combining the fragment to this folder's path """ return Folder(os.path.join(self.path, Folder(fragment).path)) def child_file(self, fragment): """ Returns a `File` object representing the `fragment`. """ return File(self.child(fragment)) def child(self, fragment): """ Returns a path of a child item represented by `fragment`. """ return os.path.join(self.path, FS(fragment).path) def make(self): """ Creates this directory and any of the missing directories in the path. Any errors that may occur are eaten. """ try: if not self.exists: logger.info("Creating %s" % self.path) os.makedirs(self.path) except os.error: pass return self def zip(self, target=None, basepath=None): """ Zips the contents of this folder. If `target` is not provided, <name>.zip is used instead. `basepath` is used to specify the base path for files in the archive. The path stored along with the files in the archive will be relative to the `basepath`. """ target = self.parent.child(target or self.name + '.zip') basepath = basepath or self.path from zipfile import ZipFile with ZipFile(target, 'w') as zip: with self.walker as walker: @walker.file_visitor def add_file(f): zip.write(f.path, f.get_relative_path(basepath)) def delete(self): """ Deletes the directory if it exists. """ if self.exists: logger.info("Deleting %s" % self.path) shutil.rmtree(self.path) def copy_to(self, destination): """ Copies this directory to the given destination. Returns a Folder object that represents the moved directory. """ target = self.__get_destination__(destination) logger.info("Copying %s to %s" % (self, target)) shutil.copytree(self.path, unicode(target)) return target def rename_to(self, destination_name): """ Moves this directory to the given destination. Returns a Folder object that represents the moved directory. """ target = self.parent.child_folder(destination_name) logger.info("Rename %s to %s" % (self, target)) shutil.move(self.path, unicode(target)) return target def _create_target_tree(self, target): """ There is a bug in dir_util that makes `copy_tree` crash if a folder in the tree has been deleted before and readded now. To workaround the bug, we first walk the tree and create directories that are needed. """ source = self with source.walker as walker: @walker.folder_visitor def visit_folder(folder): """ Create the mirror directory """ if folder != source: Folder(folder.get_mirror(target, source)).make() def copy_contents_to(self, destination): """ Copies the contents of this directory to the given destination. Returns a Folder object that represents the moved directory. """ logger.info("Copying contents of %s to %s" % (self, destination)) target = Folder(destination) target.make() self._create_target_tree(target) dir_util.copy_tree(self.path, unicode(target)) return target def get_walker(self, pattern=None): """ Return a `FolderWalker` object with a set pattern. """ return FolderWalker(self, pattern) @property def walker(self): """ Return a `FolderWalker` object """ return FolderWalker(self) def get_lister(self, pattern=None): """ Return a `FolderLister` object with a set pattern. """ return FolderLister(self, pattern) @property def lister(self): """ Return a `FolderLister` object """ return FolderLister(self)
hyde/fswrap
fswrap.py
Folder.rename_to
python
def rename_to(self, destination_name): target = self.parent.child_folder(destination_name) logger.info("Rename %s to %s" % (self, target)) shutil.move(self.path, unicode(target)) return target
Moves this directory to the given destination. Returns a Folder object that represents the moved directory.
train
https://github.com/hyde/fswrap/blob/41e4ad6f7e9ba73eabe61bd97847cd284e3edbd2/fswrap.py#L606-L614
null
class Folder(FS): """ Represents a directory. """ def __init__(self, path): super(Folder, self).__init__(path) def child_folder(self, fragment): """ Returns a folder object by combining the fragment to this folder's path """ return Folder(os.path.join(self.path, Folder(fragment).path)) def child_file(self, fragment): """ Returns a `File` object representing the `fragment`. """ return File(self.child(fragment)) def child(self, fragment): """ Returns a path of a child item represented by `fragment`. """ return os.path.join(self.path, FS(fragment).path) def make(self): """ Creates this directory and any of the missing directories in the path. Any errors that may occur are eaten. """ try: if not self.exists: logger.info("Creating %s" % self.path) os.makedirs(self.path) except os.error: pass return self def zip(self, target=None, basepath=None): """ Zips the contents of this folder. If `target` is not provided, <name>.zip is used instead. `basepath` is used to specify the base path for files in the archive. The path stored along with the files in the archive will be relative to the `basepath`. """ target = self.parent.child(target or self.name + '.zip') basepath = basepath or self.path from zipfile import ZipFile with ZipFile(target, 'w') as zip: with self.walker as walker: @walker.file_visitor def add_file(f): zip.write(f.path, f.get_relative_path(basepath)) def delete(self): """ Deletes the directory if it exists. """ if self.exists: logger.info("Deleting %s" % self.path) shutil.rmtree(self.path) def copy_to(self, destination): """ Copies this directory to the given destination. Returns a Folder object that represents the moved directory. """ target = self.__get_destination__(destination) logger.info("Copying %s to %s" % (self, target)) shutil.copytree(self.path, unicode(target)) return target def move_to(self, destination): """ Moves this directory to the given destination. Returns a Folder object that represents the moved directory. """ target = self.__get_destination__(destination) logger.info("Move %s to %s" % (self, target)) shutil.move(self.path, unicode(target)) return target def _create_target_tree(self, target): """ There is a bug in dir_util that makes `copy_tree` crash if a folder in the tree has been deleted before and readded now. To workaround the bug, we first walk the tree and create directories that are needed. """ source = self with source.walker as walker: @walker.folder_visitor def visit_folder(folder): """ Create the mirror directory """ if folder != source: Folder(folder.get_mirror(target, source)).make() def copy_contents_to(self, destination): """ Copies the contents of this directory to the given destination. Returns a Folder object that represents the moved directory. """ logger.info("Copying contents of %s to %s" % (self, destination)) target = Folder(destination) target.make() self._create_target_tree(target) dir_util.copy_tree(self.path, unicode(target)) return target def get_walker(self, pattern=None): """ Return a `FolderWalker` object with a set pattern. """ return FolderWalker(self, pattern) @property def walker(self): """ Return a `FolderWalker` object """ return FolderWalker(self) def get_lister(self, pattern=None): """ Return a `FolderLister` object with a set pattern. """ return FolderLister(self, pattern) @property def lister(self): """ Return a `FolderLister` object """ return FolderLister(self)
hyde/fswrap
fswrap.py
Folder._create_target_tree
python
def _create_target_tree(self, target): source = self with source.walker as walker: @walker.folder_visitor def visit_folder(folder): """ Create the mirror directory """ if folder != source: Folder(folder.get_mirror(target, source)).make()
There is a bug in dir_util that makes `copy_tree` crash if a folder in the tree has been deleted before and readded now. To workaround the bug, we first walk the tree and create directories that are needed.
train
https://github.com/hyde/fswrap/blob/41e4ad6f7e9ba73eabe61bd97847cd284e3edbd2/fswrap.py#L616-L631
null
class Folder(FS): """ Represents a directory. """ def __init__(self, path): super(Folder, self).__init__(path) def child_folder(self, fragment): """ Returns a folder object by combining the fragment to this folder's path """ return Folder(os.path.join(self.path, Folder(fragment).path)) def child_file(self, fragment): """ Returns a `File` object representing the `fragment`. """ return File(self.child(fragment)) def child(self, fragment): """ Returns a path of a child item represented by `fragment`. """ return os.path.join(self.path, FS(fragment).path) def make(self): """ Creates this directory and any of the missing directories in the path. Any errors that may occur are eaten. """ try: if not self.exists: logger.info("Creating %s" % self.path) os.makedirs(self.path) except os.error: pass return self def zip(self, target=None, basepath=None): """ Zips the contents of this folder. If `target` is not provided, <name>.zip is used instead. `basepath` is used to specify the base path for files in the archive. The path stored along with the files in the archive will be relative to the `basepath`. """ target = self.parent.child(target or self.name + '.zip') basepath = basepath or self.path from zipfile import ZipFile with ZipFile(target, 'w') as zip: with self.walker as walker: @walker.file_visitor def add_file(f): zip.write(f.path, f.get_relative_path(basepath)) def delete(self): """ Deletes the directory if it exists. """ if self.exists: logger.info("Deleting %s" % self.path) shutil.rmtree(self.path) def copy_to(self, destination): """ Copies this directory to the given destination. Returns a Folder object that represents the moved directory. """ target = self.__get_destination__(destination) logger.info("Copying %s to %s" % (self, target)) shutil.copytree(self.path, unicode(target)) return target def move_to(self, destination): """ Moves this directory to the given destination. Returns a Folder object that represents the moved directory. """ target = self.__get_destination__(destination) logger.info("Move %s to %s" % (self, target)) shutil.move(self.path, unicode(target)) return target def rename_to(self, destination_name): """ Moves this directory to the given destination. Returns a Folder object that represents the moved directory. """ target = self.parent.child_folder(destination_name) logger.info("Rename %s to %s" % (self, target)) shutil.move(self.path, unicode(target)) return target def copy_contents_to(self, destination): """ Copies the contents of this directory to the given destination. Returns a Folder object that represents the moved directory. """ logger.info("Copying contents of %s to %s" % (self, destination)) target = Folder(destination) target.make() self._create_target_tree(target) dir_util.copy_tree(self.path, unicode(target)) return target def get_walker(self, pattern=None): """ Return a `FolderWalker` object with a set pattern. """ return FolderWalker(self, pattern) @property def walker(self): """ Return a `FolderWalker` object """ return FolderWalker(self) def get_lister(self, pattern=None): """ Return a `FolderLister` object with a set pattern. """ return FolderLister(self, pattern) @property def lister(self): """ Return a `FolderLister` object """ return FolderLister(self)
hyde/fswrap
fswrap.py
Folder.copy_contents_to
python
def copy_contents_to(self, destination): logger.info("Copying contents of %s to %s" % (self, destination)) target = Folder(destination) target.make() self._create_target_tree(target) dir_util.copy_tree(self.path, unicode(target)) return target
Copies the contents of this directory to the given destination. Returns a Folder object that represents the moved directory.
train
https://github.com/hyde/fswrap/blob/41e4ad6f7e9ba73eabe61bd97847cd284e3edbd2/fswrap.py#L633-L643
[ "def make(self):\n \"\"\"\n Creates this directory and any of the missing directories in the path.\n Any errors that may occur are eaten.\n \"\"\"\n try:\n if not self.exists:\n logger.info(\"Creating %s\" % self.path)\n os.makedirs(self.path)\n except os.error:\n ...
class Folder(FS): """ Represents a directory. """ def __init__(self, path): super(Folder, self).__init__(path) def child_folder(self, fragment): """ Returns a folder object by combining the fragment to this folder's path """ return Folder(os.path.join(self.path, Folder(fragment).path)) def child_file(self, fragment): """ Returns a `File` object representing the `fragment`. """ return File(self.child(fragment)) def child(self, fragment): """ Returns a path of a child item represented by `fragment`. """ return os.path.join(self.path, FS(fragment).path) def make(self): """ Creates this directory and any of the missing directories in the path. Any errors that may occur are eaten. """ try: if not self.exists: logger.info("Creating %s" % self.path) os.makedirs(self.path) except os.error: pass return self def zip(self, target=None, basepath=None): """ Zips the contents of this folder. If `target` is not provided, <name>.zip is used instead. `basepath` is used to specify the base path for files in the archive. The path stored along with the files in the archive will be relative to the `basepath`. """ target = self.parent.child(target or self.name + '.zip') basepath = basepath or self.path from zipfile import ZipFile with ZipFile(target, 'w') as zip: with self.walker as walker: @walker.file_visitor def add_file(f): zip.write(f.path, f.get_relative_path(basepath)) def delete(self): """ Deletes the directory if it exists. """ if self.exists: logger.info("Deleting %s" % self.path) shutil.rmtree(self.path) def copy_to(self, destination): """ Copies this directory to the given destination. Returns a Folder object that represents the moved directory. """ target = self.__get_destination__(destination) logger.info("Copying %s to %s" % (self, target)) shutil.copytree(self.path, unicode(target)) return target def move_to(self, destination): """ Moves this directory to the given destination. Returns a Folder object that represents the moved directory. """ target = self.__get_destination__(destination) logger.info("Move %s to %s" % (self, target)) shutil.move(self.path, unicode(target)) return target def rename_to(self, destination_name): """ Moves this directory to the given destination. Returns a Folder object that represents the moved directory. """ target = self.parent.child_folder(destination_name) logger.info("Rename %s to %s" % (self, target)) shutil.move(self.path, unicode(target)) return target def _create_target_tree(self, target): """ There is a bug in dir_util that makes `copy_tree` crash if a folder in the tree has been deleted before and readded now. To workaround the bug, we first walk the tree and create directories that are needed. """ source = self with source.walker as walker: @walker.folder_visitor def visit_folder(folder): """ Create the mirror directory """ if folder != source: Folder(folder.get_mirror(target, source)).make() def get_walker(self, pattern=None): """ Return a `FolderWalker` object with a set pattern. """ return FolderWalker(self, pattern) @property def walker(self): """ Return a `FolderWalker` object """ return FolderWalker(self) def get_lister(self, pattern=None): """ Return a `FolderLister` object with a set pattern. """ return FolderLister(self, pattern) @property def lister(self): """ Return a `FolderLister` object """ return FolderLister(self)
staggerpkg/stagger
stagger/frames.py
Frame._from_frame
python
def _from_frame(cls, frame): "Copy constructor" assert frame._framespec == cls._framespec new = cls(flags=frame.flags, frameno=frame.frameno) for spec in cls._framespec: setattr(new, spec.name, getattr(frame, spec.name, None)) return new
Copy constructor
train
https://github.com/staggerpkg/stagger/blob/6530db14afc5d7d8a4599b7f3b26158fb367d786/stagger/frames.py#L91-L97
null
class Frame(metaclass=abc.ABCMeta): _framespec = tuple() _version = tuple() _allow_duplicates = False def __init__(self, frameid=None, flags=None, frameno=None, **kwargs): self.frameid = frameid if frameid else type(self).__name__ self.flags = flags if flags else set() self.frameno = frameno assert len(self._framespec) > 0 for spec in self._framespec: val = kwargs.get(spec.name, None) setattr(self, spec.name, val) def __setattr__(self, name, value): # Automatic validation on assignment for spec in self._framespec: if name == spec.name: value = spec.validate(self, value) break super().__setattr__(name, value) def __eq__(self, other): return (isinstance(other, type(self)) and self.frameid == other.frameid and self.flags == other.flags and self._framespec == other._framespec and all(getattr(self, spec.name, None) == getattr(other, spec.name, None) for spec in self._framespec)) @classmethod def _decode(cls, frameid, data, flags=None, frameno=None): frame = cls(frameid=frameid, flags=flags, frameno=frameno) if getattr(frame, "_untested", False): warn("{0}: Untested frame; please verify results".format(frameid), UntestedFrameWarning) for spec in frame._framespec: try: val, data = spec.read(frame, data) setattr(frame, spec.name, val) except EOFError: if not spec._optional: raise return frame @classmethod @classmethod def _merge(cls, frames): if cls._allow_duplicates: return frames else: if len(frames) > 1: # TODO: Research what iTunes does in this case # Mutagen displays the first frame only. warn("{0}: Duplicate frame; only the first instance is kept" .format(frames[0].frameid), DuplicateFrameWarning) return frames[0:1] @classmethod def _in_version(self, *versions): "Returns true if this frame is in any of the specified versions of ID3." for version in versions: if (self._version == version or (isinstance(self._version, collections.Container) and version in self._version)): return True return False def _to_version(self, version): if self._in_version(version): return self if version == 2 and hasattr(self, "_v2_frame"): return self._v2_frame._from_frame(self) if self._in_version(2): base = type(self).__bases__[0] if issubclass(base, Frame) and base._in_version(version): return base._from_frame(self) raise IncompatibleFrameError("Frame {0} cannot be converted " "to ID3v2.{1} format".format(self.frameid, version)) def _encode(self, encodings=("latin-1", "utf-16")): # if getattr(self, "_bozo", False): # warn("{0}: Frame type is not widely implemented, " # "its use is discouraged".format(self.frameid), # BozoFrameWarning) def encode_fields(): data = bytearray() for spec in self._framespec: if spec._optional and getattr(self, spec.name) is None: break data.extend(spec.write(self, getattr(self, spec.name))) return data def try_preferred_encodings(): orig_encoding = self.encoding try: for encoding in encodings: try: self.encoding = encoding return encode_fields() except UnicodeEncodeError: pass finally: self.encoding = orig_encoding raise ValueError("Could not encode strings") if not isinstance(self._framespec[0], EncodingSpec): return encode_fields() elif self.encoding is None: return try_preferred_encodings() else: try: # Try specified encoding before others return encode_fields() except UnicodeEncodeError: return try_preferred_encodings() def __repr__(self): stype = type(self).__name__ args = [] if stype != self.frameid: args.append("frameid={0!r}".format(self.frameid)) if self.flags: args.append("flags={0!r}".format(self.flags)) for spec in self._framespec: if isinstance(spec, BinaryDataSpec): data = getattr(self, spec.name) if isinstance(data, (bytes, bytearray)): args.append("{0}=<{1} bytes of binary data {2!r}{3}>".format( spec.name, len(data), data[:20], "..." if len(data) > 20 else "")) else: args.append(repr(data)) else: args.append("{0}={1!r}".format(spec.name, getattr(self, spec.name))) return "{0}({1})".format(stype, ", ".join(args)) def _spec(self, name): "Return the named spec." for s in self._framespec: if s.name == name: return s raise ValueError("Unknown spec: " + name) def _str_fields(self): fields = [] # Determine how many fields to show cutoff = max(i for i in range(len(self._framespec)) if i == 0 # don't call max with an the empty sequence or not self._framespec[i]._optional or getattr(self, self._framespec[i].name, None) is not None) for spec in self._framespec[:cutoff + 1]: fields.append("{0}={1}" .format(spec.name, repr(spec.to_str(getattr(self, spec.name, None))))) return ", ".join(fields) def __str__(self): flag = " " if "unknown" in self.flags: flag = "?" if isinstance(self, ErrorFrame): flag = "!" return "{0}{1}({2})".format(flag, self.frameid, self._str_fields())
staggerpkg/stagger
stagger/frames.py
Frame._in_version
python
def _in_version(self, *versions): "Returns true if this frame is in any of the specified versions of ID3." for version in versions: if (self._version == version or (isinstance(self._version, collections.Container) and version in self._version)): return True return False
Returns true if this frame is in any of the specified versions of ID3.
train
https://github.com/staggerpkg/stagger/blob/6530db14afc5d7d8a4599b7f3b26158fb367d786/stagger/frames.py#L113-L120
null
class Frame(metaclass=abc.ABCMeta): _framespec = tuple() _version = tuple() _allow_duplicates = False def __init__(self, frameid=None, flags=None, frameno=None, **kwargs): self.frameid = frameid if frameid else type(self).__name__ self.flags = flags if flags else set() self.frameno = frameno assert len(self._framespec) > 0 for spec in self._framespec: val = kwargs.get(spec.name, None) setattr(self, spec.name, val) def __setattr__(self, name, value): # Automatic validation on assignment for spec in self._framespec: if name == spec.name: value = spec.validate(self, value) break super().__setattr__(name, value) def __eq__(self, other): return (isinstance(other, type(self)) and self.frameid == other.frameid and self.flags == other.flags and self._framespec == other._framespec and all(getattr(self, spec.name, None) == getattr(other, spec.name, None) for spec in self._framespec)) @classmethod def _decode(cls, frameid, data, flags=None, frameno=None): frame = cls(frameid=frameid, flags=flags, frameno=frameno) if getattr(frame, "_untested", False): warn("{0}: Untested frame; please verify results".format(frameid), UntestedFrameWarning) for spec in frame._framespec: try: val, data = spec.read(frame, data) setattr(frame, spec.name, val) except EOFError: if not spec._optional: raise return frame @classmethod def _from_frame(cls, frame): "Copy constructor" assert frame._framespec == cls._framespec new = cls(flags=frame.flags, frameno=frame.frameno) for spec in cls._framespec: setattr(new, spec.name, getattr(frame, spec.name, None)) return new @classmethod def _merge(cls, frames): if cls._allow_duplicates: return frames else: if len(frames) > 1: # TODO: Research what iTunes does in this case # Mutagen displays the first frame only. warn("{0}: Duplicate frame; only the first instance is kept" .format(frames[0].frameid), DuplicateFrameWarning) return frames[0:1] @classmethod def _to_version(self, version): if self._in_version(version): return self if version == 2 and hasattr(self, "_v2_frame"): return self._v2_frame._from_frame(self) if self._in_version(2): base = type(self).__bases__[0] if issubclass(base, Frame) and base._in_version(version): return base._from_frame(self) raise IncompatibleFrameError("Frame {0} cannot be converted " "to ID3v2.{1} format".format(self.frameid, version)) def _encode(self, encodings=("latin-1", "utf-16")): # if getattr(self, "_bozo", False): # warn("{0}: Frame type is not widely implemented, " # "its use is discouraged".format(self.frameid), # BozoFrameWarning) def encode_fields(): data = bytearray() for spec in self._framespec: if spec._optional and getattr(self, spec.name) is None: break data.extend(spec.write(self, getattr(self, spec.name))) return data def try_preferred_encodings(): orig_encoding = self.encoding try: for encoding in encodings: try: self.encoding = encoding return encode_fields() except UnicodeEncodeError: pass finally: self.encoding = orig_encoding raise ValueError("Could not encode strings") if not isinstance(self._framespec[0], EncodingSpec): return encode_fields() elif self.encoding is None: return try_preferred_encodings() else: try: # Try specified encoding before others return encode_fields() except UnicodeEncodeError: return try_preferred_encodings() def __repr__(self): stype = type(self).__name__ args = [] if stype != self.frameid: args.append("frameid={0!r}".format(self.frameid)) if self.flags: args.append("flags={0!r}".format(self.flags)) for spec in self._framespec: if isinstance(spec, BinaryDataSpec): data = getattr(self, spec.name) if isinstance(data, (bytes, bytearray)): args.append("{0}=<{1} bytes of binary data {2!r}{3}>".format( spec.name, len(data), data[:20], "..." if len(data) > 20 else "")) else: args.append(repr(data)) else: args.append("{0}={1!r}".format(spec.name, getattr(self, spec.name))) return "{0}({1})".format(stype, ", ".join(args)) def _spec(self, name): "Return the named spec." for s in self._framespec: if s.name == name: return s raise ValueError("Unknown spec: " + name) def _str_fields(self): fields = [] # Determine how many fields to show cutoff = max(i for i in range(len(self._framespec)) if i == 0 # don't call max with an the empty sequence or not self._framespec[i]._optional or getattr(self, self._framespec[i].name, None) is not None) for spec in self._framespec[:cutoff + 1]: fields.append("{0}={1}" .format(spec.name, repr(spec.to_str(getattr(self, spec.name, None))))) return ", ".join(fields) def __str__(self): flag = " " if "unknown" in self.flags: flag = "?" if isinstance(self, ErrorFrame): flag = "!" return "{0}{1}({2})".format(flag, self.frameid, self._str_fields())
staggerpkg/stagger
stagger/frames.py
Frame._spec
python
def _spec(self, name): "Return the named spec." for s in self._framespec: if s.name == name: return s raise ValueError("Unknown spec: " + name)
Return the named spec.
train
https://github.com/staggerpkg/stagger/blob/6530db14afc5d7d8a4599b7f3b26158fb367d786/stagger/frames.py#L192-L197
null
class Frame(metaclass=abc.ABCMeta): _framespec = tuple() _version = tuple() _allow_duplicates = False def __init__(self, frameid=None, flags=None, frameno=None, **kwargs): self.frameid = frameid if frameid else type(self).__name__ self.flags = flags if flags else set() self.frameno = frameno assert len(self._framespec) > 0 for spec in self._framespec: val = kwargs.get(spec.name, None) setattr(self, spec.name, val) def __setattr__(self, name, value): # Automatic validation on assignment for spec in self._framespec: if name == spec.name: value = spec.validate(self, value) break super().__setattr__(name, value) def __eq__(self, other): return (isinstance(other, type(self)) and self.frameid == other.frameid and self.flags == other.flags and self._framespec == other._framespec and all(getattr(self, spec.name, None) == getattr(other, spec.name, None) for spec in self._framespec)) @classmethod def _decode(cls, frameid, data, flags=None, frameno=None): frame = cls(frameid=frameid, flags=flags, frameno=frameno) if getattr(frame, "_untested", False): warn("{0}: Untested frame; please verify results".format(frameid), UntestedFrameWarning) for spec in frame._framespec: try: val, data = spec.read(frame, data) setattr(frame, spec.name, val) except EOFError: if not spec._optional: raise return frame @classmethod def _from_frame(cls, frame): "Copy constructor" assert frame._framespec == cls._framespec new = cls(flags=frame.flags, frameno=frame.frameno) for spec in cls._framespec: setattr(new, spec.name, getattr(frame, spec.name, None)) return new @classmethod def _merge(cls, frames): if cls._allow_duplicates: return frames else: if len(frames) > 1: # TODO: Research what iTunes does in this case # Mutagen displays the first frame only. warn("{0}: Duplicate frame; only the first instance is kept" .format(frames[0].frameid), DuplicateFrameWarning) return frames[0:1] @classmethod def _in_version(self, *versions): "Returns true if this frame is in any of the specified versions of ID3." for version in versions: if (self._version == version or (isinstance(self._version, collections.Container) and version in self._version)): return True return False def _to_version(self, version): if self._in_version(version): return self if version == 2 and hasattr(self, "_v2_frame"): return self._v2_frame._from_frame(self) if self._in_version(2): base = type(self).__bases__[0] if issubclass(base, Frame) and base._in_version(version): return base._from_frame(self) raise IncompatibleFrameError("Frame {0} cannot be converted " "to ID3v2.{1} format".format(self.frameid, version)) def _encode(self, encodings=("latin-1", "utf-16")): # if getattr(self, "_bozo", False): # warn("{0}: Frame type is not widely implemented, " # "its use is discouraged".format(self.frameid), # BozoFrameWarning) def encode_fields(): data = bytearray() for spec in self._framespec: if spec._optional and getattr(self, spec.name) is None: break data.extend(spec.write(self, getattr(self, spec.name))) return data def try_preferred_encodings(): orig_encoding = self.encoding try: for encoding in encodings: try: self.encoding = encoding return encode_fields() except UnicodeEncodeError: pass finally: self.encoding = orig_encoding raise ValueError("Could not encode strings") if not isinstance(self._framespec[0], EncodingSpec): return encode_fields() elif self.encoding is None: return try_preferred_encodings() else: try: # Try specified encoding before others return encode_fields() except UnicodeEncodeError: return try_preferred_encodings() def __repr__(self): stype = type(self).__name__ args = [] if stype != self.frameid: args.append("frameid={0!r}".format(self.frameid)) if self.flags: args.append("flags={0!r}".format(self.flags)) for spec in self._framespec: if isinstance(spec, BinaryDataSpec): data = getattr(self, spec.name) if isinstance(data, (bytes, bytearray)): args.append("{0}=<{1} bytes of binary data {2!r}{3}>".format( spec.name, len(data), data[:20], "..." if len(data) > 20 else "")) else: args.append(repr(data)) else: args.append("{0}={1!r}".format(spec.name, getattr(self, spec.name))) return "{0}({1})".format(stype, ", ".join(args)) def _str_fields(self): fields = [] # Determine how many fields to show cutoff = max(i for i in range(len(self._framespec)) if i == 0 # don't call max with an the empty sequence or not self._framespec[i]._optional or getattr(self, self._framespec[i].name, None) is not None) for spec in self._framespec[:cutoff + 1]: fields.append("{0}={1}" .format(spec.name, repr(spec.to_str(getattr(self, spec.name, None))))) return ", ".join(fields) def __str__(self): flag = " " if "unknown" in self.flags: flag = "?" if isinstance(self, ErrorFrame): flag = "!" return "{0}{1}({2})".format(flag, self.frameid, self._str_fields())
staggerpkg/stagger
stagger/util.py
check_tag_data
python
def check_tag_data(data): "Raise a ValueError if DATA doesn't seem to be a well-formed ID3 tag." if len(data) < 10: raise ValueError("Tag too short") if data[0:3] != b"ID3": raise ValueError("Missing ID3 identifier") if data[3] >= 5 or data[4] != 0: raise ValueError("Unknown ID3 version") length = stagger.conversion.Syncsafe.decode(data[6:10]) + 10 if len(data) != length: raise ValueError("Tag size mismatch")
Raise a ValueError if DATA doesn't seem to be a well-formed ID3 tag.
train
https://github.com/staggerpkg/stagger/blob/6530db14afc5d7d8a4599b7f3b26158fb367d786/stagger/util.py#L52-L62
[ "def decode(data):\n \"Decodes a syncsafe integer\"\n value = 0\n for b in data:\n if b > 127: # iTunes bug\n raise ValueError(\"Invalid syncsafe integer\")\n value <<= 7\n value += b\n return value\n" ]
# # util.py # From the stagger project: http://code.google.com/p/stagger/ # # Copyright (c) 2009-2011 Karoly Lorentey <karoly@lorentey.hu> # All rights reserved. # # Redistribution and use in source and binary forms, with or without # modification, are permitted provided that the following conditions # are met: # # - Redistributions of source code must retain the above copyright # notice, this list of conditions and the following disclaimer. # # - Redistributions in binary form must reproduce the above copyright # notice, this list of conditions and the following disclaimer in # the documentation and/or other materials provided with the # distribution. # # THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS # "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT # LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS # FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE # COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, # INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, # BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; # LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER # CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT # LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN # ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE # POSSIBILITY OF SUCH DAMAGE. import warnings import sys from contextlib import contextmanager import stagger def python_version_check(): if sys.version_info[0:3] == (3, 1, 0): print("There are data corruption issues with Python 3.1.0's io module; \n" "please upgrade Python to at least 3.1.1 in order for Stagger \n" "to work reliably.\n\n" "For more information, see http://bugs.python.org/issue6629.", file=sys.stderr) exit(2) def verb(verbose, *args, **kwargs): if verbose: print(*args, **kwargs) def get_raw_tag_data(filename): "Return the ID3 tag in FILENAME as a raw byte string." with open(filename, "rb") as file: try: (cls, offset, length) = stagger.tags.detect_tag(file) except stagger.NoTagError: return bytes() file.seek(offset) return file.read(length) def set_raw_tag_data(filename, data, act=True, verbose=False): "Replace the ID3 tag in FILENAME with DATA." check_tag_data(data) with open(filename, "rb+") as file: try: (cls, offset, length) = stagger.tags.detect_tag(file) except stagger.NoTagError: (offset, length) = (0, 0) if length > 0: verb(verbose, "{0}: replaced tag with {1} bytes of data" .format(filename, len(data))) else: verb(verbose, "{0}: created tag with {1} bytes of data" .format(filename, len(data))) if act: stagger.fileutil.replace_chunk(file, offset, length, data) def set_frames(filename, valuedict, act=True, verbose=False): try: tag = stagger.read_tag(filename) except stagger.NoTagError: verb(verbose, "{0}: new ID3v2.{1} tag" .format(filename, stagger.default_tag.version)) tag = stagger.default_tag() for (key, value) in valuedict.items(): if key.lower() in tag._friendly_names: # Use friendly name API key = key.lower().replace("-", "_") assert hasattr(tag, key) setattr(tag, key, value) newval = repr(getattr(tag, key)) else: # Use frameid API tag[key] = value newval = tag[key] verb(verbose, "{0}: {1}: set to {2}".format(filename, key, newval)) if act: tag.write(filename) def remove_frames(filename, frameids, act=True, verbose=False): try: tag = stagger.read_tag(filename) except stagger.NoTagError: verb(verbose, "{0}: no ID3 tag".format(filename)) return for frameid in frameids: try: del tag[frameid] verb(verbose, "{0}: {1}: deleted".format(filename, frameid)) except KeyError: verb(verbose, "{0}: {1}: not in file".format(filename, frameid)) if act: tag.write(filename) @contextmanager def print_warnings(filename, options): with warnings.catch_warnings(record=True) as ws: try: yield None finally: if not options.quiet and len(ws) > 0: for w in ws: print(filename + ":warning: " + str(w.message), file=sys.stderr) sys.stderr.flush()
staggerpkg/stagger
stagger/util.py
get_raw_tag_data
python
def get_raw_tag_data(filename): "Return the ID3 tag in FILENAME as a raw byte string." with open(filename, "rb") as file: try: (cls, offset, length) = stagger.tags.detect_tag(file) except stagger.NoTagError: return bytes() file.seek(offset) return file.read(length)
Return the ID3 tag in FILENAME as a raw byte string.
train
https://github.com/staggerpkg/stagger/blob/6530db14afc5d7d8a4599b7f3b26158fb367d786/stagger/util.py#L64-L72
[ "def detect_tag(filename):\n \"\"\"Return type and position of ID3v2 tag in filename.\n Returns (tag_class, offset, length), where tag_class\n is either Tag22, Tag23, or Tag24, and (offset, length)\n is the position of the tag in the file.\n \"\"\"\n with fileutil.opened(filename, \"rb\") as file:...
# # util.py # From the stagger project: http://code.google.com/p/stagger/ # # Copyright (c) 2009-2011 Karoly Lorentey <karoly@lorentey.hu> # All rights reserved. # # Redistribution and use in source and binary forms, with or without # modification, are permitted provided that the following conditions # are met: # # - Redistributions of source code must retain the above copyright # notice, this list of conditions and the following disclaimer. # # - Redistributions in binary form must reproduce the above copyright # notice, this list of conditions and the following disclaimer in # the documentation and/or other materials provided with the # distribution. # # THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS # "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT # LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS # FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE # COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, # INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, # BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; # LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER # CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT # LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN # ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE # POSSIBILITY OF SUCH DAMAGE. import warnings import sys from contextlib import contextmanager import stagger def python_version_check(): if sys.version_info[0:3] == (3, 1, 0): print("There are data corruption issues with Python 3.1.0's io module; \n" "please upgrade Python to at least 3.1.1 in order for Stagger \n" "to work reliably.\n\n" "For more information, see http://bugs.python.org/issue6629.", file=sys.stderr) exit(2) def verb(verbose, *args, **kwargs): if verbose: print(*args, **kwargs) def check_tag_data(data): "Raise a ValueError if DATA doesn't seem to be a well-formed ID3 tag." if len(data) < 10: raise ValueError("Tag too short") if data[0:3] != b"ID3": raise ValueError("Missing ID3 identifier") if data[3] >= 5 or data[4] != 0: raise ValueError("Unknown ID3 version") length = stagger.conversion.Syncsafe.decode(data[6:10]) + 10 if len(data) != length: raise ValueError("Tag size mismatch") def set_raw_tag_data(filename, data, act=True, verbose=False): "Replace the ID3 tag in FILENAME with DATA." check_tag_data(data) with open(filename, "rb+") as file: try: (cls, offset, length) = stagger.tags.detect_tag(file) except stagger.NoTagError: (offset, length) = (0, 0) if length > 0: verb(verbose, "{0}: replaced tag with {1} bytes of data" .format(filename, len(data))) else: verb(verbose, "{0}: created tag with {1} bytes of data" .format(filename, len(data))) if act: stagger.fileutil.replace_chunk(file, offset, length, data) def set_frames(filename, valuedict, act=True, verbose=False): try: tag = stagger.read_tag(filename) except stagger.NoTagError: verb(verbose, "{0}: new ID3v2.{1} tag" .format(filename, stagger.default_tag.version)) tag = stagger.default_tag() for (key, value) in valuedict.items(): if key.lower() in tag._friendly_names: # Use friendly name API key = key.lower().replace("-", "_") assert hasattr(tag, key) setattr(tag, key, value) newval = repr(getattr(tag, key)) else: # Use frameid API tag[key] = value newval = tag[key] verb(verbose, "{0}: {1}: set to {2}".format(filename, key, newval)) if act: tag.write(filename) def remove_frames(filename, frameids, act=True, verbose=False): try: tag = stagger.read_tag(filename) except stagger.NoTagError: verb(verbose, "{0}: no ID3 tag".format(filename)) return for frameid in frameids: try: del tag[frameid] verb(verbose, "{0}: {1}: deleted".format(filename, frameid)) except KeyError: verb(verbose, "{0}: {1}: not in file".format(filename, frameid)) if act: tag.write(filename) @contextmanager def print_warnings(filename, options): with warnings.catch_warnings(record=True) as ws: try: yield None finally: if not options.quiet and len(ws) > 0: for w in ws: print(filename + ":warning: " + str(w.message), file=sys.stderr) sys.stderr.flush()
staggerpkg/stagger
stagger/util.py
set_raw_tag_data
python
def set_raw_tag_data(filename, data, act=True, verbose=False): "Replace the ID3 tag in FILENAME with DATA." check_tag_data(data) with open(filename, "rb+") as file: try: (cls, offset, length) = stagger.tags.detect_tag(file) except stagger.NoTagError: (offset, length) = (0, 0) if length > 0: verb(verbose, "{0}: replaced tag with {1} bytes of data" .format(filename, len(data))) else: verb(verbose, "{0}: created tag with {1} bytes of data" .format(filename, len(data))) if act: stagger.fileutil.replace_chunk(file, offset, length, data)
Replace the ID3 tag in FILENAME with DATA.
train
https://github.com/staggerpkg/stagger/blob/6530db14afc5d7d8a4599b7f3b26158fb367d786/stagger/util.py#L74-L89
[ "def replace_chunk(filename, offset, length, chunk, in_place=True, max_mem=5):\n \"\"\"Replace length bytes of data with chunk, starting at offset.\n Any KeyboardInterrupts arriving while replace_chunk is runnning\n are deferred until the operation is complete.\n\n If in_place is true, the operation wor...
# # util.py # From the stagger project: http://code.google.com/p/stagger/ # # Copyright (c) 2009-2011 Karoly Lorentey <karoly@lorentey.hu> # All rights reserved. # # Redistribution and use in source and binary forms, with or without # modification, are permitted provided that the following conditions # are met: # # - Redistributions of source code must retain the above copyright # notice, this list of conditions and the following disclaimer. # # - Redistributions in binary form must reproduce the above copyright # notice, this list of conditions and the following disclaimer in # the documentation and/or other materials provided with the # distribution. # # THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS # "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT # LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS # FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE # COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, # INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, # BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; # LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER # CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT # LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN # ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE # POSSIBILITY OF SUCH DAMAGE. import warnings import sys from contextlib import contextmanager import stagger def python_version_check(): if sys.version_info[0:3] == (3, 1, 0): print("There are data corruption issues with Python 3.1.0's io module; \n" "please upgrade Python to at least 3.1.1 in order for Stagger \n" "to work reliably.\n\n" "For more information, see http://bugs.python.org/issue6629.", file=sys.stderr) exit(2) def verb(verbose, *args, **kwargs): if verbose: print(*args, **kwargs) def check_tag_data(data): "Raise a ValueError if DATA doesn't seem to be a well-formed ID3 tag." if len(data) < 10: raise ValueError("Tag too short") if data[0:3] != b"ID3": raise ValueError("Missing ID3 identifier") if data[3] >= 5 or data[4] != 0: raise ValueError("Unknown ID3 version") length = stagger.conversion.Syncsafe.decode(data[6:10]) + 10 if len(data) != length: raise ValueError("Tag size mismatch") def get_raw_tag_data(filename): "Return the ID3 tag in FILENAME as a raw byte string." with open(filename, "rb") as file: try: (cls, offset, length) = stagger.tags.detect_tag(file) except stagger.NoTagError: return bytes() file.seek(offset) return file.read(length) def set_frames(filename, valuedict, act=True, verbose=False): try: tag = stagger.read_tag(filename) except stagger.NoTagError: verb(verbose, "{0}: new ID3v2.{1} tag" .format(filename, stagger.default_tag.version)) tag = stagger.default_tag() for (key, value) in valuedict.items(): if key.lower() in tag._friendly_names: # Use friendly name API key = key.lower().replace("-", "_") assert hasattr(tag, key) setattr(tag, key, value) newval = repr(getattr(tag, key)) else: # Use frameid API tag[key] = value newval = tag[key] verb(verbose, "{0}: {1}: set to {2}".format(filename, key, newval)) if act: tag.write(filename) def remove_frames(filename, frameids, act=True, verbose=False): try: tag = stagger.read_tag(filename) except stagger.NoTagError: verb(verbose, "{0}: no ID3 tag".format(filename)) return for frameid in frameids: try: del tag[frameid] verb(verbose, "{0}: {1}: deleted".format(filename, frameid)) except KeyError: verb(verbose, "{0}: {1}: not in file".format(filename, frameid)) if act: tag.write(filename) @contextmanager def print_warnings(filename, options): with warnings.catch_warnings(record=True) as ws: try: yield None finally: if not options.quiet and len(ws) > 0: for w in ws: print(filename + ":warning: " + str(w.message), file=sys.stderr) sys.stderr.flush()
staggerpkg/stagger
stagger/specs.py
SequenceSpec.read
python
def read(self, frame, data): "Returns a list of values, eats all of data." seq = [] while data: elem, data = self.spec.read(frame, data) seq.append(elem) return seq, data
Returns a list of values, eats all of data.
train
https://github.com/staggerpkg/stagger/blob/6530db14afc5d7d8a4599b7f3b26158fb367d786/stagger/specs.py#L367-L373
[ "def read(self, frame, data):\n rawstr, sep, data = data.partition(b\"\\x00\")\n return rawstr.decode('latin-1'), data\n" ]
class SequenceSpec(Spec): """Recognizes a sequence of values, all of the same spec.""" def __init__(self, name, spec): super().__init__(name) self.spec = spec def write(self, frame, values): if isinstance(values, str): return self.spec.write(frame, values) data = bytearray() for v in values: data.extend(self.spec.write(frame, v)) return data def validate(self, frame, values): if values is None: return [] if isinstance(values, str): values = [values] return [self.spec.validate(frame, v) for v in values]
staggerpkg/stagger
stagger/conversion.py
Unsync.gen_decode
python
def gen_decode(iterable): "A generator for de-unsynchronizing a byte iterable." sync = False for b in iterable: if sync and b & 0xE0: warn("Invalid unsynched data", Warning) if not (sync and b == 0x00): yield b sync = (b == 0xFF)
A generator for de-unsynchronizing a byte iterable.
train
https://github.com/staggerpkg/stagger/blob/6530db14afc5d7d8a4599b7f3b26158fb367d786/stagger/conversion.py#L39-L47
null
class Unsync: "Conversion from/to unsynchronized byte sequences." @staticmethod @staticmethod def gen_encode(data): "A generator for unsynchronizing a byte iterable." sync = False for b in data: if sync and (b == 0x00 or b & 0xE0): yield 0x00 # Insert sync char yield b sync = (b == 0xFF) if sync: yield 0x00 # Data ends on 0xFF @staticmethod def decode(data): "Remove unsynchronization bytes from data." return bytes(Unsync.gen_decode(data)) @staticmethod def encode(data): "Insert unsynchronization bytes into data." return bytes(Unsync.gen_encode(data))
staggerpkg/stagger
stagger/conversion.py
Unsync.gen_encode
python
def gen_encode(data): "A generator for unsynchronizing a byte iterable." sync = False for b in data: if sync and (b == 0x00 or b & 0xE0): yield 0x00 # Insert sync char yield b sync = (b == 0xFF) if sync: yield 0x00
A generator for unsynchronizing a byte iterable.
train
https://github.com/staggerpkg/stagger/blob/6530db14afc5d7d8a4599b7f3b26158fb367d786/stagger/conversion.py#L50-L59
null
class Unsync: "Conversion from/to unsynchronized byte sequences." @staticmethod def gen_decode(iterable): "A generator for de-unsynchronizing a byte iterable." sync = False for b in iterable: if sync and b & 0xE0: warn("Invalid unsynched data", Warning) if not (sync and b == 0x00): yield b sync = (b == 0xFF) @staticmethod # Data ends on 0xFF @staticmethod def decode(data): "Remove unsynchronization bytes from data." return bytes(Unsync.gen_decode(data)) @staticmethod def encode(data): "Insert unsynchronization bytes into data." return bytes(Unsync.gen_encode(data))
staggerpkg/stagger
stagger/conversion.py
Syncsafe.decode
python
def decode(data): "Decodes a syncsafe integer" value = 0 for b in data: if b > 127: # iTunes bug raise ValueError("Invalid syncsafe integer") value <<= 7 value += b return value
Decodes a syncsafe integer
train
https://github.com/staggerpkg/stagger/blob/6530db14afc5d7d8a4599b7f3b26158fb367d786/stagger/conversion.py#L95-L103
null
class Syncsafe: """Conversion to/from syncsafe integers. Syncsafe integers are big-endian 7-bit byte sequences. """ @staticmethod @staticmethod def encode(i, *, width=-1): """Encodes a nonnegative integer into syncsafe format When width > 0, then len(result) == width When width < 0, then len(result) >= abs(width) """ if i < 0: raise ValueError("value is negative") assert width != 0 data = bytearray() while i: data.append(i & 127) i >>= 7 if width > 0 and len(data) > width: raise ValueError("Integer too large") if len(data) < abs(width): data.extend([0] * (abs(width) - len(data))) data.reverse() return data
staggerpkg/stagger
stagger/conversion.py
Syncsafe.encode
python
def encode(i, *, width=-1): if i < 0: raise ValueError("value is negative") assert width != 0 data = bytearray() while i: data.append(i & 127) i >>= 7 if width > 0 and len(data) > width: raise ValueError("Integer too large") if len(data) < abs(width): data.extend([0] * (abs(width) - len(data))) data.reverse() return data
Encodes a nonnegative integer into syncsafe format When width > 0, then len(result) == width When width < 0, then len(result) >= abs(width)
train
https://github.com/staggerpkg/stagger/blob/6530db14afc5d7d8a4599b7f3b26158fb367d786/stagger/conversion.py#L106-L124
null
class Syncsafe: """Conversion to/from syncsafe integers. Syncsafe integers are big-endian 7-bit byte sequences. """ @staticmethod def decode(data): "Decodes a syncsafe integer" value = 0 for b in data: if b > 127: # iTunes bug raise ValueError("Invalid syncsafe integer") value <<= 7 value += b return value @staticmethod
staggerpkg/stagger
stagger/tags.py
detect_tag
python
def detect_tag(filename): with fileutil.opened(filename, "rb") as file: file.seek(0) header = file.read(10) file.seek(0) if len(header) < 10: raise NoTagError("File too short") if header[0:3] != b"ID3": raise NoTagError("ID3v2 tag not found") if header[3] not in _tag_versions or header[4] != 0: raise TagError("Unknown ID3 version: 2.{0}.{1}" .format(*header[3:5])) cls = _tag_versions[header[3]] offset = 0 length = Syncsafe.decode(header[6:10]) + 10 if header[3] == 4 and header[5] & _TAG24_FOOTER: length += 10 return (cls, offset, length)
Return type and position of ID3v2 tag in filename. Returns (tag_class, offset, length), where tag_class is either Tag22, Tag23, or Tag24, and (offset, length) is the position of the tag in the file.
train
https://github.com/staggerpkg/stagger/blob/6530db14afc5d7d8a4599b7f3b26158fb367d786/stagger/tags.py#L95-L117
null
# # tags.py # From the stagger project: http://code.google.com/p/stagger/ # # Copyright (c) 2009-2011 Karoly Lorentey <karoly@lorentey.hu> # All rights reserved. # # Redistribution and use in source and binary forms, with or without # modification, are permitted provided that the following conditions # are met: # # - Redistributions of source code must retain the above copyright # notice, this list of conditions and the following disclaimer. # # - Redistributions in binary form must reproduce the above copyright # notice, this list of conditions and the following disclaimer in # the documentation and/or other materials provided with the # distribution. # # THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS # "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT # LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS # FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE # COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, # INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, # BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; # LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER # CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT # LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN # ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE # POSSIBILITY OF SUCH DAMAGE. import abc import struct import re import collections import io import imghdr import zlib from abc import abstractmethod, abstractproperty from warnings import warn from contextlib import contextmanager from stagger.errors import * from stagger.conversion import * import stagger.frames as Frames import stagger.fileutil as fileutil _FRAME23_FORMAT_COMPRESSED = 0x0080 _FRAME23_FORMAT_ENCRYPTED = 0x0040 _FRAME23_FORMAT_GROUP = 0x0020 _FRAME23_FORMAT_UNKNOWN_MASK = 0x001F _FRAME23_STATUS_DISCARD_ON_TAG_ALTER = 0x8000 _FRAME23_STATUS_DISCARD_ON_FILE_ALTER = 0x4000 _FRAME23_STATUS_READ_ONLY = 0x2000 _FRAME23_STATUS_UNKNOWN_MASK = 0x1F00 _TAG24_UNSYNCHRONISED = 0x80 _TAG24_EXTENDED_HEADER = 0x40 _TAG24_EXPERIMENTAL = 0x20 _TAG24_FOOTER = 0x10 _TAG24_UNKNOWN_MASK = 0x0F _FRAME24_FORMAT_GROUP = 0x0040 _FRAME24_FORMAT_COMPRESSED = 0x0008 _FRAME24_FORMAT_ENCRYPTED = 0x0004 _FRAME24_FORMAT_UNSYNCHRONISED = 0x0002 _FRAME24_FORMAT_DATA_LENGTH_INDICATOR = 0x0001 _FRAME24_FORMAT_UNKNOWN_MASK = 0x00B0 _FRAME24_STATUS_DISCARD_ON_TAG_ALTER = 0x4000 _FRAME24_STATUS_DISCARD_ON_FILE_ALTER = 0x2000 _FRAME24_STATUS_READ_ONLY = 0x1000 _FRAME24_STATUS_UNKNOWN_MASK = 0x8F00 def read_tag(filename): with fileutil.opened(filename, "rb") as file: (cls, offset, length) = detect_tag(file) return cls.read(file, offset) def decode_tag(data): return read_tag(io.BytesIO(data)) def delete_tag(filename): with fileutil.opened(filename, "rb+") as file: try: (cls, offset, length) = detect_tag(file) fileutil.replace_chunk(file, offset, length, bytes()) except NoTagError: pass def frameclass(cls): """Register cls as a class representing an ID3 frame. Sets cls.frameid and cls._version if not present, and registers the new frame in Tag's known_frames dictionary. To be used as a decorator on the class definition: @frameclass class UFID(Frame): _framespec = (NullTerminatedStringSpec("owner"), BinaryDataSpec("data")) """ assert issubclass(cls, Frames.Frame) # Register v2.2 versions of v2.3/v2.4 frames if encoded by inheritance. if len(cls.__name__) == 3: base = cls.__bases__[0] if issubclass(base, Frames.Frame) and base._in_version(3, 4): assert not hasattr(base, "_v2_frame") base._v2_frame = cls # Override frameid from base with v2.2 name if base.frameid == cls.frameid: cls.frameid = cls.__name__ # Add frameid. if not hasattr(cls, "frameid"): cls.frameid = cls.__name__ assert Tag._is_frame_id(cls.frameid) # Supply _version attribute if missing. if len(cls.frameid) == 3: cls._version = 2 if len(cls.frameid) == 4 and not cls._version: cls._version = (3, 4) # Register cls as a known frame. assert cls.frameid not in Tag.known_frames Tag.known_frames[cls.frameid] = cls return cls class FrameOrder: """Order frames based on their position in a predefined list of patterns, and their original position in the source tag. A pattern may be a frame class, or a regular expression that is to be matched against the frame id. >>> order = FrameOrder(TIT1, "T.*", TXXX) >>> order.key(TIT1()) (0, 1) >>> order.key(TPE1()) (1, 1) >>> order.key(TXXX()) (2, 1) >>> order.key(APIC()) (3, 1) >>> order.key(APIC(frameno=3)) (3, 0, 3) """ def __init__(self, *patterns): self.re_keys = [] self.frame_keys = dict() i = -1 for (i, pattern) in zip(range(len(patterns)), patterns): if isinstance(pattern, str): self.re_keys.append((pattern, i)) else: assert issubclass(pattern, Frames.Frame) self.frame_keys[pattern] = i self.unknown_key = i + 1 def key(self, frame): "Return the sort key for the given frame." def keytuple(primary): if frame.frameno is None: return (primary, 1) return (primary, 0, frame.frameno) # Look up frame by exact match if type(frame) in self.frame_keys: return keytuple(self.frame_keys[type(frame)]) # Look up parent frame for v2.2 frames if frame._in_version(2) and type(frame).__bases__[0] in self.frame_keys: return keytuple(self.frame_keys[type(frame).__bases__[0]]) # Try each pattern for (pattern, key) in self.re_keys: if re.match(pattern, frame.frameid): return keytuple(key) return keytuple(self.unknown_key) def __repr__(self): order = [] order.extend((repr(pair[0]), pair[1]) for pair in self.re_keys) order.extend((cls.__name__, self.frame_keys[cls]) for cls in self.frame_keys) order.sort(key=lambda pair: pair[1]) return "<FrameOrder: {0}>".format(", ".join(pair[0] for pair in order)) class Tag(collections.MutableMapping, metaclass=abc.ABCMeta): known_frames = { } # Maps known frameids to Frame class objects frame_order = None # Initialized by stagger.id3 def __init__(self): self.flags = set() self._frames = dict() self._filename = None # Primary accessor (no magic) def frames(self, key=None, orig_order=False): """Returns a list of frames in this tag. If KEY is None, returns all frames in the tag; otherwise returns all frames whose frameid matches KEY. If ORIG_ORDER is True, then the frames are returned in their original order. Otherwise the frames are sorted in canonical order according to the frame_order field of this tag. """ if key is not None: # If there are multiple frames, then they are already in original order. key = self._normalize_key(key) if len(self._frames[key]) == 0: raise KeyError("Key not found: " + repr(key)) return self._frames[key] frames = [] for frameid in self._frames.keys(): for frame in self._frames[frameid]: frames.append(frame) if orig_order: key = (lambda frame: (0, frame.frameno) if frame.frameno is not None else (1,)) else: key = self.frame_order.key frames.sort(key=key) return frames # MutableMapping API def __iter__(self): for frameid in self._frames: yield frameid def __len__(self): return sum(len(self._frames[l]) for l in self._frames) def __eq__(self, other): return (self.version == other.version and self.flags == other.flags and self._frames == other._frames) def _normalize_key(self, key, unknown_ok=True): """Return the normalized version of KEY. KEY may be a frameid (a string), or a Frame class object. If KEY corresponds to a registered frameid, then that frameid is returned. Otherwise, either KeyError is raised, or KEY is returned verbatim, depending on the value of UNKNOWN_OK. """ if Frames.is_frame_class(key): key = key.frameid if isinstance(key, str): if not self._is_frame_id(key): raise KeyError("{0}: Invalid frame id".format(key)) if key not in self.known_frames: if unknown_ok: warn("{0}: Unknown frame id".format(key), UnknownFrameWarning) else: raise KeyError("{0}: Unknown frame id".format(key)) return key # Mapping accessor (with extra magic, for convenience) def __getitem__(self, key): key = self._normalize_key(key) fs = self.frames(key) allow_duplicates = (key not in self.known_frames or self.known_frames[key]._allow_duplicates) if allow_duplicates: return fs if len(fs) > 1: # Merge duplicates into one ephemeral frame, and return that. # This may break users' expectations when they try to make changes # to the attributes of the returned frame; however, I think # sometimes returning a list, sometimes a single frame for the same # frame id would be even worse. fs = fs[0]._merge(fs) assert len(fs) == 1 return fs[0] def __setitem__(self, key, value): key = self._normalize_key(key, unknown_ok=False) if isinstance(value, self.known_frames[key]): self._frames[key] = [value] return if self.known_frames[key]._allow_duplicates: if not isinstance(value, collections.Iterable) or isinstance(value, str): raise ValueError("{0} requires a list of frame values".format(key)) self._frames[key] = [val if isinstance(val, self.known_frames[key]) else self.known_frames[key](val) for val in value] else: # not _allow_duplicates self._frames[key] = [self.known_frames[key](value)] def __delitem__(self, key): del self._frames[self._normalize_key(key)] def values(self): for frameid in self._frames.keys(): for frame in self._frames[frameid]: yield frame # Friendly names API _friendly_names = [ "title", "artist", "date", "album-artist", "album", "track", "track-total", "disc", "disc-total", "grouping", "composer", "genre", "comment", #"compilation", "picture", "sort-title", "sort-artist", "sort-album-artist", "sort-album", "sort-composer", ] title = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) artist = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) date = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) album_artist = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) album = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) track = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) track_total = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) disc = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) disc_total = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) composer = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) genre = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) comment = abstractproperty(fget=lambda self: Non, fset=lambda self, value: None) grouping = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) picture = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) sort_title = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) sort_artist = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) sort_album_artist = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) sort_album = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) sort_composer = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) def __friendly_text_collect(self, frameid): """Collect text values from all instances of FRAMEID into a single list. Returns an empty list if there are no instances of FRAMEID with a text attribute. """ try: return self[frameid].text except (KeyError, AttributeError): return [] @classmethod def _friendly_text_frame(cls, frameid): def getter(self): return " / ".join(self.__friendly_text_collect(frameid)) def setter(self, value): if isinstance(value, str): if len(value): # For non-empty strings, split value self[frameid] = value.split(" / ") elif frameid in self: # For empty strings, delete frame del self[frameid] else: self[frameid] = value return (getter, setter) @classmethod def _friendly_track(cls, frameid, totalattr): def getter(self): ts = self.__friendly_text_collect(frameid) try: return int(ts[0].partition("/")[0]) except (ValueError, IndexError): return 0 def setter(self, value): value = int(value) total = getattr(self, totalattr) if total > 0: self[frameid] = "{0}/{1}".format(value, total) elif value: self[frameid] = str(value) elif frameid in self: del self[frameid] return (getter, setter) @classmethod def _friendly_track_total(cls, frameid, trackattr): def getter(self): ts = self.__friendly_text_collect(frameid) try: return int(ts[0].partition("/")[2]) except (ValueError, IndexError): return 0 def setter(self, value): value = int(value) track = getattr(self, trackattr) if value: self[frameid] = "{0}/{1}".format(track, value) elif track: self[frameid] = str(track) elif frameid in self: del self[frameid] return (getter, setter) __date_pattern = re.compile(r"""(?x)\s* ((?P<year>[0-9]{4}) # YYYY (-(?P<month>[01][0-9]) # -MM (-(?P<day>[0-3][0-9]) # -DD )?)?)? [ T]? ((?P<hour>[0-2][0-9]) # HH (:(?P<min>[0-6][0-9]) # :MM (:(?P<sec>[0-6][0-9]) # :SS )?)?)?\s* """) @classmethod def _validate_friendly_date(cls, string): m = cls.__date_pattern.match(string) if m is None or m.end() != len(string): raise ValueError("date must be in 'YYYY-MM-DD HH:MM:SS' format") @classmethod def _get_friendly_date(cls, string): m = cls.__date_pattern.match(string) if m is None: return (None, None, None, None, None, None) res = [] for field in ("year", "month", "day", "hour", "min", "sec"): v = m.group(field) res.append(int(v) if v is not None else None) return res def _get_date(self, yearframe, dateframe, timeframe): year = month = day = hour = minute = second = None # Parse year. try: year = int(self.__friendly_text_collect(yearframe)[0]) except (IndexError, ValueError): pass # Parse month and date. try: date = self.__friendly_text_collect(dateframe)[0] m = re.match(r"\s*(?P<month>[01][0-9])\s*-?\s*(?P<day>[0-3][0-9])?\s*$", date) if m is not None: month = int(m.group("month")) day = int(m.group("day")) except IndexError: pass # Parse time. try: time = self.__friendly_text_collect(timeframe)[0] m = re.match(r"\s*(?P<hour>[0-2][0-9])\s*:?\s*" "(?P<minute>[0-5][0-9])\s*:?\s*" "(?P<second>[0-5][0-9])?\s*$", time) if m is not None: hour = int(m.group("hour")) minute = int(m.group("minute")) s = m.group("second") second = int(s) if s is not None else None except IndexError: pass return (year, month, day, hour, minute, second) def _friendly_date_string(self, *fields): seps = ("", "-", "-", " ", ":", ":") formats = ("04", "02", "02", "02", "02", "02") res = [] for i in range(len(fields)): if fields[i] is None: break res.append(seps[i]) res.append("{0:{1}}".format(fields[i], formats[i])) return "".join(res) @classmethod def _friendly_picture(cls, frameid): def getter(self): if frameid not in self: return "" else: return ", ".join("{0}:{1}:<{2} bytes of {3} data>" .format(f._spec("type").to_str(f.type), f.desc, len(f.data), imghdr.what(None, f.data[:32])) for f in self[frameid]) def setter(self, value): if len(value) > 0: self[frameid] = [self.known_frames[frameid](value=value)] elif frameid in self: del self[frameid] return (getter, setter) @classmethod def _friendly_comment(cls, frameid): def comment_frame_index(self): if frameid not in self: return None # Return comment with lang="eng", desc="", if present. # Otherwise return the first comment with no description, # regardless of language. icmt = None for i in range(len(self[frameid])): f = self[frameid][i] if f.desc == "": if f.lang == "eng": return i if icmt is None: icmt = i return icmt def getter(self): i = comment_frame_index(self) if i is None: return "" else: return self[frameid][i].text def setter(self, value): assert isinstance(value, str) i = comment_frame_index(self) if i is not None: del self._frames[frameid][i] if len(value) > 0: frame = self.known_frames[frameid](lang="eng", desc="", text=value) if frameid not in self._frames: self._frames[frameid] = [] self._frames[frameid].append(frame) return (getter, setter) # Misc def __repr__(self): return "<{0}: ID3v2.{1} tag{2} with {3} frames>".format( type(self).__name__, self.version, ("({0})".format(", ".join(self.flags)) if len(self.flags) > 0 else ""), len(self._frames)) # Reading tags @classmethod def read(cls, filename, offset=0): """Read an ID3v2 tag from a file.""" i = 0 with fileutil.opened(filename, "rb") as file: file.seek(offset) tag = cls() tag._read_header(file) for (frameid, bflags, data) in tag._read_frames(file): if len(data) == 0: warn("{0}: Ignoring empty frame".format(frameid), EmptyFrameWarning) else: frame = tag._decode_frame(frameid, bflags, data, i) if frame is not None: l = tag._frames.setdefault(frame.frameid, []) l.append(frame) if file.tell() > tag.offset + tag.size: break i += 1 try: tag._filename = file.name except AttributeError: pass return tag @classmethod def decode(cls, data): return cls.read(io.BytesIO(data)) def _decode_frame(self, frameid, bflags, data, frameno=None): try: (flags, data) = self._interpret_frame_flags(frameid, bflags, data) if flags is None: flags = set() if frameid in self.known_frames: return self.known_frames[frameid]._decode(frameid, data, flags, frameno=frameno) else: # Unknown frame flags.add("unknown") warn("{0}: Unknown frame".format(frameid), UnknownFrameWarning) if frameid.startswith('T'): # Unknown text frame return Frames.TextFrame._decode(frameid, data, flags, frameno=frameno) elif frameid.startswith('W'): # Unknown URL frame return Frames.URLFrame._decode(frameid, data, flags, frameno=frameno) else: return Frames.UnknownFrame._decode(frameid, data, flags, frameno=frameno) except (FrameError, ValueError, EOFError) as e: warn("{0}: Invalid frame".format(frameid), ErrorFrameWarning) return Frames.ErrorFrame(frameid, data, exception=e, frameno=frameno) @abstractmethod def _read_header(self, file): pass @abstractmethod def _read_frames(self, file): pass @abstractmethod def _interpret_frame_flags(self, frameid, bflags, data): pass # Writing tags def write(self, filename=None): if not filename: filename = self._filename if not filename: raise TypeError("invalid file: {0}".format(filename)) with fileutil.opened(filename, "rb+") as file: try: (offset, length) = detect_tag(file)[1:3] except NoTagError: (offset, length) = (0, 0) if offset > 0: delete_tag(file) (offset, length) = (0, 0) tag_data = self.encode(size_hint=length) fileutil.replace_chunk(file, offset, length, tag_data) @abstractmethod def encode(self, size_hint=None): pass padding_default = 128 padding_max = 1024 def _get_size_with_padding(self, size_desired, size_actual): size = size_actual if (size_desired is not None and size < size_desired and (self.padding_max is None or size_desired - size_actual <= self.padding_max)): size = size_desired elif self.padding_default: size += min(self.padding_default, self.padding_max) return size @staticmethod def _is_frame_id(data): if isinstance(data, str): try: data = data.encode("ASCII") except UnicodeEncodeError: return false # Allow a single space at end of four-character ids # Some programs (e.g. iTunes 8.2) generate such frames when converting # from 2.2 to 2.3/2.4 tags. pattern = re.compile(b"^[A-Z][A-Z0-9]{2}[A-Z0-9 ]?$") return pattern.match(data) def _prepare_frames_hook(self): pass def _prepare_frames(self): # Generate dictionary of frames d = self._frames # Merge duplicate frames for frameid in self._frames.keys(): fs = self._frames[frameid] if len(fs) > 1: d[frameid] = fs[0]._merge(fs) self._prepare_frames_hook() # Convert frames newframes = [] for frameid in self._frames.keys(): for frame in self._frames[frameid]: try: newframes.append(frame._to_version(self.version)) except IncompatibleFrameError: warn("{0}: Ignoring incompatible frame".format(frameid), FrameWarning) except ValueError as e: warn("{0}: Ignoring invalid frame ({1})".format(frameid, e), FrameWarning) # Sort frames newframes.sort(key=self.frame_order.key) return newframes class Tag22(Tag): version = 2 encodings = ("latin-1", "utf-16") def __init__(self): super().__init__() title = property(*Tag._friendly_text_frame("TT2")) artist = property(*Tag._friendly_text_frame("TP1")) @property def date(self): fields = self._get_date("TYE", "TDA", "TIM") return self._friendly_date_string(*fields) @date.setter def date(self, value): self._validate_friendly_date(value) (year, month, day, hour, minute, second) = self._get_friendly_date(value) for f in "TYE", "TDA", "TIM": if f in self: del self[f] if year is not None: self["TYE"] = "{0:04}".format(year) if month is not None and day is not None: self["TDA"] = "{0:02}{1:02}".format(month, day) if hour is not None and minute is not None: self["TIM"] = "{0:02}{1:02}".format(hour, minute) album_artist = property(*Tag._friendly_text_frame("TP2")) album = property(*Tag._friendly_text_frame("TAL")) track = property(*Tag._friendly_track("TRK", "track_total")) track_total = property(*Tag._friendly_track_total("TRK", "track")) disc = property(*Tag._friendly_track("TPA", "disc_total")) disc_total = property(*Tag._friendly_track_total("TPA", "disc")) composer = property(*Tag._friendly_text_frame("TCM")) genre = property(*Tag._friendly_text_frame("TCO")) comment = property(*Tag._friendly_comment("COM")) grouping = property(*Tag._friendly_text_frame("TT1")) # TODO: compilation picture = property(*Tag._friendly_picture("PIC")) sort_title = property(*Tag._friendly_text_frame("TST")) sort_artist = property(*Tag._friendly_text_frame("TSP")) sort_album_artist = property(*Tag._friendly_text_frame("TS2")) sort_album = property(*Tag._friendly_text_frame("TSA")) sort_composer = property(*Tag._friendly_text_frame("TSC")) def _read_header(self, file): self.offset = file.tell() header = fileutil.xread(file, 10) if header[0:5] != b"ID3\x02\00": raise NoTagError("ID3v2.2 header not found") if header[5] & 0x80: self.flags.add("unsynchronised") if header[5] & 0x40: # Compression bit is ill-defined in standard raise TagError("ID3v2.2 tag compression is not supported") if header[5] & 0x3F: warn("Unknown ID3v2.2 flags", TagWarning) self.size = Syncsafe.decode(header[6:10]) + 10 def _read_frames(self, file): if "unsynchronised" in self.flags: ufile = UnsyncReader(file) else: ufile = file while file.tell() < self.offset + self.size: header = fileutil.xread(ufile, 6) if not self._is_frame_id(header[0:3]): break frameid = header[0:3].decode("ASCII") size = Int8.decode(header[3:6]) data = fileutil.xread(ufile, size) yield (frameid, None, data) def _interpret_frame_flags(self, frameid, bflags, data): # No frame flags in v2.2 return (None, data) def __encode_one_frame(self, frame): framedata = frame._encode(encodings=self.encodings) data = bytearray() # Frame id if len(frame.frameid) != 3 or not self._is_frame_id(frame.frameid): raise "Invalid ID3v2.2 frame id {0}".format(repr(frame.frameid)) data.extend(frame.frameid.encode("ASCII")) # Size data.extend(Int8.encode(len(framedata), width=3)) assert(len(data) == 6) data.extend(framedata) return data def _prepare_frames_hook(self): for frameid in self._frames.keys(): for frame in self._frames[frameid]: if isinstance(frame, Frames.TextFrame): # ID3v2.2 doesn't support multiple text values if len(frame.text) > 1: warn("{0}: merged multiple text strings into one value" .format(frame.frameid), FrameWarning) frame.text = [" / ".join(frame.text)] def encode(self, size_hint=None): if len(self) == 0: # No frames -> no tag return b"" frames = self._prepare_frames() framedata = bytearray().join(self.__encode_one_frame(frame) for frame in frames) if "unsynchronised" in self.flags: framedata = Unsync.encode(framedata) size = self._get_size_with_padding(size_hint, len(framedata) + 10) data = bytearray() data.extend(b"ID3\x02\x00") data.append(0x80 if "unsynchronised" in self.flags else 0x00) data.extend(Syncsafe.encode(size - 10, width=4)) assert len(data) == 10 data.extend(framedata) if size > len(data): data.extend(b"\x00" * (size - len(data))) assert len(data) == size return data class Tag23(Tag): version = 3 encodings = ("latin-1", "utf-16") def __init__(self): super().__init__() title = property(*Tag._friendly_text_frame("TIT2")) artist = property(*Tag._friendly_text_frame("TPE1")) @property def date(self): fields = self._get_date("TYER", "TDAT", "TIME") return self._friendly_date_string(*fields) @date.setter def date(self, value): self._validate_friendly_date(value) (year, month, day, hour, minute, second) = self._get_friendly_date(value) for f in "TYER", "TDAT", "TIME": if f in self: del self[f] if year is not None: self["TYER"] = "{0:04}".format(year) if month is not None and day is not None: self["TDAT"] = "{0:02}{1:02}".format(month, day) if hour is not None and minute is not None: self["TIME"] = "{0:02}{1:02}".format(hour, minute) album_artist = property(*Tag._friendly_text_frame("TPE2")) album = property(*Tag._friendly_text_frame("TALB")) track = property(*Tag._friendly_track("TRCK", "track_total")) track_total = property(*Tag._friendly_track_total("TRCK", "track")) disc = property(*Tag._friendly_track("TPOS", "disc_total")) disc_total = property(*Tag._friendly_track_total("TPOS", "disc")) composer = property(*Tag._friendly_text_frame("TCOM")) genre = property(*Tag._friendly_text_frame("TCON")) comment = property(*Tag._friendly_comment("COMM")) grouping = property(*Tag._friendly_text_frame("TIT1")) # TODO: compilation picture = property(*Tag._friendly_picture("APIC")) sort_title = property(*Tag._friendly_text_frame("TSOT")) sort_artist = property(*Tag._friendly_text_frame("TSOP")) sort_album_artist = property(*Tag._friendly_text_frame("TSO2")) sort_album = property(*Tag._friendly_text_frame("TSOA")) sort_composer = property(*Tag._friendly_text_frame("TSOC")) def _read_header(self, file): self.offset = file.tell() header = fileutil.xread(file, 10) if header[0:5] != b"ID3\x03\x00": raise NoTagError("ID3v2.3 header not found") if header[5] & 0x80: self.flags.add("unsynchronised") if header[5] & 0x40: self.flags.add("extended_header") if header[5] & 0x20: self.flags.add("experimental") if header[5] & 0x1F: warn("Unknown ID3v2.3 flags", TagWarning) self.size = Syncsafe.decode(header[6:10]) + 10 if "extended_header" in self.flags: self.__read_extended_header(file) def __read_extended_header(self, file): (size, ext_flags, self.padding_size) = \ struct.unpack("!IHI", fileutil.xread(file, 10)) if size != 6 and size != 10: warn("Unexpected size of ID3v2.3 extended header: {0}".format(size), TagWarning) if ext_flags & 32768: if size < 10: warn("Extended header is too short for a CRC field: {0} bytes instead of 10".format(size), TagWarning) else: self.flags.add("ext:crc_present") (self.crc32,) = struct.unpack("!I", fileutil.xread(file, 4)) size -= 6 if size > 6: fileutil.xread(file, size - 6) def _read_frames(self, file): if "unsynchronised" in self.flags: ufile = UnsyncReader(file) else: ufile = file while file.tell() < self.offset + self.size: header = fileutil.xread(ufile, 10) if not self._is_frame_id(header[0:4]): break frameid = header[0:4].decode("ASCII") size = Int8.decode(header[4:8]) bflags = Int8.decode(header[8:10]) data = fileutil.xread(ufile, size) yield (frameid, bflags, data) def _interpret_frame_flags(self, frameid, bflags, data): flags = set() # Frame encoding flags if bflags & _FRAME23_FORMAT_UNKNOWN_MASK: raise FrameError("{0}: Invalid ID3v2.3 frame encoding flags: 0x{0:X}".format(frameid, bflags)) if bflags & _FRAME23_FORMAT_COMPRESSED: flags.add("compressed") expanded_size = Int8.decode(data[0:4]) data = zlib.decompress(data[4:]) if bflags & _FRAME23_FORMAT_ENCRYPTED: raise FrameError("{0}: Can't read ID3v2.3 encrypted frames".format(frameid)) if bflags & _FRAME23_FORMAT_GROUP: flags.add("group") flags.add("group={0}".format(data[0])) # Hack data = data[1:] # Frame status messages if bflags & _FRAME23_STATUS_DISCARD_ON_TAG_ALTER: flags.add("discard_on_tag_alter") if bflags & _FRAME23_STATUS_DISCARD_ON_FILE_ALTER: flags.add("discard_on_file_alter") if bflags & _FRAME23_STATUS_READ_ONLY: flags.add("read_only") if bflags & _FRAME23_STATUS_UNKNOWN_MASK: warn("{0}: Unexpected ID3v2.3 frame status flags: 0x{1:X}".format(frameid, bflags), TagWarning) return flags, data def __encode_one_frame(self, frame): framedata = frame._encode(encodings=self.encodings) origlen = len(framedata) flagval = 0 frameinfo = bytearray() if "compressed" in frame.flags: framedata = zlib.compress(framedata) flagval |= _FRAME23_FORMAT_COMPRESSED frameinfo.extend(Int8.encode(origlen, width=4)) if "group" in frame.flags: grp = 0 for flag in frame.flags: if flag.startswith("group="): grp = int(flag[6:]) frameinfo.append(grp) flagval |= _FRAME23_FORMAT_GROUP if "discard_on_tag_alter" in frame.flags: flagval |= _FRAME23_STATUS_DISCARD_ON_TAG_ALTER if "discard_on_file_alter" in frame.flags: flagval |= _FRAME23_STATUS_DISCARD_ON_FILE_ALTER if "read_only" in frame.flags: flagval |= _FRAME23_STATUS_READ_ONLY data = bytearray() # Frame id if len(frame.frameid) != 4 or not self._is_frame_id(frame.frameid.encode("ASCII")): raise ValueError("Invalid ID3v2.3 frame id {0}".format(repr(frame.frameid))) data.extend(frame.frameid.encode("ASCII")) # Size data.extend(Int8.encode(len(frameinfo) + len(framedata), width=4)) # Flags data.extend(Int8.encode(flagval, width=2)) assert len(data) == 10 # Format info data.extend(frameinfo) # Frame data data.extend(framedata) return data def encode(self, size_hint=None): if len(self) == 0: # No frames -> no tag return b"" frames = self._prepare_frames() framedata = bytearray().join(self.__encode_one_frame(frame) for frame in frames) if "unsynchronised" in self.flags: framedata = Unsync.encode(framedata) size = self._get_size_with_padding(size_hint, len(framedata) + 10) data = bytearray() data.extend(b"ID3\x03\x00") flagval = 0x00 if "unsynchronised" in self.flags: flagval |= 0x80 data.append(flagval) data.extend(Syncsafe.encode(size - 10, width=4)) assert len(data) == 10 data.extend(framedata) if size > len(data): data.extend(b"\x00" * (size - len(data))) assert len(data) == size return data class Tag24(Tag): ITUNES_WORKAROUND = False version = 4 encodings = ("latin-1", "utf-8") def __init__(self): super().__init__() title = property(*Tag._friendly_text_frame("TIT2")) artist = property(*Tag._friendly_text_frame("TPE1")) @property def date(self): try: frame = self["TDRC"] except KeyError: return "" else: return frame.text[0] @date.setter def date(self, value): self._validate_friendly_date(value) fields = self._get_friendly_date(value) val = self._friendly_date_string(*fields) if val: self["TDRC"] = val elif "TDRC" in self: del self["TDRC"] album = property(*Tag._friendly_text_frame("TALB")) album_artist = property(*Tag._friendly_text_frame("TPE2")) track = property(*Tag._friendly_track("TRCK", "track_total")) track_total = property(*Tag._friendly_track_total("TRCK", "track")) disc = property(*Tag._friendly_track("TPOS", "disc_total")) disc_total = property(*Tag._friendly_track_total("TPOS", "disc")) composer = property(*Tag._friendly_text_frame("TCOM")) genre = property(*Tag._friendly_text_frame("TCON")) comment = property(*Tag._friendly_comment("COMM")) grouping = property(*Tag._friendly_text_frame("TIT1")) # TODO: compilation picture = property(*Tag._friendly_picture("APIC")) sort_title = property(*Tag._friendly_text_frame("TSOT")) sort_artist = property(*Tag._friendly_text_frame("TSOP")) sort_album_artist = property(*Tag._friendly_text_frame("TSO2")) sort_album = property(*Tag._friendly_text_frame("TSOA")) sort_composer = property(*Tag._friendly_text_frame("TSOC")) def _read_header(self, file): self.offset = file.tell() header = fileutil.xread(file, 10) if header[0:5] != b"ID3\x04\x00": raise NoTagError("ID3v2 header not found") if header[5] & _TAG24_UNSYNCHRONISED: self.flags.add("unsynchronised") if header[5] & _TAG24_EXTENDED_HEADER: self.flags.add("extended_header") if header[5] & _TAG24_EXPERIMENTAL: self.flags.add("experimental") if header[5] & _TAG24_FOOTER: self.flags.add("footer") if header[5] & _TAG24_UNKNOWN_MASK: warn("Unknown ID3v2.4 flags", TagWarning) self.size = (Syncsafe.decode(header[6:10]) + 10 + (10 if "footer" in self.flags else 0)) if "extended_header" in self.flags: self.__read_extended_header(file) def __read_extended_header_flag_data(self, data): # 1-byte length + data length = data[0] if length & 128: raise TagError("Invalid size of extended header field") return (data[1:1+length], data[1+length:]) def __read_extended_header(self, file): size = Syncsafe.decode(fileutil.xread(file, 4)) if size < 6: warn("Unexpected size of ID3v2.4 extended header: {0}".format(size), TagWarning) data = fileutil.xread(file, size - 4) numflags = data[0] if numflags != 1: warn("Unexpected number of ID3v2.4 extended flag bytes: {0}" .format(numflags), TagWarning) flags = data[1] data = data[1+numflags:] if flags & 0x40: self.flags.add("ext:update") (dummy, data) = self.__read_extended_header_flag_data(data) if flags & 0x20: self.flags.add("ext:crc_present") (self.crc32, data) = self.__read_extended_header_flag_data(data) self.crc32 = Syncsafe.decode(self.crc32) if flags & 0x10: self.flags.add("ext:restrictions") (self.restrictions, data) = self.__read_extended_header_flag_data(data) def _read_frames(self, file, syncsafe_workaround = None): # Older versions of iTunes stored frame sizes as straight 8bit integers, # not syncsafe values as the spec requires. # (The bug is known to be fixed in iTunes 8.2.) # # To work around such an erroneous encoding, we re-read the entire tag # in non-syncsafe mode when we detect a frame with a bad size. # This heuristic does not detect all badly encoded tags; # it fails when the 8-bit frame size happens to be in syncsafe format. # # We could improve detection by parsing the tag both ways and see which # interpretation produces more frames. However, the extra effort doesn't # seem worthwhile to do by default. # # If you have many files with iTunes-encoded tags, you can force stagger # to read them in non-syncsafe mode setting the ITUNES_WORKAROUND # class attribute to True and let stagger reencode your tags. (Stagger # will never produce a 2.4 tag with non-syncsafe frame lengths.) if syncsafe_workaround is None: syncsafe_workaround = self.ITUNES_WORKAROUND origfpos = file.tell() frames = [] while file.tell() < self.offset + self.size: header = fileutil.xread(file, 10) if not self._is_frame_id(header[0:4]): break frameid = header[0:4].decode("ASCII") if syncsafe_workaround: size = Int8.decode(header[4:8]) else: try: size = Syncsafe.decode(header[4:8]) except ValueError: if syncsafe_workaround: raise warn("Invalid syncsafe frame size; switching to 8-bit mode") file.seek(origfpos) return self._read_frames(file, True) bflags = Int8.decode(header[8:10]) data = fileutil.xread(file, size) frames.append((frameid, bflags, data)) return frames def _interpret_frame_flags(self, frameid, bflags, data): flags = set() # Frame format flags if bflags & _FRAME24_FORMAT_UNKNOWN_MASK: raise FrameError("{0}: Unknown frame encoding flags: 0x{1:X}".format(frameid, bflags)) if bflags & _FRAME24_FORMAT_GROUP: flags.add("group") flags.add("group={0}".format(data[0])) # hack data = data[1:] if bflags & _FRAME24_FORMAT_COMPRESSED: flags.add("compressed") if bflags & _FRAME24_FORMAT_ENCRYPTED: raise FrameError("{0}: Can't read encrypted frames".format(frameid)) if bflags & _FRAME24_FORMAT_UNSYNCHRONISED: flags.add("unsynchronised") expanded_size = len(data) if bflags & _FRAME24_FORMAT_DATA_LENGTH_INDICATOR: flags.add("data_length_indicator") expanded_size = Syncsafe.decode(data[0:4]) data = data[4:] if "unsynchronised" in self.flags: data = Unsync.decode(data) if "compressed" in self.flags: data = zlib.decompress(data) # Frame status flags if bflags & _FRAME24_STATUS_DISCARD_ON_TAG_ALTER: flags.add("discard_on_tag_alter") if bflags & _FRAME24_STATUS_DISCARD_ON_FILE_ALTER: flags.add("discard_on_file_alter") if bflags & _FRAME24_STATUS_READ_ONLY: flags.add("read_only") if bflags & _FRAME24_STATUS_UNKNOWN_MASK: warn("{0}: Unexpected status flags: 0x{1:X}" .format(frameid, bflags), FrameWarning) return flags, data def __encode_one_frame(self, frame): framedata = frame._encode(encodings=self.encodings) origlen = len(framedata) flagval = 0 frameinfo = bytearray() if "group" in frame.flags: grp = 0 for flag in frame.flags: if flag.startswith("group="): grp = int(flag[6:]) frameinfo.append(grp) flagval |= _FRAME24_FORMAT_GROUP if "compressed" in frame.flags: frame.flags.add("data_length_indicator") framedata = zlib.compress(framedata) flagval |= _FRAME24_FORMAT_COMPRESSED if "unsynchronised" in frame.flags: frame.flags.add("data_length_indicator") framedata = Unsync.encode(framedata) flagval |= _FRAME24_FORMAT_UNSYNCHRONISED if "data_length_indicator" in frame.flags: frameinfo.extend(Syncsafe.encode(origlen, width=4)) flagval |= _FRAME24_FORMAT_DATA_LENGTH_INDICATOR if "discard_on_tag_alter" in frame.flags: flagval |= _FRAME24_STATUS_DISCARD_ON_TAG_ALTER if "discard_on_file_alter" in frame.flags: flagval |= _FRAME24_STATUS_DISCARD_ON_FILE_ALTER if "read_only" in frame.flags: flagval |= _FRAME24_STATUS_READ_ONLY data = bytearray() # Frame id if len(frame.frameid) != 4 or not self._is_frame_id(frame.frameid): raise "{0}: Invalid frame id".format(repr(frame.frameid)) data.extend(frame.frameid.encode("ASCII")) # Size data.extend(Syncsafe.encode(len(frameinfo) + len(framedata), width=4)) # Flags data.extend(Int8.encode(flagval, width=2)) assert len(data) == 10 # Format info data.extend(frameinfo) # Frame data data.extend(framedata) return data def encode(self, size_hint=None): if len(self) == 0: # No frames -> no tag return b"" frames = self._prepare_frames() if "unsynchronised" in self.flags: for frame in frames: frame.flags.add("unsynchronised") framedata = bytearray().join(self.__encode_one_frame(frame) for frame in frames) size = self._get_size_with_padding(size_hint, len(framedata) + 10) data = bytearray() data.extend(b"ID3\x04\x00") flagval = 0x00 if "unsynchronised" in self.flags: flagval |= 0x80 data.append(flagval) data.extend(Syncsafe.encode(size - 10, width=4)) assert len(data) == 10 data.extend(framedata) if size > len(data): data.extend(b"\x00" * (size - len(data))) assert len(data) == size return data _tag_versions = { 2: Tag22, 3: Tag23, 4: Tag24, }
staggerpkg/stagger
stagger/tags.py
frameclass
python
def frameclass(cls): assert issubclass(cls, Frames.Frame) # Register v2.2 versions of v2.3/v2.4 frames if encoded by inheritance. if len(cls.__name__) == 3: base = cls.__bases__[0] if issubclass(base, Frames.Frame) and base._in_version(3, 4): assert not hasattr(base, "_v2_frame") base._v2_frame = cls # Override frameid from base with v2.2 name if base.frameid == cls.frameid: cls.frameid = cls.__name__ # Add frameid. if not hasattr(cls, "frameid"): cls.frameid = cls.__name__ assert Tag._is_frame_id(cls.frameid) # Supply _version attribute if missing. if len(cls.frameid) == 3: cls._version = 2 if len(cls.frameid) == 4 and not cls._version: cls._version = (3, 4) # Register cls as a known frame. assert cls.frameid not in Tag.known_frames Tag.known_frames[cls.frameid] = cls return cls
Register cls as a class representing an ID3 frame. Sets cls.frameid and cls._version if not present, and registers the new frame in Tag's known_frames dictionary. To be used as a decorator on the class definition: @frameclass class UFID(Frame): _framespec = (NullTerminatedStringSpec("owner"), BinaryDataSpec("data"))
train
https://github.com/staggerpkg/stagger/blob/6530db14afc5d7d8a4599b7f3b26158fb367d786/stagger/tags.py#L119-L158
null
# # tags.py # From the stagger project: http://code.google.com/p/stagger/ # # Copyright (c) 2009-2011 Karoly Lorentey <karoly@lorentey.hu> # All rights reserved. # # Redistribution and use in source and binary forms, with or without # modification, are permitted provided that the following conditions # are met: # # - Redistributions of source code must retain the above copyright # notice, this list of conditions and the following disclaimer. # # - Redistributions in binary form must reproduce the above copyright # notice, this list of conditions and the following disclaimer in # the documentation and/or other materials provided with the # distribution. # # THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS # "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT # LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS # FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE # COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, # INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, # BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; # LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER # CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT # LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN # ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE # POSSIBILITY OF SUCH DAMAGE. import abc import struct import re import collections import io import imghdr import zlib from abc import abstractmethod, abstractproperty from warnings import warn from contextlib import contextmanager from stagger.errors import * from stagger.conversion import * import stagger.frames as Frames import stagger.fileutil as fileutil _FRAME23_FORMAT_COMPRESSED = 0x0080 _FRAME23_FORMAT_ENCRYPTED = 0x0040 _FRAME23_FORMAT_GROUP = 0x0020 _FRAME23_FORMAT_UNKNOWN_MASK = 0x001F _FRAME23_STATUS_DISCARD_ON_TAG_ALTER = 0x8000 _FRAME23_STATUS_DISCARD_ON_FILE_ALTER = 0x4000 _FRAME23_STATUS_READ_ONLY = 0x2000 _FRAME23_STATUS_UNKNOWN_MASK = 0x1F00 _TAG24_UNSYNCHRONISED = 0x80 _TAG24_EXTENDED_HEADER = 0x40 _TAG24_EXPERIMENTAL = 0x20 _TAG24_FOOTER = 0x10 _TAG24_UNKNOWN_MASK = 0x0F _FRAME24_FORMAT_GROUP = 0x0040 _FRAME24_FORMAT_COMPRESSED = 0x0008 _FRAME24_FORMAT_ENCRYPTED = 0x0004 _FRAME24_FORMAT_UNSYNCHRONISED = 0x0002 _FRAME24_FORMAT_DATA_LENGTH_INDICATOR = 0x0001 _FRAME24_FORMAT_UNKNOWN_MASK = 0x00B0 _FRAME24_STATUS_DISCARD_ON_TAG_ALTER = 0x4000 _FRAME24_STATUS_DISCARD_ON_FILE_ALTER = 0x2000 _FRAME24_STATUS_READ_ONLY = 0x1000 _FRAME24_STATUS_UNKNOWN_MASK = 0x8F00 def read_tag(filename): with fileutil.opened(filename, "rb") as file: (cls, offset, length) = detect_tag(file) return cls.read(file, offset) def decode_tag(data): return read_tag(io.BytesIO(data)) def delete_tag(filename): with fileutil.opened(filename, "rb+") as file: try: (cls, offset, length) = detect_tag(file) fileutil.replace_chunk(file, offset, length, bytes()) except NoTagError: pass def detect_tag(filename): """Return type and position of ID3v2 tag in filename. Returns (tag_class, offset, length), where tag_class is either Tag22, Tag23, or Tag24, and (offset, length) is the position of the tag in the file. """ with fileutil.opened(filename, "rb") as file: file.seek(0) header = file.read(10) file.seek(0) if len(header) < 10: raise NoTagError("File too short") if header[0:3] != b"ID3": raise NoTagError("ID3v2 tag not found") if header[3] not in _tag_versions or header[4] != 0: raise TagError("Unknown ID3 version: 2.{0}.{1}" .format(*header[3:5])) cls = _tag_versions[header[3]] offset = 0 length = Syncsafe.decode(header[6:10]) + 10 if header[3] == 4 and header[5] & _TAG24_FOOTER: length += 10 return (cls, offset, length) def frameclass(cls): """Register cls as a class representing an ID3 frame. Sets cls.frameid and cls._version if not present, and registers the new frame in Tag's known_frames dictionary. To be used as a decorator on the class definition: @frameclass class UFID(Frame): _framespec = (NullTerminatedStringSpec("owner"), BinaryDataSpec("data")) """ assert issubclass(cls, Frames.Frame) # Register v2.2 versions of v2.3/v2.4 frames if encoded by inheritance. if len(cls.__name__) == 3: base = cls.__bases__[0] if issubclass(base, Frames.Frame) and base._in_version(3, 4): assert not hasattr(base, "_v2_frame") base._v2_frame = cls # Override frameid from base with v2.2 name if base.frameid == cls.frameid: cls.frameid = cls.__name__ # Add frameid. if not hasattr(cls, "frameid"): cls.frameid = cls.__name__ assert Tag._is_frame_id(cls.frameid) # Supply _version attribute if missing. if len(cls.frameid) == 3: cls._version = 2 if len(cls.frameid) == 4 and not cls._version: cls._version = (3, 4) # Register cls as a known frame. assert cls.frameid not in Tag.known_frames Tag.known_frames[cls.frameid] = cls return cls class FrameOrder: """Order frames based on their position in a predefined list of patterns, and their original position in the source tag. A pattern may be a frame class, or a regular expression that is to be matched against the frame id. >>> order = FrameOrder(TIT1, "T.*", TXXX) >>> order.key(TIT1()) (0, 1) >>> order.key(TPE1()) (1, 1) >>> order.key(TXXX()) (2, 1) >>> order.key(APIC()) (3, 1) >>> order.key(APIC(frameno=3)) (3, 0, 3) """ def __init__(self, *patterns): self.re_keys = [] self.frame_keys = dict() i = -1 for (i, pattern) in zip(range(len(patterns)), patterns): if isinstance(pattern, str): self.re_keys.append((pattern, i)) else: assert issubclass(pattern, Frames.Frame) self.frame_keys[pattern] = i self.unknown_key = i + 1 def key(self, frame): "Return the sort key for the given frame." def keytuple(primary): if frame.frameno is None: return (primary, 1) return (primary, 0, frame.frameno) # Look up frame by exact match if type(frame) in self.frame_keys: return keytuple(self.frame_keys[type(frame)]) # Look up parent frame for v2.2 frames if frame._in_version(2) and type(frame).__bases__[0] in self.frame_keys: return keytuple(self.frame_keys[type(frame).__bases__[0]]) # Try each pattern for (pattern, key) in self.re_keys: if re.match(pattern, frame.frameid): return keytuple(key) return keytuple(self.unknown_key) def __repr__(self): order = [] order.extend((repr(pair[0]), pair[1]) for pair in self.re_keys) order.extend((cls.__name__, self.frame_keys[cls]) for cls in self.frame_keys) order.sort(key=lambda pair: pair[1]) return "<FrameOrder: {0}>".format(", ".join(pair[0] for pair in order)) class Tag(collections.MutableMapping, metaclass=abc.ABCMeta): known_frames = { } # Maps known frameids to Frame class objects frame_order = None # Initialized by stagger.id3 def __init__(self): self.flags = set() self._frames = dict() self._filename = None # Primary accessor (no magic) def frames(self, key=None, orig_order=False): """Returns a list of frames in this tag. If KEY is None, returns all frames in the tag; otherwise returns all frames whose frameid matches KEY. If ORIG_ORDER is True, then the frames are returned in their original order. Otherwise the frames are sorted in canonical order according to the frame_order field of this tag. """ if key is not None: # If there are multiple frames, then they are already in original order. key = self._normalize_key(key) if len(self._frames[key]) == 0: raise KeyError("Key not found: " + repr(key)) return self._frames[key] frames = [] for frameid in self._frames.keys(): for frame in self._frames[frameid]: frames.append(frame) if orig_order: key = (lambda frame: (0, frame.frameno) if frame.frameno is not None else (1,)) else: key = self.frame_order.key frames.sort(key=key) return frames # MutableMapping API def __iter__(self): for frameid in self._frames: yield frameid def __len__(self): return sum(len(self._frames[l]) for l in self._frames) def __eq__(self, other): return (self.version == other.version and self.flags == other.flags and self._frames == other._frames) def _normalize_key(self, key, unknown_ok=True): """Return the normalized version of KEY. KEY may be a frameid (a string), or a Frame class object. If KEY corresponds to a registered frameid, then that frameid is returned. Otherwise, either KeyError is raised, or KEY is returned verbatim, depending on the value of UNKNOWN_OK. """ if Frames.is_frame_class(key): key = key.frameid if isinstance(key, str): if not self._is_frame_id(key): raise KeyError("{0}: Invalid frame id".format(key)) if key not in self.known_frames: if unknown_ok: warn("{0}: Unknown frame id".format(key), UnknownFrameWarning) else: raise KeyError("{0}: Unknown frame id".format(key)) return key # Mapping accessor (with extra magic, for convenience) def __getitem__(self, key): key = self._normalize_key(key) fs = self.frames(key) allow_duplicates = (key not in self.known_frames or self.known_frames[key]._allow_duplicates) if allow_duplicates: return fs if len(fs) > 1: # Merge duplicates into one ephemeral frame, and return that. # This may break users' expectations when they try to make changes # to the attributes of the returned frame; however, I think # sometimes returning a list, sometimes a single frame for the same # frame id would be even worse. fs = fs[0]._merge(fs) assert len(fs) == 1 return fs[0] def __setitem__(self, key, value): key = self._normalize_key(key, unknown_ok=False) if isinstance(value, self.known_frames[key]): self._frames[key] = [value] return if self.known_frames[key]._allow_duplicates: if not isinstance(value, collections.Iterable) or isinstance(value, str): raise ValueError("{0} requires a list of frame values".format(key)) self._frames[key] = [val if isinstance(val, self.known_frames[key]) else self.known_frames[key](val) for val in value] else: # not _allow_duplicates self._frames[key] = [self.known_frames[key](value)] def __delitem__(self, key): del self._frames[self._normalize_key(key)] def values(self): for frameid in self._frames.keys(): for frame in self._frames[frameid]: yield frame # Friendly names API _friendly_names = [ "title", "artist", "date", "album-artist", "album", "track", "track-total", "disc", "disc-total", "grouping", "composer", "genre", "comment", #"compilation", "picture", "sort-title", "sort-artist", "sort-album-artist", "sort-album", "sort-composer", ] title = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) artist = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) date = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) album_artist = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) album = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) track = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) track_total = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) disc = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) disc_total = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) composer = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) genre = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) comment = abstractproperty(fget=lambda self: Non, fset=lambda self, value: None) grouping = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) picture = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) sort_title = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) sort_artist = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) sort_album_artist = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) sort_album = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) sort_composer = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) def __friendly_text_collect(self, frameid): """Collect text values from all instances of FRAMEID into a single list. Returns an empty list if there are no instances of FRAMEID with a text attribute. """ try: return self[frameid].text except (KeyError, AttributeError): return [] @classmethod def _friendly_text_frame(cls, frameid): def getter(self): return " / ".join(self.__friendly_text_collect(frameid)) def setter(self, value): if isinstance(value, str): if len(value): # For non-empty strings, split value self[frameid] = value.split(" / ") elif frameid in self: # For empty strings, delete frame del self[frameid] else: self[frameid] = value return (getter, setter) @classmethod def _friendly_track(cls, frameid, totalattr): def getter(self): ts = self.__friendly_text_collect(frameid) try: return int(ts[0].partition("/")[0]) except (ValueError, IndexError): return 0 def setter(self, value): value = int(value) total = getattr(self, totalattr) if total > 0: self[frameid] = "{0}/{1}".format(value, total) elif value: self[frameid] = str(value) elif frameid in self: del self[frameid] return (getter, setter) @classmethod def _friendly_track_total(cls, frameid, trackattr): def getter(self): ts = self.__friendly_text_collect(frameid) try: return int(ts[0].partition("/")[2]) except (ValueError, IndexError): return 0 def setter(self, value): value = int(value) track = getattr(self, trackattr) if value: self[frameid] = "{0}/{1}".format(track, value) elif track: self[frameid] = str(track) elif frameid in self: del self[frameid] return (getter, setter) __date_pattern = re.compile(r"""(?x)\s* ((?P<year>[0-9]{4}) # YYYY (-(?P<month>[01][0-9]) # -MM (-(?P<day>[0-3][0-9]) # -DD )?)?)? [ T]? ((?P<hour>[0-2][0-9]) # HH (:(?P<min>[0-6][0-9]) # :MM (:(?P<sec>[0-6][0-9]) # :SS )?)?)?\s* """) @classmethod def _validate_friendly_date(cls, string): m = cls.__date_pattern.match(string) if m is None or m.end() != len(string): raise ValueError("date must be in 'YYYY-MM-DD HH:MM:SS' format") @classmethod def _get_friendly_date(cls, string): m = cls.__date_pattern.match(string) if m is None: return (None, None, None, None, None, None) res = [] for field in ("year", "month", "day", "hour", "min", "sec"): v = m.group(field) res.append(int(v) if v is not None else None) return res def _get_date(self, yearframe, dateframe, timeframe): year = month = day = hour = minute = second = None # Parse year. try: year = int(self.__friendly_text_collect(yearframe)[0]) except (IndexError, ValueError): pass # Parse month and date. try: date = self.__friendly_text_collect(dateframe)[0] m = re.match(r"\s*(?P<month>[01][0-9])\s*-?\s*(?P<day>[0-3][0-9])?\s*$", date) if m is not None: month = int(m.group("month")) day = int(m.group("day")) except IndexError: pass # Parse time. try: time = self.__friendly_text_collect(timeframe)[0] m = re.match(r"\s*(?P<hour>[0-2][0-9])\s*:?\s*" "(?P<minute>[0-5][0-9])\s*:?\s*" "(?P<second>[0-5][0-9])?\s*$", time) if m is not None: hour = int(m.group("hour")) minute = int(m.group("minute")) s = m.group("second") second = int(s) if s is not None else None except IndexError: pass return (year, month, day, hour, minute, second) def _friendly_date_string(self, *fields): seps = ("", "-", "-", " ", ":", ":") formats = ("04", "02", "02", "02", "02", "02") res = [] for i in range(len(fields)): if fields[i] is None: break res.append(seps[i]) res.append("{0:{1}}".format(fields[i], formats[i])) return "".join(res) @classmethod def _friendly_picture(cls, frameid): def getter(self): if frameid not in self: return "" else: return ", ".join("{0}:{1}:<{2} bytes of {3} data>" .format(f._spec("type").to_str(f.type), f.desc, len(f.data), imghdr.what(None, f.data[:32])) for f in self[frameid]) def setter(self, value): if len(value) > 0: self[frameid] = [self.known_frames[frameid](value=value)] elif frameid in self: del self[frameid] return (getter, setter) @classmethod def _friendly_comment(cls, frameid): def comment_frame_index(self): if frameid not in self: return None # Return comment with lang="eng", desc="", if present. # Otherwise return the first comment with no description, # regardless of language. icmt = None for i in range(len(self[frameid])): f = self[frameid][i] if f.desc == "": if f.lang == "eng": return i if icmt is None: icmt = i return icmt def getter(self): i = comment_frame_index(self) if i is None: return "" else: return self[frameid][i].text def setter(self, value): assert isinstance(value, str) i = comment_frame_index(self) if i is not None: del self._frames[frameid][i] if len(value) > 0: frame = self.known_frames[frameid](lang="eng", desc="", text=value) if frameid not in self._frames: self._frames[frameid] = [] self._frames[frameid].append(frame) return (getter, setter) # Misc def __repr__(self): return "<{0}: ID3v2.{1} tag{2} with {3} frames>".format( type(self).__name__, self.version, ("({0})".format(", ".join(self.flags)) if len(self.flags) > 0 else ""), len(self._frames)) # Reading tags @classmethod def read(cls, filename, offset=0): """Read an ID3v2 tag from a file.""" i = 0 with fileutil.opened(filename, "rb") as file: file.seek(offset) tag = cls() tag._read_header(file) for (frameid, bflags, data) in tag._read_frames(file): if len(data) == 0: warn("{0}: Ignoring empty frame".format(frameid), EmptyFrameWarning) else: frame = tag._decode_frame(frameid, bflags, data, i) if frame is not None: l = tag._frames.setdefault(frame.frameid, []) l.append(frame) if file.tell() > tag.offset + tag.size: break i += 1 try: tag._filename = file.name except AttributeError: pass return tag @classmethod def decode(cls, data): return cls.read(io.BytesIO(data)) def _decode_frame(self, frameid, bflags, data, frameno=None): try: (flags, data) = self._interpret_frame_flags(frameid, bflags, data) if flags is None: flags = set() if frameid in self.known_frames: return self.known_frames[frameid]._decode(frameid, data, flags, frameno=frameno) else: # Unknown frame flags.add("unknown") warn("{0}: Unknown frame".format(frameid), UnknownFrameWarning) if frameid.startswith('T'): # Unknown text frame return Frames.TextFrame._decode(frameid, data, flags, frameno=frameno) elif frameid.startswith('W'): # Unknown URL frame return Frames.URLFrame._decode(frameid, data, flags, frameno=frameno) else: return Frames.UnknownFrame._decode(frameid, data, flags, frameno=frameno) except (FrameError, ValueError, EOFError) as e: warn("{0}: Invalid frame".format(frameid), ErrorFrameWarning) return Frames.ErrorFrame(frameid, data, exception=e, frameno=frameno) @abstractmethod def _read_header(self, file): pass @abstractmethod def _read_frames(self, file): pass @abstractmethod def _interpret_frame_flags(self, frameid, bflags, data): pass # Writing tags def write(self, filename=None): if not filename: filename = self._filename if not filename: raise TypeError("invalid file: {0}".format(filename)) with fileutil.opened(filename, "rb+") as file: try: (offset, length) = detect_tag(file)[1:3] except NoTagError: (offset, length) = (0, 0) if offset > 0: delete_tag(file) (offset, length) = (0, 0) tag_data = self.encode(size_hint=length) fileutil.replace_chunk(file, offset, length, tag_data) @abstractmethod def encode(self, size_hint=None): pass padding_default = 128 padding_max = 1024 def _get_size_with_padding(self, size_desired, size_actual): size = size_actual if (size_desired is not None and size < size_desired and (self.padding_max is None or size_desired - size_actual <= self.padding_max)): size = size_desired elif self.padding_default: size += min(self.padding_default, self.padding_max) return size @staticmethod def _is_frame_id(data): if isinstance(data, str): try: data = data.encode("ASCII") except UnicodeEncodeError: return false # Allow a single space at end of four-character ids # Some programs (e.g. iTunes 8.2) generate such frames when converting # from 2.2 to 2.3/2.4 tags. pattern = re.compile(b"^[A-Z][A-Z0-9]{2}[A-Z0-9 ]?$") return pattern.match(data) def _prepare_frames_hook(self): pass def _prepare_frames(self): # Generate dictionary of frames d = self._frames # Merge duplicate frames for frameid in self._frames.keys(): fs = self._frames[frameid] if len(fs) > 1: d[frameid] = fs[0]._merge(fs) self._prepare_frames_hook() # Convert frames newframes = [] for frameid in self._frames.keys(): for frame in self._frames[frameid]: try: newframes.append(frame._to_version(self.version)) except IncompatibleFrameError: warn("{0}: Ignoring incompatible frame".format(frameid), FrameWarning) except ValueError as e: warn("{0}: Ignoring invalid frame ({1})".format(frameid, e), FrameWarning) # Sort frames newframes.sort(key=self.frame_order.key) return newframes class Tag22(Tag): version = 2 encodings = ("latin-1", "utf-16") def __init__(self): super().__init__() title = property(*Tag._friendly_text_frame("TT2")) artist = property(*Tag._friendly_text_frame("TP1")) @property def date(self): fields = self._get_date("TYE", "TDA", "TIM") return self._friendly_date_string(*fields) @date.setter def date(self, value): self._validate_friendly_date(value) (year, month, day, hour, minute, second) = self._get_friendly_date(value) for f in "TYE", "TDA", "TIM": if f in self: del self[f] if year is not None: self["TYE"] = "{0:04}".format(year) if month is not None and day is not None: self["TDA"] = "{0:02}{1:02}".format(month, day) if hour is not None and minute is not None: self["TIM"] = "{0:02}{1:02}".format(hour, minute) album_artist = property(*Tag._friendly_text_frame("TP2")) album = property(*Tag._friendly_text_frame("TAL")) track = property(*Tag._friendly_track("TRK", "track_total")) track_total = property(*Tag._friendly_track_total("TRK", "track")) disc = property(*Tag._friendly_track("TPA", "disc_total")) disc_total = property(*Tag._friendly_track_total("TPA", "disc")) composer = property(*Tag._friendly_text_frame("TCM")) genre = property(*Tag._friendly_text_frame("TCO")) comment = property(*Tag._friendly_comment("COM")) grouping = property(*Tag._friendly_text_frame("TT1")) # TODO: compilation picture = property(*Tag._friendly_picture("PIC")) sort_title = property(*Tag._friendly_text_frame("TST")) sort_artist = property(*Tag._friendly_text_frame("TSP")) sort_album_artist = property(*Tag._friendly_text_frame("TS2")) sort_album = property(*Tag._friendly_text_frame("TSA")) sort_composer = property(*Tag._friendly_text_frame("TSC")) def _read_header(self, file): self.offset = file.tell() header = fileutil.xread(file, 10) if header[0:5] != b"ID3\x02\00": raise NoTagError("ID3v2.2 header not found") if header[5] & 0x80: self.flags.add("unsynchronised") if header[5] & 0x40: # Compression bit is ill-defined in standard raise TagError("ID3v2.2 tag compression is not supported") if header[5] & 0x3F: warn("Unknown ID3v2.2 flags", TagWarning) self.size = Syncsafe.decode(header[6:10]) + 10 def _read_frames(self, file): if "unsynchronised" in self.flags: ufile = UnsyncReader(file) else: ufile = file while file.tell() < self.offset + self.size: header = fileutil.xread(ufile, 6) if not self._is_frame_id(header[0:3]): break frameid = header[0:3].decode("ASCII") size = Int8.decode(header[3:6]) data = fileutil.xread(ufile, size) yield (frameid, None, data) def _interpret_frame_flags(self, frameid, bflags, data): # No frame flags in v2.2 return (None, data) def __encode_one_frame(self, frame): framedata = frame._encode(encodings=self.encodings) data = bytearray() # Frame id if len(frame.frameid) != 3 or not self._is_frame_id(frame.frameid): raise "Invalid ID3v2.2 frame id {0}".format(repr(frame.frameid)) data.extend(frame.frameid.encode("ASCII")) # Size data.extend(Int8.encode(len(framedata), width=3)) assert(len(data) == 6) data.extend(framedata) return data def _prepare_frames_hook(self): for frameid in self._frames.keys(): for frame in self._frames[frameid]: if isinstance(frame, Frames.TextFrame): # ID3v2.2 doesn't support multiple text values if len(frame.text) > 1: warn("{0}: merged multiple text strings into one value" .format(frame.frameid), FrameWarning) frame.text = [" / ".join(frame.text)] def encode(self, size_hint=None): if len(self) == 0: # No frames -> no tag return b"" frames = self._prepare_frames() framedata = bytearray().join(self.__encode_one_frame(frame) for frame in frames) if "unsynchronised" in self.flags: framedata = Unsync.encode(framedata) size = self._get_size_with_padding(size_hint, len(framedata) + 10) data = bytearray() data.extend(b"ID3\x02\x00") data.append(0x80 if "unsynchronised" in self.flags else 0x00) data.extend(Syncsafe.encode(size - 10, width=4)) assert len(data) == 10 data.extend(framedata) if size > len(data): data.extend(b"\x00" * (size - len(data))) assert len(data) == size return data class Tag23(Tag): version = 3 encodings = ("latin-1", "utf-16") def __init__(self): super().__init__() title = property(*Tag._friendly_text_frame("TIT2")) artist = property(*Tag._friendly_text_frame("TPE1")) @property def date(self): fields = self._get_date("TYER", "TDAT", "TIME") return self._friendly_date_string(*fields) @date.setter def date(self, value): self._validate_friendly_date(value) (year, month, day, hour, minute, second) = self._get_friendly_date(value) for f in "TYER", "TDAT", "TIME": if f in self: del self[f] if year is not None: self["TYER"] = "{0:04}".format(year) if month is not None and day is not None: self["TDAT"] = "{0:02}{1:02}".format(month, day) if hour is not None and minute is not None: self["TIME"] = "{0:02}{1:02}".format(hour, minute) album_artist = property(*Tag._friendly_text_frame("TPE2")) album = property(*Tag._friendly_text_frame("TALB")) track = property(*Tag._friendly_track("TRCK", "track_total")) track_total = property(*Tag._friendly_track_total("TRCK", "track")) disc = property(*Tag._friendly_track("TPOS", "disc_total")) disc_total = property(*Tag._friendly_track_total("TPOS", "disc")) composer = property(*Tag._friendly_text_frame("TCOM")) genre = property(*Tag._friendly_text_frame("TCON")) comment = property(*Tag._friendly_comment("COMM")) grouping = property(*Tag._friendly_text_frame("TIT1")) # TODO: compilation picture = property(*Tag._friendly_picture("APIC")) sort_title = property(*Tag._friendly_text_frame("TSOT")) sort_artist = property(*Tag._friendly_text_frame("TSOP")) sort_album_artist = property(*Tag._friendly_text_frame("TSO2")) sort_album = property(*Tag._friendly_text_frame("TSOA")) sort_composer = property(*Tag._friendly_text_frame("TSOC")) def _read_header(self, file): self.offset = file.tell() header = fileutil.xread(file, 10) if header[0:5] != b"ID3\x03\x00": raise NoTagError("ID3v2.3 header not found") if header[5] & 0x80: self.flags.add("unsynchronised") if header[5] & 0x40: self.flags.add("extended_header") if header[5] & 0x20: self.flags.add("experimental") if header[5] & 0x1F: warn("Unknown ID3v2.3 flags", TagWarning) self.size = Syncsafe.decode(header[6:10]) + 10 if "extended_header" in self.flags: self.__read_extended_header(file) def __read_extended_header(self, file): (size, ext_flags, self.padding_size) = \ struct.unpack("!IHI", fileutil.xread(file, 10)) if size != 6 and size != 10: warn("Unexpected size of ID3v2.3 extended header: {0}".format(size), TagWarning) if ext_flags & 32768: if size < 10: warn("Extended header is too short for a CRC field: {0} bytes instead of 10".format(size), TagWarning) else: self.flags.add("ext:crc_present") (self.crc32,) = struct.unpack("!I", fileutil.xread(file, 4)) size -= 6 if size > 6: fileutil.xread(file, size - 6) def _read_frames(self, file): if "unsynchronised" in self.flags: ufile = UnsyncReader(file) else: ufile = file while file.tell() < self.offset + self.size: header = fileutil.xread(ufile, 10) if not self._is_frame_id(header[0:4]): break frameid = header[0:4].decode("ASCII") size = Int8.decode(header[4:8]) bflags = Int8.decode(header[8:10]) data = fileutil.xread(ufile, size) yield (frameid, bflags, data) def _interpret_frame_flags(self, frameid, bflags, data): flags = set() # Frame encoding flags if bflags & _FRAME23_FORMAT_UNKNOWN_MASK: raise FrameError("{0}: Invalid ID3v2.3 frame encoding flags: 0x{0:X}".format(frameid, bflags)) if bflags & _FRAME23_FORMAT_COMPRESSED: flags.add("compressed") expanded_size = Int8.decode(data[0:4]) data = zlib.decompress(data[4:]) if bflags & _FRAME23_FORMAT_ENCRYPTED: raise FrameError("{0}: Can't read ID3v2.3 encrypted frames".format(frameid)) if bflags & _FRAME23_FORMAT_GROUP: flags.add("group") flags.add("group={0}".format(data[0])) # Hack data = data[1:] # Frame status messages if bflags & _FRAME23_STATUS_DISCARD_ON_TAG_ALTER: flags.add("discard_on_tag_alter") if bflags & _FRAME23_STATUS_DISCARD_ON_FILE_ALTER: flags.add("discard_on_file_alter") if bflags & _FRAME23_STATUS_READ_ONLY: flags.add("read_only") if bflags & _FRAME23_STATUS_UNKNOWN_MASK: warn("{0}: Unexpected ID3v2.3 frame status flags: 0x{1:X}".format(frameid, bflags), TagWarning) return flags, data def __encode_one_frame(self, frame): framedata = frame._encode(encodings=self.encodings) origlen = len(framedata) flagval = 0 frameinfo = bytearray() if "compressed" in frame.flags: framedata = zlib.compress(framedata) flagval |= _FRAME23_FORMAT_COMPRESSED frameinfo.extend(Int8.encode(origlen, width=4)) if "group" in frame.flags: grp = 0 for flag in frame.flags: if flag.startswith("group="): grp = int(flag[6:]) frameinfo.append(grp) flagval |= _FRAME23_FORMAT_GROUP if "discard_on_tag_alter" in frame.flags: flagval |= _FRAME23_STATUS_DISCARD_ON_TAG_ALTER if "discard_on_file_alter" in frame.flags: flagval |= _FRAME23_STATUS_DISCARD_ON_FILE_ALTER if "read_only" in frame.flags: flagval |= _FRAME23_STATUS_READ_ONLY data = bytearray() # Frame id if len(frame.frameid) != 4 or not self._is_frame_id(frame.frameid.encode("ASCII")): raise ValueError("Invalid ID3v2.3 frame id {0}".format(repr(frame.frameid))) data.extend(frame.frameid.encode("ASCII")) # Size data.extend(Int8.encode(len(frameinfo) + len(framedata), width=4)) # Flags data.extend(Int8.encode(flagval, width=2)) assert len(data) == 10 # Format info data.extend(frameinfo) # Frame data data.extend(framedata) return data def encode(self, size_hint=None): if len(self) == 0: # No frames -> no tag return b"" frames = self._prepare_frames() framedata = bytearray().join(self.__encode_one_frame(frame) for frame in frames) if "unsynchronised" in self.flags: framedata = Unsync.encode(framedata) size = self._get_size_with_padding(size_hint, len(framedata) + 10) data = bytearray() data.extend(b"ID3\x03\x00") flagval = 0x00 if "unsynchronised" in self.flags: flagval |= 0x80 data.append(flagval) data.extend(Syncsafe.encode(size - 10, width=4)) assert len(data) == 10 data.extend(framedata) if size > len(data): data.extend(b"\x00" * (size - len(data))) assert len(data) == size return data class Tag24(Tag): ITUNES_WORKAROUND = False version = 4 encodings = ("latin-1", "utf-8") def __init__(self): super().__init__() title = property(*Tag._friendly_text_frame("TIT2")) artist = property(*Tag._friendly_text_frame("TPE1")) @property def date(self): try: frame = self["TDRC"] except KeyError: return "" else: return frame.text[0] @date.setter def date(self, value): self._validate_friendly_date(value) fields = self._get_friendly_date(value) val = self._friendly_date_string(*fields) if val: self["TDRC"] = val elif "TDRC" in self: del self["TDRC"] album = property(*Tag._friendly_text_frame("TALB")) album_artist = property(*Tag._friendly_text_frame("TPE2")) track = property(*Tag._friendly_track("TRCK", "track_total")) track_total = property(*Tag._friendly_track_total("TRCK", "track")) disc = property(*Tag._friendly_track("TPOS", "disc_total")) disc_total = property(*Tag._friendly_track_total("TPOS", "disc")) composer = property(*Tag._friendly_text_frame("TCOM")) genre = property(*Tag._friendly_text_frame("TCON")) comment = property(*Tag._friendly_comment("COMM")) grouping = property(*Tag._friendly_text_frame("TIT1")) # TODO: compilation picture = property(*Tag._friendly_picture("APIC")) sort_title = property(*Tag._friendly_text_frame("TSOT")) sort_artist = property(*Tag._friendly_text_frame("TSOP")) sort_album_artist = property(*Tag._friendly_text_frame("TSO2")) sort_album = property(*Tag._friendly_text_frame("TSOA")) sort_composer = property(*Tag._friendly_text_frame("TSOC")) def _read_header(self, file): self.offset = file.tell() header = fileutil.xread(file, 10) if header[0:5] != b"ID3\x04\x00": raise NoTagError("ID3v2 header not found") if header[5] & _TAG24_UNSYNCHRONISED: self.flags.add("unsynchronised") if header[5] & _TAG24_EXTENDED_HEADER: self.flags.add("extended_header") if header[5] & _TAG24_EXPERIMENTAL: self.flags.add("experimental") if header[5] & _TAG24_FOOTER: self.flags.add("footer") if header[5] & _TAG24_UNKNOWN_MASK: warn("Unknown ID3v2.4 flags", TagWarning) self.size = (Syncsafe.decode(header[6:10]) + 10 + (10 if "footer" in self.flags else 0)) if "extended_header" in self.flags: self.__read_extended_header(file) def __read_extended_header_flag_data(self, data): # 1-byte length + data length = data[0] if length & 128: raise TagError("Invalid size of extended header field") return (data[1:1+length], data[1+length:]) def __read_extended_header(self, file): size = Syncsafe.decode(fileutil.xread(file, 4)) if size < 6: warn("Unexpected size of ID3v2.4 extended header: {0}".format(size), TagWarning) data = fileutil.xread(file, size - 4) numflags = data[0] if numflags != 1: warn("Unexpected number of ID3v2.4 extended flag bytes: {0}" .format(numflags), TagWarning) flags = data[1] data = data[1+numflags:] if flags & 0x40: self.flags.add("ext:update") (dummy, data) = self.__read_extended_header_flag_data(data) if flags & 0x20: self.flags.add("ext:crc_present") (self.crc32, data) = self.__read_extended_header_flag_data(data) self.crc32 = Syncsafe.decode(self.crc32) if flags & 0x10: self.flags.add("ext:restrictions") (self.restrictions, data) = self.__read_extended_header_flag_data(data) def _read_frames(self, file, syncsafe_workaround = None): # Older versions of iTunes stored frame sizes as straight 8bit integers, # not syncsafe values as the spec requires. # (The bug is known to be fixed in iTunes 8.2.) # # To work around such an erroneous encoding, we re-read the entire tag # in non-syncsafe mode when we detect a frame with a bad size. # This heuristic does not detect all badly encoded tags; # it fails when the 8-bit frame size happens to be in syncsafe format. # # We could improve detection by parsing the tag both ways and see which # interpretation produces more frames. However, the extra effort doesn't # seem worthwhile to do by default. # # If you have many files with iTunes-encoded tags, you can force stagger # to read them in non-syncsafe mode setting the ITUNES_WORKAROUND # class attribute to True and let stagger reencode your tags. (Stagger # will never produce a 2.4 tag with non-syncsafe frame lengths.) if syncsafe_workaround is None: syncsafe_workaround = self.ITUNES_WORKAROUND origfpos = file.tell() frames = [] while file.tell() < self.offset + self.size: header = fileutil.xread(file, 10) if not self._is_frame_id(header[0:4]): break frameid = header[0:4].decode("ASCII") if syncsafe_workaround: size = Int8.decode(header[4:8]) else: try: size = Syncsafe.decode(header[4:8]) except ValueError: if syncsafe_workaround: raise warn("Invalid syncsafe frame size; switching to 8-bit mode") file.seek(origfpos) return self._read_frames(file, True) bflags = Int8.decode(header[8:10]) data = fileutil.xread(file, size) frames.append((frameid, bflags, data)) return frames def _interpret_frame_flags(self, frameid, bflags, data): flags = set() # Frame format flags if bflags & _FRAME24_FORMAT_UNKNOWN_MASK: raise FrameError("{0}: Unknown frame encoding flags: 0x{1:X}".format(frameid, bflags)) if bflags & _FRAME24_FORMAT_GROUP: flags.add("group") flags.add("group={0}".format(data[0])) # hack data = data[1:] if bflags & _FRAME24_FORMAT_COMPRESSED: flags.add("compressed") if bflags & _FRAME24_FORMAT_ENCRYPTED: raise FrameError("{0}: Can't read encrypted frames".format(frameid)) if bflags & _FRAME24_FORMAT_UNSYNCHRONISED: flags.add("unsynchronised") expanded_size = len(data) if bflags & _FRAME24_FORMAT_DATA_LENGTH_INDICATOR: flags.add("data_length_indicator") expanded_size = Syncsafe.decode(data[0:4]) data = data[4:] if "unsynchronised" in self.flags: data = Unsync.decode(data) if "compressed" in self.flags: data = zlib.decompress(data) # Frame status flags if bflags & _FRAME24_STATUS_DISCARD_ON_TAG_ALTER: flags.add("discard_on_tag_alter") if bflags & _FRAME24_STATUS_DISCARD_ON_FILE_ALTER: flags.add("discard_on_file_alter") if bflags & _FRAME24_STATUS_READ_ONLY: flags.add("read_only") if bflags & _FRAME24_STATUS_UNKNOWN_MASK: warn("{0}: Unexpected status flags: 0x{1:X}" .format(frameid, bflags), FrameWarning) return flags, data def __encode_one_frame(self, frame): framedata = frame._encode(encodings=self.encodings) origlen = len(framedata) flagval = 0 frameinfo = bytearray() if "group" in frame.flags: grp = 0 for flag in frame.flags: if flag.startswith("group="): grp = int(flag[6:]) frameinfo.append(grp) flagval |= _FRAME24_FORMAT_GROUP if "compressed" in frame.flags: frame.flags.add("data_length_indicator") framedata = zlib.compress(framedata) flagval |= _FRAME24_FORMAT_COMPRESSED if "unsynchronised" in frame.flags: frame.flags.add("data_length_indicator") framedata = Unsync.encode(framedata) flagval |= _FRAME24_FORMAT_UNSYNCHRONISED if "data_length_indicator" in frame.flags: frameinfo.extend(Syncsafe.encode(origlen, width=4)) flagval |= _FRAME24_FORMAT_DATA_LENGTH_INDICATOR if "discard_on_tag_alter" in frame.flags: flagval |= _FRAME24_STATUS_DISCARD_ON_TAG_ALTER if "discard_on_file_alter" in frame.flags: flagval |= _FRAME24_STATUS_DISCARD_ON_FILE_ALTER if "read_only" in frame.flags: flagval |= _FRAME24_STATUS_READ_ONLY data = bytearray() # Frame id if len(frame.frameid) != 4 or not self._is_frame_id(frame.frameid): raise "{0}: Invalid frame id".format(repr(frame.frameid)) data.extend(frame.frameid.encode("ASCII")) # Size data.extend(Syncsafe.encode(len(frameinfo) + len(framedata), width=4)) # Flags data.extend(Int8.encode(flagval, width=2)) assert len(data) == 10 # Format info data.extend(frameinfo) # Frame data data.extend(framedata) return data def encode(self, size_hint=None): if len(self) == 0: # No frames -> no tag return b"" frames = self._prepare_frames() if "unsynchronised" in self.flags: for frame in frames: frame.flags.add("unsynchronised") framedata = bytearray().join(self.__encode_one_frame(frame) for frame in frames) size = self._get_size_with_padding(size_hint, len(framedata) + 10) data = bytearray() data.extend(b"ID3\x04\x00") flagval = 0x00 if "unsynchronised" in self.flags: flagval |= 0x80 data.append(flagval) data.extend(Syncsafe.encode(size - 10, width=4)) assert len(data) == 10 data.extend(framedata) if size > len(data): data.extend(b"\x00" * (size - len(data))) assert len(data) == size return data _tag_versions = { 2: Tag22, 3: Tag23, 4: Tag24, }
staggerpkg/stagger
stagger/tags.py
FrameOrder.key
python
def key(self, frame): "Return the sort key for the given frame." def keytuple(primary): if frame.frameno is None: return (primary, 1) return (primary, 0, frame.frameno) # Look up frame by exact match if type(frame) in self.frame_keys: return keytuple(self.frame_keys[type(frame)]) # Look up parent frame for v2.2 frames if frame._in_version(2) and type(frame).__bases__[0] in self.frame_keys: return keytuple(self.frame_keys[type(frame).__bases__[0]]) # Try each pattern for (pattern, key) in self.re_keys: if re.match(pattern, frame.frameid): return keytuple(key) return keytuple(self.unknown_key)
Return the sort key for the given frame.
train
https://github.com/staggerpkg/stagger/blob/6530db14afc5d7d8a4599b7f3b26158fb367d786/stagger/tags.py#L191-L211
[ "def keytuple(primary):\n if frame.frameno is None:\n return (primary, 1)\n return (primary, 0, frame.frameno)\n" ]
class FrameOrder: """Order frames based on their position in a predefined list of patterns, and their original position in the source tag. A pattern may be a frame class, or a regular expression that is to be matched against the frame id. >>> order = FrameOrder(TIT1, "T.*", TXXX) >>> order.key(TIT1()) (0, 1) >>> order.key(TPE1()) (1, 1) >>> order.key(TXXX()) (2, 1) >>> order.key(APIC()) (3, 1) >>> order.key(APIC(frameno=3)) (3, 0, 3) """ def __init__(self, *patterns): self.re_keys = [] self.frame_keys = dict() i = -1 for (i, pattern) in zip(range(len(patterns)), patterns): if isinstance(pattern, str): self.re_keys.append((pattern, i)) else: assert issubclass(pattern, Frames.Frame) self.frame_keys[pattern] = i self.unknown_key = i + 1 def __repr__(self): order = [] order.extend((repr(pair[0]), pair[1]) for pair in self.re_keys) order.extend((cls.__name__, self.frame_keys[cls]) for cls in self.frame_keys) order.sort(key=lambda pair: pair[1]) return "<FrameOrder: {0}>".format(", ".join(pair[0] for pair in order))
staggerpkg/stagger
stagger/tags.py
Tag.frames
python
def frames(self, key=None, orig_order=False): if key is not None: # If there are multiple frames, then they are already in original order. key = self._normalize_key(key) if len(self._frames[key]) == 0: raise KeyError("Key not found: " + repr(key)) return self._frames[key] frames = [] for frameid in self._frames.keys(): for frame in self._frames[frameid]: frames.append(frame) if orig_order: key = (lambda frame: (0, frame.frameno) if frame.frameno is not None else (1,)) else: key = self.frame_order.key frames.sort(key=key) return frames
Returns a list of frames in this tag. If KEY is None, returns all frames in the tag; otherwise returns all frames whose frameid matches KEY. If ORIG_ORDER is True, then the frames are returned in their original order. Otherwise the frames are sorted in canonical order according to the frame_order field of this tag.
train
https://github.com/staggerpkg/stagger/blob/6530db14afc5d7d8a4599b7f3b26158fb367d786/stagger/tags.py#L233-L261
null
class Tag(collections.MutableMapping, metaclass=abc.ABCMeta): known_frames = { } # Maps known frameids to Frame class objects frame_order = None # Initialized by stagger.id3 def __init__(self): self.flags = set() self._frames = dict() self._filename = None # Primary accessor (no magic) def frames(self, key=None, orig_order=False): """Returns a list of frames in this tag. If KEY is None, returns all frames in the tag; otherwise returns all frames whose frameid matches KEY. If ORIG_ORDER is True, then the frames are returned in their original order. Otherwise the frames are sorted in canonical order according to the frame_order field of this tag. """ if key is not None: # If there are multiple frames, then they are already in original order. key = self._normalize_key(key) if len(self._frames[key]) == 0: raise KeyError("Key not found: " + repr(key)) return self._frames[key] frames = [] for frameid in self._frames.keys(): for frame in self._frames[frameid]: frames.append(frame) if orig_order: key = (lambda frame: (0, frame.frameno) if frame.frameno is not None else (1,)) else: key = self.frame_order.key frames.sort(key=key) return frames # MutableMapping API def __iter__(self): for frameid in self._frames: yield frameid def __len__(self): return sum(len(self._frames[l]) for l in self._frames) def __eq__(self, other): return (self.version == other.version and self.flags == other.flags and self._frames == other._frames) def _normalize_key(self, key, unknown_ok=True): """Return the normalized version of KEY. KEY may be a frameid (a string), or a Frame class object. If KEY corresponds to a registered frameid, then that frameid is returned. Otherwise, either KeyError is raised, or KEY is returned verbatim, depending on the value of UNKNOWN_OK. """ if Frames.is_frame_class(key): key = key.frameid if isinstance(key, str): if not self._is_frame_id(key): raise KeyError("{0}: Invalid frame id".format(key)) if key not in self.known_frames: if unknown_ok: warn("{0}: Unknown frame id".format(key), UnknownFrameWarning) else: raise KeyError("{0}: Unknown frame id".format(key)) return key # Mapping accessor (with extra magic, for convenience) def __getitem__(self, key): key = self._normalize_key(key) fs = self.frames(key) allow_duplicates = (key not in self.known_frames or self.known_frames[key]._allow_duplicates) if allow_duplicates: return fs if len(fs) > 1: # Merge duplicates into one ephemeral frame, and return that. # This may break users' expectations when they try to make changes # to the attributes of the returned frame; however, I think # sometimes returning a list, sometimes a single frame for the same # frame id would be even worse. fs = fs[0]._merge(fs) assert len(fs) == 1 return fs[0] def __setitem__(self, key, value): key = self._normalize_key(key, unknown_ok=False) if isinstance(value, self.known_frames[key]): self._frames[key] = [value] return if self.known_frames[key]._allow_duplicates: if not isinstance(value, collections.Iterable) or isinstance(value, str): raise ValueError("{0} requires a list of frame values".format(key)) self._frames[key] = [val if isinstance(val, self.known_frames[key]) else self.known_frames[key](val) for val in value] else: # not _allow_duplicates self._frames[key] = [self.known_frames[key](value)] def __delitem__(self, key): del self._frames[self._normalize_key(key)] def values(self): for frameid in self._frames.keys(): for frame in self._frames[frameid]: yield frame # Friendly names API _friendly_names = [ "title", "artist", "date", "album-artist", "album", "track", "track-total", "disc", "disc-total", "grouping", "composer", "genre", "comment", #"compilation", "picture", "sort-title", "sort-artist", "sort-album-artist", "sort-album", "sort-composer", ] title = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) artist = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) date = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) album_artist = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) album = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) track = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) track_total = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) disc = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) disc_total = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) composer = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) genre = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) comment = abstractproperty(fget=lambda self: Non, fset=lambda self, value: None) grouping = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) picture = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) sort_title = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) sort_artist = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) sort_album_artist = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) sort_album = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) sort_composer = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) def __friendly_text_collect(self, frameid): """Collect text values from all instances of FRAMEID into a single list. Returns an empty list if there are no instances of FRAMEID with a text attribute. """ try: return self[frameid].text except (KeyError, AttributeError): return [] @classmethod def _friendly_text_frame(cls, frameid): def getter(self): return " / ".join(self.__friendly_text_collect(frameid)) def setter(self, value): if isinstance(value, str): if len(value): # For non-empty strings, split value self[frameid] = value.split(" / ") elif frameid in self: # For empty strings, delete frame del self[frameid] else: self[frameid] = value return (getter, setter) @classmethod def _friendly_track(cls, frameid, totalattr): def getter(self): ts = self.__friendly_text_collect(frameid) try: return int(ts[0].partition("/")[0]) except (ValueError, IndexError): return 0 def setter(self, value): value = int(value) total = getattr(self, totalattr) if total > 0: self[frameid] = "{0}/{1}".format(value, total) elif value: self[frameid] = str(value) elif frameid in self: del self[frameid] return (getter, setter) @classmethod def _friendly_track_total(cls, frameid, trackattr): def getter(self): ts = self.__friendly_text_collect(frameid) try: return int(ts[0].partition("/")[2]) except (ValueError, IndexError): return 0 def setter(self, value): value = int(value) track = getattr(self, trackattr) if value: self[frameid] = "{0}/{1}".format(track, value) elif track: self[frameid] = str(track) elif frameid in self: del self[frameid] return (getter, setter) __date_pattern = re.compile(r"""(?x)\s* ((?P<year>[0-9]{4}) # YYYY (-(?P<month>[01][0-9]) # -MM (-(?P<day>[0-3][0-9]) # -DD )?)?)? [ T]? ((?P<hour>[0-2][0-9]) # HH (:(?P<min>[0-6][0-9]) # :MM (:(?P<sec>[0-6][0-9]) # :SS )?)?)?\s* """) @classmethod def _validate_friendly_date(cls, string): m = cls.__date_pattern.match(string) if m is None or m.end() != len(string): raise ValueError("date must be in 'YYYY-MM-DD HH:MM:SS' format") @classmethod def _get_friendly_date(cls, string): m = cls.__date_pattern.match(string) if m is None: return (None, None, None, None, None, None) res = [] for field in ("year", "month", "day", "hour", "min", "sec"): v = m.group(field) res.append(int(v) if v is not None else None) return res def _get_date(self, yearframe, dateframe, timeframe): year = month = day = hour = minute = second = None # Parse year. try: year = int(self.__friendly_text_collect(yearframe)[0]) except (IndexError, ValueError): pass # Parse month and date. try: date = self.__friendly_text_collect(dateframe)[0] m = re.match(r"\s*(?P<month>[01][0-9])\s*-?\s*(?P<day>[0-3][0-9])?\s*$", date) if m is not None: month = int(m.group("month")) day = int(m.group("day")) except IndexError: pass # Parse time. try: time = self.__friendly_text_collect(timeframe)[0] m = re.match(r"\s*(?P<hour>[0-2][0-9])\s*:?\s*" "(?P<minute>[0-5][0-9])\s*:?\s*" "(?P<second>[0-5][0-9])?\s*$", time) if m is not None: hour = int(m.group("hour")) minute = int(m.group("minute")) s = m.group("second") second = int(s) if s is not None else None except IndexError: pass return (year, month, day, hour, minute, second) def _friendly_date_string(self, *fields): seps = ("", "-", "-", " ", ":", ":") formats = ("04", "02", "02", "02", "02", "02") res = [] for i in range(len(fields)): if fields[i] is None: break res.append(seps[i]) res.append("{0:{1}}".format(fields[i], formats[i])) return "".join(res) @classmethod def _friendly_picture(cls, frameid): def getter(self): if frameid not in self: return "" else: return ", ".join("{0}:{1}:<{2} bytes of {3} data>" .format(f._spec("type").to_str(f.type), f.desc, len(f.data), imghdr.what(None, f.data[:32])) for f in self[frameid]) def setter(self, value): if len(value) > 0: self[frameid] = [self.known_frames[frameid](value=value)] elif frameid in self: del self[frameid] return (getter, setter) @classmethod def _friendly_comment(cls, frameid): def comment_frame_index(self): if frameid not in self: return None # Return comment with lang="eng", desc="", if present. # Otherwise return the first comment with no description, # regardless of language. icmt = None for i in range(len(self[frameid])): f = self[frameid][i] if f.desc == "": if f.lang == "eng": return i if icmt is None: icmt = i return icmt def getter(self): i = comment_frame_index(self) if i is None: return "" else: return self[frameid][i].text def setter(self, value): assert isinstance(value, str) i = comment_frame_index(self) if i is not None: del self._frames[frameid][i] if len(value) > 0: frame = self.known_frames[frameid](lang="eng", desc="", text=value) if frameid not in self._frames: self._frames[frameid] = [] self._frames[frameid].append(frame) return (getter, setter) # Misc def __repr__(self): return "<{0}: ID3v2.{1} tag{2} with {3} frames>".format( type(self).__name__, self.version, ("({0})".format(", ".join(self.flags)) if len(self.flags) > 0 else ""), len(self._frames)) # Reading tags @classmethod def read(cls, filename, offset=0): """Read an ID3v2 tag from a file.""" i = 0 with fileutil.opened(filename, "rb") as file: file.seek(offset) tag = cls() tag._read_header(file) for (frameid, bflags, data) in tag._read_frames(file): if len(data) == 0: warn("{0}: Ignoring empty frame".format(frameid), EmptyFrameWarning) else: frame = tag._decode_frame(frameid, bflags, data, i) if frame is not None: l = tag._frames.setdefault(frame.frameid, []) l.append(frame) if file.tell() > tag.offset + tag.size: break i += 1 try: tag._filename = file.name except AttributeError: pass return tag @classmethod def decode(cls, data): return cls.read(io.BytesIO(data)) def _decode_frame(self, frameid, bflags, data, frameno=None): try: (flags, data) = self._interpret_frame_flags(frameid, bflags, data) if flags is None: flags = set() if frameid in self.known_frames: return self.known_frames[frameid]._decode(frameid, data, flags, frameno=frameno) else: # Unknown frame flags.add("unknown") warn("{0}: Unknown frame".format(frameid), UnknownFrameWarning) if frameid.startswith('T'): # Unknown text frame return Frames.TextFrame._decode(frameid, data, flags, frameno=frameno) elif frameid.startswith('W'): # Unknown URL frame return Frames.URLFrame._decode(frameid, data, flags, frameno=frameno) else: return Frames.UnknownFrame._decode(frameid, data, flags, frameno=frameno) except (FrameError, ValueError, EOFError) as e: warn("{0}: Invalid frame".format(frameid), ErrorFrameWarning) return Frames.ErrorFrame(frameid, data, exception=e, frameno=frameno) @abstractmethod def _read_header(self, file): pass @abstractmethod def _read_frames(self, file): pass @abstractmethod def _interpret_frame_flags(self, frameid, bflags, data): pass # Writing tags def write(self, filename=None): if not filename: filename = self._filename if not filename: raise TypeError("invalid file: {0}".format(filename)) with fileutil.opened(filename, "rb+") as file: try: (offset, length) = detect_tag(file)[1:3] except NoTagError: (offset, length) = (0, 0) if offset > 0: delete_tag(file) (offset, length) = (0, 0) tag_data = self.encode(size_hint=length) fileutil.replace_chunk(file, offset, length, tag_data) @abstractmethod def encode(self, size_hint=None): pass padding_default = 128 padding_max = 1024 def _get_size_with_padding(self, size_desired, size_actual): size = size_actual if (size_desired is not None and size < size_desired and (self.padding_max is None or size_desired - size_actual <= self.padding_max)): size = size_desired elif self.padding_default: size += min(self.padding_default, self.padding_max) return size @staticmethod def _is_frame_id(data): if isinstance(data, str): try: data = data.encode("ASCII") except UnicodeEncodeError: return false # Allow a single space at end of four-character ids # Some programs (e.g. iTunes 8.2) generate such frames when converting # from 2.2 to 2.3/2.4 tags. pattern = re.compile(b"^[A-Z][A-Z0-9]{2}[A-Z0-9 ]?$") return pattern.match(data) def _prepare_frames_hook(self): pass def _prepare_frames(self): # Generate dictionary of frames d = self._frames # Merge duplicate frames for frameid in self._frames.keys(): fs = self._frames[frameid] if len(fs) > 1: d[frameid] = fs[0]._merge(fs) self._prepare_frames_hook() # Convert frames newframes = [] for frameid in self._frames.keys(): for frame in self._frames[frameid]: try: newframes.append(frame._to_version(self.version)) except IncompatibleFrameError: warn("{0}: Ignoring incompatible frame".format(frameid), FrameWarning) except ValueError as e: warn("{0}: Ignoring invalid frame ({1})".format(frameid, e), FrameWarning) # Sort frames newframes.sort(key=self.frame_order.key) return newframes
staggerpkg/stagger
stagger/tags.py
Tag._normalize_key
python
def _normalize_key(self, key, unknown_ok=True): if Frames.is_frame_class(key): key = key.frameid if isinstance(key, str): if not self._is_frame_id(key): raise KeyError("{0}: Invalid frame id".format(key)) if key not in self.known_frames: if unknown_ok: warn("{0}: Unknown frame id".format(key), UnknownFrameWarning) else: raise KeyError("{0}: Unknown frame id".format(key)) return key
Return the normalized version of KEY. KEY may be a frameid (a string), or a Frame class object. If KEY corresponds to a registered frameid, then that frameid is returned. Otherwise, either KeyError is raised, or KEY is returned verbatim, depending on the value of UNKNOWN_OK.
train
https://github.com/staggerpkg/stagger/blob/6530db14afc5d7d8a4599b7f3b26158fb367d786/stagger/tags.py#L276-L293
null
class Tag(collections.MutableMapping, metaclass=abc.ABCMeta): known_frames = { } # Maps known frameids to Frame class objects frame_order = None # Initialized by stagger.id3 def __init__(self): self.flags = set() self._frames = dict() self._filename = None # Primary accessor (no magic) def frames(self, key=None, orig_order=False): """Returns a list of frames in this tag. If KEY is None, returns all frames in the tag; otherwise returns all frames whose frameid matches KEY. If ORIG_ORDER is True, then the frames are returned in their original order. Otherwise the frames are sorted in canonical order according to the frame_order field of this tag. """ if key is not None: # If there are multiple frames, then they are already in original order. key = self._normalize_key(key) if len(self._frames[key]) == 0: raise KeyError("Key not found: " + repr(key)) return self._frames[key] frames = [] for frameid in self._frames.keys(): for frame in self._frames[frameid]: frames.append(frame) if orig_order: key = (lambda frame: (0, frame.frameno) if frame.frameno is not None else (1,)) else: key = self.frame_order.key frames.sort(key=key) return frames # MutableMapping API def __iter__(self): for frameid in self._frames: yield frameid def __len__(self): return sum(len(self._frames[l]) for l in self._frames) def __eq__(self, other): return (self.version == other.version and self.flags == other.flags and self._frames == other._frames) # Mapping accessor (with extra magic, for convenience) def __getitem__(self, key): key = self._normalize_key(key) fs = self.frames(key) allow_duplicates = (key not in self.known_frames or self.known_frames[key]._allow_duplicates) if allow_duplicates: return fs if len(fs) > 1: # Merge duplicates into one ephemeral frame, and return that. # This may break users' expectations when they try to make changes # to the attributes of the returned frame; however, I think # sometimes returning a list, sometimes a single frame for the same # frame id would be even worse. fs = fs[0]._merge(fs) assert len(fs) == 1 return fs[0] def __setitem__(self, key, value): key = self._normalize_key(key, unknown_ok=False) if isinstance(value, self.known_frames[key]): self._frames[key] = [value] return if self.known_frames[key]._allow_duplicates: if not isinstance(value, collections.Iterable) or isinstance(value, str): raise ValueError("{0} requires a list of frame values".format(key)) self._frames[key] = [val if isinstance(val, self.known_frames[key]) else self.known_frames[key](val) for val in value] else: # not _allow_duplicates self._frames[key] = [self.known_frames[key](value)] def __delitem__(self, key): del self._frames[self._normalize_key(key)] def values(self): for frameid in self._frames.keys(): for frame in self._frames[frameid]: yield frame # Friendly names API _friendly_names = [ "title", "artist", "date", "album-artist", "album", "track", "track-total", "disc", "disc-total", "grouping", "composer", "genre", "comment", #"compilation", "picture", "sort-title", "sort-artist", "sort-album-artist", "sort-album", "sort-composer", ] title = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) artist = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) date = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) album_artist = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) album = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) track = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) track_total = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) disc = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) disc_total = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) composer = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) genre = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) comment = abstractproperty(fget=lambda self: Non, fset=lambda self, value: None) grouping = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) picture = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) sort_title = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) sort_artist = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) sort_album_artist = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) sort_album = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) sort_composer = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) def __friendly_text_collect(self, frameid): """Collect text values from all instances of FRAMEID into a single list. Returns an empty list if there are no instances of FRAMEID with a text attribute. """ try: return self[frameid].text except (KeyError, AttributeError): return [] @classmethod def _friendly_text_frame(cls, frameid): def getter(self): return " / ".join(self.__friendly_text_collect(frameid)) def setter(self, value): if isinstance(value, str): if len(value): # For non-empty strings, split value self[frameid] = value.split(" / ") elif frameid in self: # For empty strings, delete frame del self[frameid] else: self[frameid] = value return (getter, setter) @classmethod def _friendly_track(cls, frameid, totalattr): def getter(self): ts = self.__friendly_text_collect(frameid) try: return int(ts[0].partition("/")[0]) except (ValueError, IndexError): return 0 def setter(self, value): value = int(value) total = getattr(self, totalattr) if total > 0: self[frameid] = "{0}/{1}".format(value, total) elif value: self[frameid] = str(value) elif frameid in self: del self[frameid] return (getter, setter) @classmethod def _friendly_track_total(cls, frameid, trackattr): def getter(self): ts = self.__friendly_text_collect(frameid) try: return int(ts[0].partition("/")[2]) except (ValueError, IndexError): return 0 def setter(self, value): value = int(value) track = getattr(self, trackattr) if value: self[frameid] = "{0}/{1}".format(track, value) elif track: self[frameid] = str(track) elif frameid in self: del self[frameid] return (getter, setter) __date_pattern = re.compile(r"""(?x)\s* ((?P<year>[0-9]{4}) # YYYY (-(?P<month>[01][0-9]) # -MM (-(?P<day>[0-3][0-9]) # -DD )?)?)? [ T]? ((?P<hour>[0-2][0-9]) # HH (:(?P<min>[0-6][0-9]) # :MM (:(?P<sec>[0-6][0-9]) # :SS )?)?)?\s* """) @classmethod def _validate_friendly_date(cls, string): m = cls.__date_pattern.match(string) if m is None or m.end() != len(string): raise ValueError("date must be in 'YYYY-MM-DD HH:MM:SS' format") @classmethod def _get_friendly_date(cls, string): m = cls.__date_pattern.match(string) if m is None: return (None, None, None, None, None, None) res = [] for field in ("year", "month", "day", "hour", "min", "sec"): v = m.group(field) res.append(int(v) if v is not None else None) return res def _get_date(self, yearframe, dateframe, timeframe): year = month = day = hour = minute = second = None # Parse year. try: year = int(self.__friendly_text_collect(yearframe)[0]) except (IndexError, ValueError): pass # Parse month and date. try: date = self.__friendly_text_collect(dateframe)[0] m = re.match(r"\s*(?P<month>[01][0-9])\s*-?\s*(?P<day>[0-3][0-9])?\s*$", date) if m is not None: month = int(m.group("month")) day = int(m.group("day")) except IndexError: pass # Parse time. try: time = self.__friendly_text_collect(timeframe)[0] m = re.match(r"\s*(?P<hour>[0-2][0-9])\s*:?\s*" "(?P<minute>[0-5][0-9])\s*:?\s*" "(?P<second>[0-5][0-9])?\s*$", time) if m is not None: hour = int(m.group("hour")) minute = int(m.group("minute")) s = m.group("second") second = int(s) if s is not None else None except IndexError: pass return (year, month, day, hour, minute, second) def _friendly_date_string(self, *fields): seps = ("", "-", "-", " ", ":", ":") formats = ("04", "02", "02", "02", "02", "02") res = [] for i in range(len(fields)): if fields[i] is None: break res.append(seps[i]) res.append("{0:{1}}".format(fields[i], formats[i])) return "".join(res) @classmethod def _friendly_picture(cls, frameid): def getter(self): if frameid not in self: return "" else: return ", ".join("{0}:{1}:<{2} bytes of {3} data>" .format(f._spec("type").to_str(f.type), f.desc, len(f.data), imghdr.what(None, f.data[:32])) for f in self[frameid]) def setter(self, value): if len(value) > 0: self[frameid] = [self.known_frames[frameid](value=value)] elif frameid in self: del self[frameid] return (getter, setter) @classmethod def _friendly_comment(cls, frameid): def comment_frame_index(self): if frameid not in self: return None # Return comment with lang="eng", desc="", if present. # Otherwise return the first comment with no description, # regardless of language. icmt = None for i in range(len(self[frameid])): f = self[frameid][i] if f.desc == "": if f.lang == "eng": return i if icmt is None: icmt = i return icmt def getter(self): i = comment_frame_index(self) if i is None: return "" else: return self[frameid][i].text def setter(self, value): assert isinstance(value, str) i = comment_frame_index(self) if i is not None: del self._frames[frameid][i] if len(value) > 0: frame = self.known_frames[frameid](lang="eng", desc="", text=value) if frameid not in self._frames: self._frames[frameid] = [] self._frames[frameid].append(frame) return (getter, setter) # Misc def __repr__(self): return "<{0}: ID3v2.{1} tag{2} with {3} frames>".format( type(self).__name__, self.version, ("({0})".format(", ".join(self.flags)) if len(self.flags) > 0 else ""), len(self._frames)) # Reading tags @classmethod def read(cls, filename, offset=0): """Read an ID3v2 tag from a file.""" i = 0 with fileutil.opened(filename, "rb") as file: file.seek(offset) tag = cls() tag._read_header(file) for (frameid, bflags, data) in tag._read_frames(file): if len(data) == 0: warn("{0}: Ignoring empty frame".format(frameid), EmptyFrameWarning) else: frame = tag._decode_frame(frameid, bflags, data, i) if frame is not None: l = tag._frames.setdefault(frame.frameid, []) l.append(frame) if file.tell() > tag.offset + tag.size: break i += 1 try: tag._filename = file.name except AttributeError: pass return tag @classmethod def decode(cls, data): return cls.read(io.BytesIO(data)) def _decode_frame(self, frameid, bflags, data, frameno=None): try: (flags, data) = self._interpret_frame_flags(frameid, bflags, data) if flags is None: flags = set() if frameid in self.known_frames: return self.known_frames[frameid]._decode(frameid, data, flags, frameno=frameno) else: # Unknown frame flags.add("unknown") warn("{0}: Unknown frame".format(frameid), UnknownFrameWarning) if frameid.startswith('T'): # Unknown text frame return Frames.TextFrame._decode(frameid, data, flags, frameno=frameno) elif frameid.startswith('W'): # Unknown URL frame return Frames.URLFrame._decode(frameid, data, flags, frameno=frameno) else: return Frames.UnknownFrame._decode(frameid, data, flags, frameno=frameno) except (FrameError, ValueError, EOFError) as e: warn("{0}: Invalid frame".format(frameid), ErrorFrameWarning) return Frames.ErrorFrame(frameid, data, exception=e, frameno=frameno) @abstractmethod def _read_header(self, file): pass @abstractmethod def _read_frames(self, file): pass @abstractmethod def _interpret_frame_flags(self, frameid, bflags, data): pass # Writing tags def write(self, filename=None): if not filename: filename = self._filename if not filename: raise TypeError("invalid file: {0}".format(filename)) with fileutil.opened(filename, "rb+") as file: try: (offset, length) = detect_tag(file)[1:3] except NoTagError: (offset, length) = (0, 0) if offset > 0: delete_tag(file) (offset, length) = (0, 0) tag_data = self.encode(size_hint=length) fileutil.replace_chunk(file, offset, length, tag_data) @abstractmethod def encode(self, size_hint=None): pass padding_default = 128 padding_max = 1024 def _get_size_with_padding(self, size_desired, size_actual): size = size_actual if (size_desired is not None and size < size_desired and (self.padding_max is None or size_desired - size_actual <= self.padding_max)): size = size_desired elif self.padding_default: size += min(self.padding_default, self.padding_max) return size @staticmethod def _is_frame_id(data): if isinstance(data, str): try: data = data.encode("ASCII") except UnicodeEncodeError: return false # Allow a single space at end of four-character ids # Some programs (e.g. iTunes 8.2) generate such frames when converting # from 2.2 to 2.3/2.4 tags. pattern = re.compile(b"^[A-Z][A-Z0-9]{2}[A-Z0-9 ]?$") return pattern.match(data) def _prepare_frames_hook(self): pass def _prepare_frames(self): # Generate dictionary of frames d = self._frames # Merge duplicate frames for frameid in self._frames.keys(): fs = self._frames[frameid] if len(fs) > 1: d[frameid] = fs[0]._merge(fs) self._prepare_frames_hook() # Convert frames newframes = [] for frameid in self._frames.keys(): for frame in self._frames[frameid]: try: newframes.append(frame._to_version(self.version)) except IncompatibleFrameError: warn("{0}: Ignoring incompatible frame".format(frameid), FrameWarning) except ValueError as e: warn("{0}: Ignoring invalid frame ({1})".format(frameid, e), FrameWarning) # Sort frames newframes.sort(key=self.frame_order.key) return newframes
staggerpkg/stagger
stagger/tags.py
Tag.read
python
def read(cls, filename, offset=0): i = 0 with fileutil.opened(filename, "rb") as file: file.seek(offset) tag = cls() tag._read_header(file) for (frameid, bflags, data) in tag._read_frames(file): if len(data) == 0: warn("{0}: Ignoring empty frame".format(frameid), EmptyFrameWarning) else: frame = tag._decode_frame(frameid, bflags, data, i) if frame is not None: l = tag._frames.setdefault(frame.frameid, []) l.append(frame) if file.tell() > tag.offset + tag.size: break i += 1 try: tag._filename = file.name except AttributeError: pass return tag
Read an ID3v2 tag from a file.
train
https://github.com/staggerpkg/stagger/blob/6530db14afc5d7d8a4599b7f3b26158fb367d786/stagger/tags.py#L575-L598
null
class Tag(collections.MutableMapping, metaclass=abc.ABCMeta): known_frames = { } # Maps known frameids to Frame class objects frame_order = None # Initialized by stagger.id3 def __init__(self): self.flags = set() self._frames = dict() self._filename = None # Primary accessor (no magic) def frames(self, key=None, orig_order=False): """Returns a list of frames in this tag. If KEY is None, returns all frames in the tag; otherwise returns all frames whose frameid matches KEY. If ORIG_ORDER is True, then the frames are returned in their original order. Otherwise the frames are sorted in canonical order according to the frame_order field of this tag. """ if key is not None: # If there are multiple frames, then they are already in original order. key = self._normalize_key(key) if len(self._frames[key]) == 0: raise KeyError("Key not found: " + repr(key)) return self._frames[key] frames = [] for frameid in self._frames.keys(): for frame in self._frames[frameid]: frames.append(frame) if orig_order: key = (lambda frame: (0, frame.frameno) if frame.frameno is not None else (1,)) else: key = self.frame_order.key frames.sort(key=key) return frames # MutableMapping API def __iter__(self): for frameid in self._frames: yield frameid def __len__(self): return sum(len(self._frames[l]) for l in self._frames) def __eq__(self, other): return (self.version == other.version and self.flags == other.flags and self._frames == other._frames) def _normalize_key(self, key, unknown_ok=True): """Return the normalized version of KEY. KEY may be a frameid (a string), or a Frame class object. If KEY corresponds to a registered frameid, then that frameid is returned. Otherwise, either KeyError is raised, or KEY is returned verbatim, depending on the value of UNKNOWN_OK. """ if Frames.is_frame_class(key): key = key.frameid if isinstance(key, str): if not self._is_frame_id(key): raise KeyError("{0}: Invalid frame id".format(key)) if key not in self.known_frames: if unknown_ok: warn("{0}: Unknown frame id".format(key), UnknownFrameWarning) else: raise KeyError("{0}: Unknown frame id".format(key)) return key # Mapping accessor (with extra magic, for convenience) def __getitem__(self, key): key = self._normalize_key(key) fs = self.frames(key) allow_duplicates = (key not in self.known_frames or self.known_frames[key]._allow_duplicates) if allow_duplicates: return fs if len(fs) > 1: # Merge duplicates into one ephemeral frame, and return that. # This may break users' expectations when they try to make changes # to the attributes of the returned frame; however, I think # sometimes returning a list, sometimes a single frame for the same # frame id would be even worse. fs = fs[0]._merge(fs) assert len(fs) == 1 return fs[0] def __setitem__(self, key, value): key = self._normalize_key(key, unknown_ok=False) if isinstance(value, self.known_frames[key]): self._frames[key] = [value] return if self.known_frames[key]._allow_duplicates: if not isinstance(value, collections.Iterable) or isinstance(value, str): raise ValueError("{0} requires a list of frame values".format(key)) self._frames[key] = [val if isinstance(val, self.known_frames[key]) else self.known_frames[key](val) for val in value] else: # not _allow_duplicates self._frames[key] = [self.known_frames[key](value)] def __delitem__(self, key): del self._frames[self._normalize_key(key)] def values(self): for frameid in self._frames.keys(): for frame in self._frames[frameid]: yield frame # Friendly names API _friendly_names = [ "title", "artist", "date", "album-artist", "album", "track", "track-total", "disc", "disc-total", "grouping", "composer", "genre", "comment", #"compilation", "picture", "sort-title", "sort-artist", "sort-album-artist", "sort-album", "sort-composer", ] title = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) artist = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) date = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) album_artist = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) album = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) track = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) track_total = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) disc = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) disc_total = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) composer = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) genre = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) comment = abstractproperty(fget=lambda self: Non, fset=lambda self, value: None) grouping = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) picture = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) sort_title = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) sort_artist = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) sort_album_artist = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) sort_album = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) sort_composer = abstractproperty(fget=lambda self: None, fset=lambda self, value: None) def __friendly_text_collect(self, frameid): """Collect text values from all instances of FRAMEID into a single list. Returns an empty list if there are no instances of FRAMEID with a text attribute. """ try: return self[frameid].text except (KeyError, AttributeError): return [] @classmethod def _friendly_text_frame(cls, frameid): def getter(self): return " / ".join(self.__friendly_text_collect(frameid)) def setter(self, value): if isinstance(value, str): if len(value): # For non-empty strings, split value self[frameid] = value.split(" / ") elif frameid in self: # For empty strings, delete frame del self[frameid] else: self[frameid] = value return (getter, setter) @classmethod def _friendly_track(cls, frameid, totalattr): def getter(self): ts = self.__friendly_text_collect(frameid) try: return int(ts[0].partition("/")[0]) except (ValueError, IndexError): return 0 def setter(self, value): value = int(value) total = getattr(self, totalattr) if total > 0: self[frameid] = "{0}/{1}".format(value, total) elif value: self[frameid] = str(value) elif frameid in self: del self[frameid] return (getter, setter) @classmethod def _friendly_track_total(cls, frameid, trackattr): def getter(self): ts = self.__friendly_text_collect(frameid) try: return int(ts[0].partition("/")[2]) except (ValueError, IndexError): return 0 def setter(self, value): value = int(value) track = getattr(self, trackattr) if value: self[frameid] = "{0}/{1}".format(track, value) elif track: self[frameid] = str(track) elif frameid in self: del self[frameid] return (getter, setter) __date_pattern = re.compile(r"""(?x)\s* ((?P<year>[0-9]{4}) # YYYY (-(?P<month>[01][0-9]) # -MM (-(?P<day>[0-3][0-9]) # -DD )?)?)? [ T]? ((?P<hour>[0-2][0-9]) # HH (:(?P<min>[0-6][0-9]) # :MM (:(?P<sec>[0-6][0-9]) # :SS )?)?)?\s* """) @classmethod def _validate_friendly_date(cls, string): m = cls.__date_pattern.match(string) if m is None or m.end() != len(string): raise ValueError("date must be in 'YYYY-MM-DD HH:MM:SS' format") @classmethod def _get_friendly_date(cls, string): m = cls.__date_pattern.match(string) if m is None: return (None, None, None, None, None, None) res = [] for field in ("year", "month", "day", "hour", "min", "sec"): v = m.group(field) res.append(int(v) if v is not None else None) return res def _get_date(self, yearframe, dateframe, timeframe): year = month = day = hour = minute = second = None # Parse year. try: year = int(self.__friendly_text_collect(yearframe)[0]) except (IndexError, ValueError): pass # Parse month and date. try: date = self.__friendly_text_collect(dateframe)[0] m = re.match(r"\s*(?P<month>[01][0-9])\s*-?\s*(?P<day>[0-3][0-9])?\s*$", date) if m is not None: month = int(m.group("month")) day = int(m.group("day")) except IndexError: pass # Parse time. try: time = self.__friendly_text_collect(timeframe)[0] m = re.match(r"\s*(?P<hour>[0-2][0-9])\s*:?\s*" "(?P<minute>[0-5][0-9])\s*:?\s*" "(?P<second>[0-5][0-9])?\s*$", time) if m is not None: hour = int(m.group("hour")) minute = int(m.group("minute")) s = m.group("second") second = int(s) if s is not None else None except IndexError: pass return (year, month, day, hour, minute, second) def _friendly_date_string(self, *fields): seps = ("", "-", "-", " ", ":", ":") formats = ("04", "02", "02", "02", "02", "02") res = [] for i in range(len(fields)): if fields[i] is None: break res.append(seps[i]) res.append("{0:{1}}".format(fields[i], formats[i])) return "".join(res) @classmethod def _friendly_picture(cls, frameid): def getter(self): if frameid not in self: return "" else: return ", ".join("{0}:{1}:<{2} bytes of {3} data>" .format(f._spec("type").to_str(f.type), f.desc, len(f.data), imghdr.what(None, f.data[:32])) for f in self[frameid]) def setter(self, value): if len(value) > 0: self[frameid] = [self.known_frames[frameid](value=value)] elif frameid in self: del self[frameid] return (getter, setter) @classmethod def _friendly_comment(cls, frameid): def comment_frame_index(self): if frameid not in self: return None # Return comment with lang="eng", desc="", if present. # Otherwise return the first comment with no description, # regardless of language. icmt = None for i in range(len(self[frameid])): f = self[frameid][i] if f.desc == "": if f.lang == "eng": return i if icmt is None: icmt = i return icmt def getter(self): i = comment_frame_index(self) if i is None: return "" else: return self[frameid][i].text def setter(self, value): assert isinstance(value, str) i = comment_frame_index(self) if i is not None: del self._frames[frameid][i] if len(value) > 0: frame = self.known_frames[frameid](lang="eng", desc="", text=value) if frameid not in self._frames: self._frames[frameid] = [] self._frames[frameid].append(frame) return (getter, setter) # Misc def __repr__(self): return "<{0}: ID3v2.{1} tag{2} with {3} frames>".format( type(self).__name__, self.version, ("({0})".format(", ".join(self.flags)) if len(self.flags) > 0 else ""), len(self._frames)) # Reading tags @classmethod @classmethod def decode(cls, data): return cls.read(io.BytesIO(data)) def _decode_frame(self, frameid, bflags, data, frameno=None): try: (flags, data) = self._interpret_frame_flags(frameid, bflags, data) if flags is None: flags = set() if frameid in self.known_frames: return self.known_frames[frameid]._decode(frameid, data, flags, frameno=frameno) else: # Unknown frame flags.add("unknown") warn("{0}: Unknown frame".format(frameid), UnknownFrameWarning) if frameid.startswith('T'): # Unknown text frame return Frames.TextFrame._decode(frameid, data, flags, frameno=frameno) elif frameid.startswith('W'): # Unknown URL frame return Frames.URLFrame._decode(frameid, data, flags, frameno=frameno) else: return Frames.UnknownFrame._decode(frameid, data, flags, frameno=frameno) except (FrameError, ValueError, EOFError) as e: warn("{0}: Invalid frame".format(frameid), ErrorFrameWarning) return Frames.ErrorFrame(frameid, data, exception=e, frameno=frameno) @abstractmethod def _read_header(self, file): pass @abstractmethod def _read_frames(self, file): pass @abstractmethod def _interpret_frame_flags(self, frameid, bflags, data): pass # Writing tags def write(self, filename=None): if not filename: filename = self._filename if not filename: raise TypeError("invalid file: {0}".format(filename)) with fileutil.opened(filename, "rb+") as file: try: (offset, length) = detect_tag(file)[1:3] except NoTagError: (offset, length) = (0, 0) if offset > 0: delete_tag(file) (offset, length) = (0, 0) tag_data = self.encode(size_hint=length) fileutil.replace_chunk(file, offset, length, tag_data) @abstractmethod def encode(self, size_hint=None): pass padding_default = 128 padding_max = 1024 def _get_size_with_padding(self, size_desired, size_actual): size = size_actual if (size_desired is not None and size < size_desired and (self.padding_max is None or size_desired - size_actual <= self.padding_max)): size = size_desired elif self.padding_default: size += min(self.padding_default, self.padding_max) return size @staticmethod def _is_frame_id(data): if isinstance(data, str): try: data = data.encode("ASCII") except UnicodeEncodeError: return false # Allow a single space at end of four-character ids # Some programs (e.g. iTunes 8.2) generate such frames when converting # from 2.2 to 2.3/2.4 tags. pattern = re.compile(b"^[A-Z][A-Z0-9]{2}[A-Z0-9 ]?$") return pattern.match(data) def _prepare_frames_hook(self): pass def _prepare_frames(self): # Generate dictionary of frames d = self._frames # Merge duplicate frames for frameid in self._frames.keys(): fs = self._frames[frameid] if len(fs) > 1: d[frameid] = fs[0]._merge(fs) self._prepare_frames_hook() # Convert frames newframes = [] for frameid in self._frames.keys(): for frame in self._frames[frameid]: try: newframes.append(frame._to_version(self.version)) except IncompatibleFrameError: warn("{0}: Ignoring incompatible frame".format(frameid), FrameWarning) except ValueError as e: warn("{0}: Ignoring invalid frame ({1})".format(frameid, e), FrameWarning) # Sort frames newframes.sort(key=self.frame_order.key) return newframes
staggerpkg/stagger
stagger/id3v1.py
Tag1.read
python
def read(cls, filename, offset=None, encoding="iso-8859-1"): with fileutil.opened(filename, "rb") as file: if offset is None: file.seek(-128, 2) else: file.seek(offset) data = file.read(128) return cls.decode(data, encoding=encoding)
Read an ID3v1 tag from a file.
train
https://github.com/staggerpkg/stagger/blob/6530db14afc5d7d8a4599b7f3b26158fb367d786/stagger/id3v1.py#L120-L128
[ "def decode(cls, data, encoding=\"iso-8859-1\"):\n def decode_field(data):\n try:\n data = data[:data.index(b\"\\x00\")]\n except ValueError:\n pass\n return data.decode(encoding).strip(string.whitespace)\n if data[:3] != b\"TAG\" or len(data) < 128:\n raise N...
class Tag1(): @property def genre(self): if self._genre < len(genres): return "{0} ({1})".format(self._genre, genres[self._genre]) else: return "{0} (unknown)".format(self._genre) @genre.setter def genre(self, value): if value is None: self._genre = 0 return if type(value) == int: if value in range(256): self._genre = value else: raise ValueError("Genre must be between 0 and 255") return if type(value) == str: if value.lower() == "unknown": self._genre = 255 return try: self._genre = genres.index(value) return except ValueError: raise ValueError("Unknown genre") raise TypeError("Invalid genre") def __str__(self): return "\n".join(["title={0}".format(repr(self.title)), "artist={0}".format(repr(self.artist)), "album={0}".format(repr(self.album)), "year={0}".format(repr(self.year)), "comment={0}".format(repr(self.comment)), "genre={0}".format(self.genre)]) def __repr__(self): return "Tag1({0})".format(", ".join(["title={0}".format(repr(self.title)), "artist={0}".format(repr(self.artist)), "album={0}".format(repr(self.album)), "year={0}".format(repr(self.year)), "comment={0}".format(repr(self.comment)), "genre={0}".format(self._genre)])) def __eq__(self, other): return (isinstance(other, Tag1) and self.title == other.title and self.artist == other.artist and self.album == other.album and self.year == other.year and self.comment == other.comment and self._genre == other._genre) @classmethod def decode(cls, data, encoding="iso-8859-1"): def decode_field(data): try: data = data[:data.index(b"\x00")] except ValueError: pass return data.decode(encoding).strip(string.whitespace) if data[:3] != b"TAG" or len(data) < 128: raise NoTagError("ID3v1 tag not found") tag = Tag1() tag.title = decode_field(data[3:33]) tag.artist = decode_field(data[33:63]) tag.album = decode_field(data[63:93]) tag.year = decode_field(data[93:97]) if data[125] == 0: tag.comment = decode_field(data[97:125]) tag.track = data[126] else: tag.comment = decode_field(data[97:127]) tag.track = 0 tag._genre = data[127] return tag @classmethod @classmethod def delete(cls, filename, offset=None): """Delete ID3v1 tag from a file (if present).""" with fileutil.opened(filename, "rb+") as file: if offset is None: file.seek(-128, 2) else: file.seek(offset) offset = file.tell() data = file.read(128) if data[:3] == b"TAG": fileutil.replace_chunk(file, offset, 128, b"", in_place=True) def encode(self, encoding="iso-8859-1", errors="strict"): def encode_field(field, width): data = field.encode(encoding, errors) if len(data) < width: data = data + b"\x00" * (width - len(data)) return data[:width] data = bytearray(b"TAG") data.extend(encode_field(self.title, 30)) data.extend(encode_field(self.artist, 30)) data.extend(encode_field(self.album, 30)) data.extend(encode_field(self.year, 4)) if self.track: data.extend(encode_field(self.comment, 28)) data.append(0) data.append(self.track) else: data.extend(encode_field(self.comment, 30)) data.append(self._genre) assert len(data) == 128 return data def write(self, filename, encoding="iso-8859-1", errors="strict"): with fileutil.opened(filename, "rb+") as file: file.seek(-128, 2) data = file.read(128) if data[:3] == b"TAG": file.seek(-128, 2) else: file.seek(0, 2) file.write(self.encode(encoding, errors))
staggerpkg/stagger
stagger/id3v1.py
Tag1.delete
python
def delete(cls, filename, offset=None): with fileutil.opened(filename, "rb+") as file: if offset is None: file.seek(-128, 2) else: file.seek(offset) offset = file.tell() data = file.read(128) if data[:3] == b"TAG": fileutil.replace_chunk(file, offset, 128, b"", in_place=True)
Delete ID3v1 tag from a file (if present).
train
https://github.com/staggerpkg/stagger/blob/6530db14afc5d7d8a4599b7f3b26158fb367d786/stagger/id3v1.py#L131-L141
null
class Tag1(): @property def genre(self): if self._genre < len(genres): return "{0} ({1})".format(self._genre, genres[self._genre]) else: return "{0} (unknown)".format(self._genre) @genre.setter def genre(self, value): if value is None: self._genre = 0 return if type(value) == int: if value in range(256): self._genre = value else: raise ValueError("Genre must be between 0 and 255") return if type(value) == str: if value.lower() == "unknown": self._genre = 255 return try: self._genre = genres.index(value) return except ValueError: raise ValueError("Unknown genre") raise TypeError("Invalid genre") def __str__(self): return "\n".join(["title={0}".format(repr(self.title)), "artist={0}".format(repr(self.artist)), "album={0}".format(repr(self.album)), "year={0}".format(repr(self.year)), "comment={0}".format(repr(self.comment)), "genre={0}".format(self.genre)]) def __repr__(self): return "Tag1({0})".format(", ".join(["title={0}".format(repr(self.title)), "artist={0}".format(repr(self.artist)), "album={0}".format(repr(self.album)), "year={0}".format(repr(self.year)), "comment={0}".format(repr(self.comment)), "genre={0}".format(self._genre)])) def __eq__(self, other): return (isinstance(other, Tag1) and self.title == other.title and self.artist == other.artist and self.album == other.album and self.year == other.year and self.comment == other.comment and self._genre == other._genre) @classmethod def decode(cls, data, encoding="iso-8859-1"): def decode_field(data): try: data = data[:data.index(b"\x00")] except ValueError: pass return data.decode(encoding).strip(string.whitespace) if data[:3] != b"TAG" or len(data) < 128: raise NoTagError("ID3v1 tag not found") tag = Tag1() tag.title = decode_field(data[3:33]) tag.artist = decode_field(data[33:63]) tag.album = decode_field(data[63:93]) tag.year = decode_field(data[93:97]) if data[125] == 0: tag.comment = decode_field(data[97:125]) tag.track = data[126] else: tag.comment = decode_field(data[97:127]) tag.track = 0 tag._genre = data[127] return tag @classmethod def read(cls, filename, offset=None, encoding="iso-8859-1"): """Read an ID3v1 tag from a file.""" with fileutil.opened(filename, "rb") as file: if offset is None: file.seek(-128, 2) else: file.seek(offset) data = file.read(128) return cls.decode(data, encoding=encoding) @classmethod def encode(self, encoding="iso-8859-1", errors="strict"): def encode_field(field, width): data = field.encode(encoding, errors) if len(data) < width: data = data + b"\x00" * (width - len(data)) return data[:width] data = bytearray(b"TAG") data.extend(encode_field(self.title, 30)) data.extend(encode_field(self.artist, 30)) data.extend(encode_field(self.album, 30)) data.extend(encode_field(self.year, 4)) if self.track: data.extend(encode_field(self.comment, 28)) data.append(0) data.append(self.track) else: data.extend(encode_field(self.comment, 30)) data.append(self._genre) assert len(data) == 128 return data def write(self, filename, encoding="iso-8859-1", errors="strict"): with fileutil.opened(filename, "rb+") as file: file.seek(-128, 2) data = file.read(128) if data[:3] == b"TAG": file.seek(-128, 2) else: file.seek(0, 2) file.write(self.encode(encoding, errors))
staggerpkg/stagger
stagger/fileutil.py
xread
python
def xread(file, length): "Read exactly length bytes from file; raise EOFError if file ends sooner." data = file.read(length) if len(data) != length: raise EOFError return data
Read exactly length bytes from file; raise EOFError if file ends sooner.
train
https://github.com/staggerpkg/stagger/blob/6530db14afc5d7d8a4599b7f3b26158fb367d786/stagger/fileutil.py#L43-L48
null
# # fileutil.py # From the stagger project: http://code.google.com/p/stagger/ # # Copyright (c) 2009-2011 Karoly Lorentey <karoly@lorentey.hu> # All rights reserved. # # Redistribution and use in source and binary forms, with or without # modification, are permitted provided that the following conditions # are met: # # - Redistributions of source code must retain the above copyright # notice, this list of conditions and the following disclaimer. # # - Redistributions in binary form must reproduce the above copyright # notice, this list of conditions and the following disclaimer in # the documentation and/or other materials provided with the # distribution. # # THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS # "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT # LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS # FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE # COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, # INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, # BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; # LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER # CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT # LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN # ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE # POSSIBILITY OF SUCH DAMAGE. """File manipulation utilities.""" import io import os.path import shutil import tempfile import signal from contextlib import contextmanager @contextmanager def opened(filename, mode): "Open filename, or do nothing if filename is already an open file object" if isinstance(filename, str): file = open(filename, mode) try: yield file finally: if not file.closed: file.close() else: yield filename @contextmanager def suppress_interrupt(): """Suppress KeyboardInterrupt exceptions while the context is active. The suppressed interrupt (if any) is raised when the context is exited. """ interrupted = False def sigint_handler(signum, frame): nonlocal interrupted interrupted = True s = signal.signal(signal.SIGINT, sigint_handler) try: yield None finally: signal.signal(signal.SIGINT, s) if interrupted: raise KeyboardInterrupt() def replace_chunk(filename, offset, length, chunk, in_place=True, max_mem=5): """Replace length bytes of data with chunk, starting at offset. Any KeyboardInterrupts arriving while replace_chunk is runnning are deferred until the operation is complete. If in_place is true, the operation works directly on the original file; this is fast and works on files that are already open, but an error or interrupt may lead to corrupt file contents. If in_place is false, the function prepares a copy first, then renames it back over the original file. This method is slower, but it prevents corruption on systems with atomic renames (UNIX), and reduces the window of vulnerability elsewhere (Windows). If there is no need to move data that is not being replaced, then we use the direct method irrespective of in_place. (In this case an interrupt may only corrupt the chunk being replaced.) """ with suppress_interrupt(): _replace_chunk(filename, offset, length, chunk, in_place, max_mem) def _replace_chunk(filename, offset, length, chunk, in_place, max_mem): assert isinstance(filename, str) or in_place with opened(filename, "rb+") as file: # If the sizes match, we can simply overwrite the original data. if length == len(chunk): file.seek(offset) file.write(chunk) return oldsize = file.seek(0, 2) newsize = oldsize - length + len(chunk) # If the orig chunk is exactly at the end of the file, we can # simply truncate the file and then append the new chunk. if offset + length == oldsize: file.seek(offset) file.truncate() file.write(chunk) return if in_place: _replace_chunk_in_place(file, offset, length, chunk, oldsize, newsize) else: # not in_place temp = tempfile.NamedTemporaryFile(dir=os.path.dirname(filename), prefix="stagger-", suffix=".tmp", delete=False) try: file.seek(0) _copy_chunk(file, temp, offset) temp.write(chunk) file.seek(offset + length) _copy_chunk(file, temp, oldsize - offset - length) finally: temp.close() file.close() shutil.copymode(filename, temp.name) shutil.move(temp.name, filename) return def _copy_chunk(src, dst, length): "Copy length bytes from file src to file dst." BUFSIZE = 128 * 1024 while length > 0: l = min(BUFSIZE, length) buf = src.read(l) assert len(buf) == l dst.write(buf) length -= l def _replace_chunk_in_place(file, offset, length, chunk, oldsize, newsize): if newsize > oldsize: file.seek(0, 2) file.write(b"\x00" * (len(chunk) - length)) file.seek(0) try: import mmap m = mmap.mmap(file.fileno(), max(oldsize, newsize)) try: m.move(offset + len(chunk), offset + length, oldsize - offset - length) m[offset:offset + len(chunk)] = chunk finally: m.close() except (ImportError, EnvironmentError, ValueError): # mmap didn't work. Let's load the tail into a tempfile # and construct the result from there. file.seek(offset + length) temp = tempfile.SpooledTemporaryFile( max_size=max_mem * (1<<20), prefix="stagger-", suffix=".tmp") try: _copy_chunk(file, temp, oldsize - offset - length) file.seek(offset) file.truncate() file.write(chunk) temp.seek(0) _copy_chunk(temp, file, oldsize - offset - length) finally: temp.close() return else: # mmap did work, we just need to truncate any leftover parts # at the end file.truncate(newsize) return
staggerpkg/stagger
stagger/fileutil.py
opened
python
def opened(filename, mode): "Open filename, or do nothing if filename is already an open file object" if isinstance(filename, str): file = open(filename, mode) try: yield file finally: if not file.closed: file.close() else: yield filename
Open filename, or do nothing if filename is already an open file object
train
https://github.com/staggerpkg/stagger/blob/6530db14afc5d7d8a4599b7f3b26158fb367d786/stagger/fileutil.py#L51-L61
null
# # fileutil.py # From the stagger project: http://code.google.com/p/stagger/ # # Copyright (c) 2009-2011 Karoly Lorentey <karoly@lorentey.hu> # All rights reserved. # # Redistribution and use in source and binary forms, with or without # modification, are permitted provided that the following conditions # are met: # # - Redistributions of source code must retain the above copyright # notice, this list of conditions and the following disclaimer. # # - Redistributions in binary form must reproduce the above copyright # notice, this list of conditions and the following disclaimer in # the documentation and/or other materials provided with the # distribution. # # THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS # "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT # LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS # FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE # COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, # INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, # BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; # LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER # CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT # LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN # ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE # POSSIBILITY OF SUCH DAMAGE. """File manipulation utilities.""" import io import os.path import shutil import tempfile import signal from contextlib import contextmanager def xread(file, length): "Read exactly length bytes from file; raise EOFError if file ends sooner." data = file.read(length) if len(data) != length: raise EOFError return data @contextmanager @contextmanager def suppress_interrupt(): """Suppress KeyboardInterrupt exceptions while the context is active. The suppressed interrupt (if any) is raised when the context is exited. """ interrupted = False def sigint_handler(signum, frame): nonlocal interrupted interrupted = True s = signal.signal(signal.SIGINT, sigint_handler) try: yield None finally: signal.signal(signal.SIGINT, s) if interrupted: raise KeyboardInterrupt() def replace_chunk(filename, offset, length, chunk, in_place=True, max_mem=5): """Replace length bytes of data with chunk, starting at offset. Any KeyboardInterrupts arriving while replace_chunk is runnning are deferred until the operation is complete. If in_place is true, the operation works directly on the original file; this is fast and works on files that are already open, but an error or interrupt may lead to corrupt file contents. If in_place is false, the function prepares a copy first, then renames it back over the original file. This method is slower, but it prevents corruption on systems with atomic renames (UNIX), and reduces the window of vulnerability elsewhere (Windows). If there is no need to move data that is not being replaced, then we use the direct method irrespective of in_place. (In this case an interrupt may only corrupt the chunk being replaced.) """ with suppress_interrupt(): _replace_chunk(filename, offset, length, chunk, in_place, max_mem) def _replace_chunk(filename, offset, length, chunk, in_place, max_mem): assert isinstance(filename, str) or in_place with opened(filename, "rb+") as file: # If the sizes match, we can simply overwrite the original data. if length == len(chunk): file.seek(offset) file.write(chunk) return oldsize = file.seek(0, 2) newsize = oldsize - length + len(chunk) # If the orig chunk is exactly at the end of the file, we can # simply truncate the file and then append the new chunk. if offset + length == oldsize: file.seek(offset) file.truncate() file.write(chunk) return if in_place: _replace_chunk_in_place(file, offset, length, chunk, oldsize, newsize) else: # not in_place temp = tempfile.NamedTemporaryFile(dir=os.path.dirname(filename), prefix="stagger-", suffix=".tmp", delete=False) try: file.seek(0) _copy_chunk(file, temp, offset) temp.write(chunk) file.seek(offset + length) _copy_chunk(file, temp, oldsize - offset - length) finally: temp.close() file.close() shutil.copymode(filename, temp.name) shutil.move(temp.name, filename) return def _copy_chunk(src, dst, length): "Copy length bytes from file src to file dst." BUFSIZE = 128 * 1024 while length > 0: l = min(BUFSIZE, length) buf = src.read(l) assert len(buf) == l dst.write(buf) length -= l def _replace_chunk_in_place(file, offset, length, chunk, oldsize, newsize): if newsize > oldsize: file.seek(0, 2) file.write(b"\x00" * (len(chunk) - length)) file.seek(0) try: import mmap m = mmap.mmap(file.fileno(), max(oldsize, newsize)) try: m.move(offset + len(chunk), offset + length, oldsize - offset - length) m[offset:offset + len(chunk)] = chunk finally: m.close() except (ImportError, EnvironmentError, ValueError): # mmap didn't work. Let's load the tail into a tempfile # and construct the result from there. file.seek(offset + length) temp = tempfile.SpooledTemporaryFile( max_size=max_mem * (1<<20), prefix="stagger-", suffix=".tmp") try: _copy_chunk(file, temp, oldsize - offset - length) file.seek(offset) file.truncate() file.write(chunk) temp.seek(0) _copy_chunk(temp, file, oldsize - offset - length) finally: temp.close() return else: # mmap did work, we just need to truncate any leftover parts # at the end file.truncate(newsize) return
staggerpkg/stagger
stagger/fileutil.py
suppress_interrupt
python
def suppress_interrupt(): interrupted = False def sigint_handler(signum, frame): nonlocal interrupted interrupted = True s = signal.signal(signal.SIGINT, sigint_handler) try: yield None finally: signal.signal(signal.SIGINT, s) if interrupted: raise KeyboardInterrupt()
Suppress KeyboardInterrupt exceptions while the context is active. The suppressed interrupt (if any) is raised when the context is exited.
train
https://github.com/staggerpkg/stagger/blob/6530db14afc5d7d8a4599b7f3b26158fb367d786/stagger/fileutil.py#L64-L81
null
# # fileutil.py # From the stagger project: http://code.google.com/p/stagger/ # # Copyright (c) 2009-2011 Karoly Lorentey <karoly@lorentey.hu> # All rights reserved. # # Redistribution and use in source and binary forms, with or without # modification, are permitted provided that the following conditions # are met: # # - Redistributions of source code must retain the above copyright # notice, this list of conditions and the following disclaimer. # # - Redistributions in binary form must reproduce the above copyright # notice, this list of conditions and the following disclaimer in # the documentation and/or other materials provided with the # distribution. # # THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS # "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT # LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS # FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE # COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, # INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, # BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; # LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER # CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT # LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN # ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE # POSSIBILITY OF SUCH DAMAGE. """File manipulation utilities.""" import io import os.path import shutil import tempfile import signal from contextlib import contextmanager def xread(file, length): "Read exactly length bytes from file; raise EOFError if file ends sooner." data = file.read(length) if len(data) != length: raise EOFError return data @contextmanager def opened(filename, mode): "Open filename, or do nothing if filename is already an open file object" if isinstance(filename, str): file = open(filename, mode) try: yield file finally: if not file.closed: file.close() else: yield filename @contextmanager def suppress_interrupt(): """Suppress KeyboardInterrupt exceptions while the context is active. The suppressed interrupt (if any) is raised when the context is exited. """ interrupted = False def sigint_handler(signum, frame): nonlocal interrupted interrupted = True s = signal.signal(signal.SIGINT, sigint_handler) try: yield None finally: signal.signal(signal.SIGINT, s) if interrupted: raise KeyboardInterrupt() def replace_chunk(filename, offset, length, chunk, in_place=True, max_mem=5): """Replace length bytes of data with chunk, starting at offset. Any KeyboardInterrupts arriving while replace_chunk is runnning are deferred until the operation is complete. If in_place is true, the operation works directly on the original file; this is fast and works on files that are already open, but an error or interrupt may lead to corrupt file contents. If in_place is false, the function prepares a copy first, then renames it back over the original file. This method is slower, but it prevents corruption on systems with atomic renames (UNIX), and reduces the window of vulnerability elsewhere (Windows). If there is no need to move data that is not being replaced, then we use the direct method irrespective of in_place. (In this case an interrupt may only corrupt the chunk being replaced.) """ with suppress_interrupt(): _replace_chunk(filename, offset, length, chunk, in_place, max_mem) def _replace_chunk(filename, offset, length, chunk, in_place, max_mem): assert isinstance(filename, str) or in_place with opened(filename, "rb+") as file: # If the sizes match, we can simply overwrite the original data. if length == len(chunk): file.seek(offset) file.write(chunk) return oldsize = file.seek(0, 2) newsize = oldsize - length + len(chunk) # If the orig chunk is exactly at the end of the file, we can # simply truncate the file and then append the new chunk. if offset + length == oldsize: file.seek(offset) file.truncate() file.write(chunk) return if in_place: _replace_chunk_in_place(file, offset, length, chunk, oldsize, newsize) else: # not in_place temp = tempfile.NamedTemporaryFile(dir=os.path.dirname(filename), prefix="stagger-", suffix=".tmp", delete=False) try: file.seek(0) _copy_chunk(file, temp, offset) temp.write(chunk) file.seek(offset + length) _copy_chunk(file, temp, oldsize - offset - length) finally: temp.close() file.close() shutil.copymode(filename, temp.name) shutil.move(temp.name, filename) return def _copy_chunk(src, dst, length): "Copy length bytes from file src to file dst." BUFSIZE = 128 * 1024 while length > 0: l = min(BUFSIZE, length) buf = src.read(l) assert len(buf) == l dst.write(buf) length -= l def _replace_chunk_in_place(file, offset, length, chunk, oldsize, newsize): if newsize > oldsize: file.seek(0, 2) file.write(b"\x00" * (len(chunk) - length)) file.seek(0) try: import mmap m = mmap.mmap(file.fileno(), max(oldsize, newsize)) try: m.move(offset + len(chunk), offset + length, oldsize - offset - length) m[offset:offset + len(chunk)] = chunk finally: m.close() except (ImportError, EnvironmentError, ValueError): # mmap didn't work. Let's load the tail into a tempfile # and construct the result from there. file.seek(offset + length) temp = tempfile.SpooledTemporaryFile( max_size=max_mem * (1<<20), prefix="stagger-", suffix=".tmp") try: _copy_chunk(file, temp, oldsize - offset - length) file.seek(offset) file.truncate() file.write(chunk) temp.seek(0) _copy_chunk(temp, file, oldsize - offset - length) finally: temp.close() return else: # mmap did work, we just need to truncate any leftover parts # at the end file.truncate(newsize) return
staggerpkg/stagger
stagger/fileutil.py
replace_chunk
python
def replace_chunk(filename, offset, length, chunk, in_place=True, max_mem=5): with suppress_interrupt(): _replace_chunk(filename, offset, length, chunk, in_place, max_mem)
Replace length bytes of data with chunk, starting at offset. Any KeyboardInterrupts arriving while replace_chunk is runnning are deferred until the operation is complete. If in_place is true, the operation works directly on the original file; this is fast and works on files that are already open, but an error or interrupt may lead to corrupt file contents. If in_place is false, the function prepares a copy first, then renames it back over the original file. This method is slower, but it prevents corruption on systems with atomic renames (UNIX), and reduces the window of vulnerability elsewhere (Windows). If there is no need to move data that is not being replaced, then we use the direct method irrespective of in_place. (In this case an interrupt may only corrupt the chunk being replaced.)
train
https://github.com/staggerpkg/stagger/blob/6530db14afc5d7d8a4599b7f3b26158fb367d786/stagger/fileutil.py#L83-L102
[ "def _replace_chunk(filename, offset, length, chunk, in_place, max_mem):\n assert isinstance(filename, str) or in_place\n with opened(filename, \"rb+\") as file:\n # If the sizes match, we can simply overwrite the original data.\n if length == len(chunk):\n file.seek(offset)\n ...
# # fileutil.py # From the stagger project: http://code.google.com/p/stagger/ # # Copyright (c) 2009-2011 Karoly Lorentey <karoly@lorentey.hu> # All rights reserved. # # Redistribution and use in source and binary forms, with or without # modification, are permitted provided that the following conditions # are met: # # - Redistributions of source code must retain the above copyright # notice, this list of conditions and the following disclaimer. # # - Redistributions in binary form must reproduce the above copyright # notice, this list of conditions and the following disclaimer in # the documentation and/or other materials provided with the # distribution. # # THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS # "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT # LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS # FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE # COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, # INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, # BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; # LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER # CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT # LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN # ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE # POSSIBILITY OF SUCH DAMAGE. """File manipulation utilities.""" import io import os.path import shutil import tempfile import signal from contextlib import contextmanager def xread(file, length): "Read exactly length bytes from file; raise EOFError if file ends sooner." data = file.read(length) if len(data) != length: raise EOFError return data @contextmanager def opened(filename, mode): "Open filename, or do nothing if filename is already an open file object" if isinstance(filename, str): file = open(filename, mode) try: yield file finally: if not file.closed: file.close() else: yield filename @contextmanager def suppress_interrupt(): """Suppress KeyboardInterrupt exceptions while the context is active. The suppressed interrupt (if any) is raised when the context is exited. """ interrupted = False def sigint_handler(signum, frame): nonlocal interrupted interrupted = True s = signal.signal(signal.SIGINT, sigint_handler) try: yield None finally: signal.signal(signal.SIGINT, s) if interrupted: raise KeyboardInterrupt() def _replace_chunk(filename, offset, length, chunk, in_place, max_mem): assert isinstance(filename, str) or in_place with opened(filename, "rb+") as file: # If the sizes match, we can simply overwrite the original data. if length == len(chunk): file.seek(offset) file.write(chunk) return oldsize = file.seek(0, 2) newsize = oldsize - length + len(chunk) # If the orig chunk is exactly at the end of the file, we can # simply truncate the file and then append the new chunk. if offset + length == oldsize: file.seek(offset) file.truncate() file.write(chunk) return if in_place: _replace_chunk_in_place(file, offset, length, chunk, oldsize, newsize) else: # not in_place temp = tempfile.NamedTemporaryFile(dir=os.path.dirname(filename), prefix="stagger-", suffix=".tmp", delete=False) try: file.seek(0) _copy_chunk(file, temp, offset) temp.write(chunk) file.seek(offset + length) _copy_chunk(file, temp, oldsize - offset - length) finally: temp.close() file.close() shutil.copymode(filename, temp.name) shutil.move(temp.name, filename) return def _copy_chunk(src, dst, length): "Copy length bytes from file src to file dst." BUFSIZE = 128 * 1024 while length > 0: l = min(BUFSIZE, length) buf = src.read(l) assert len(buf) == l dst.write(buf) length -= l def _replace_chunk_in_place(file, offset, length, chunk, oldsize, newsize): if newsize > oldsize: file.seek(0, 2) file.write(b"\x00" * (len(chunk) - length)) file.seek(0) try: import mmap m = mmap.mmap(file.fileno(), max(oldsize, newsize)) try: m.move(offset + len(chunk), offset + length, oldsize - offset - length) m[offset:offset + len(chunk)] = chunk finally: m.close() except (ImportError, EnvironmentError, ValueError): # mmap didn't work. Let's load the tail into a tempfile # and construct the result from there. file.seek(offset + length) temp = tempfile.SpooledTemporaryFile( max_size=max_mem * (1<<20), prefix="stagger-", suffix=".tmp") try: _copy_chunk(file, temp, oldsize - offset - length) file.seek(offset) file.truncate() file.write(chunk) temp.seek(0) _copy_chunk(temp, file, oldsize - offset - length) finally: temp.close() return else: # mmap did work, we just need to truncate any leftover parts # at the end file.truncate(newsize) return
staggerpkg/stagger
stagger/fileutil.py
_copy_chunk
python
def _copy_chunk(src, dst, length): "Copy length bytes from file src to file dst." BUFSIZE = 128 * 1024 while length > 0: l = min(BUFSIZE, length) buf = src.read(l) assert len(buf) == l dst.write(buf) length -= l
Copy length bytes from file src to file dst.
train
https://github.com/staggerpkg/stagger/blob/6530db14afc5d7d8a4599b7f3b26158fb367d786/stagger/fileutil.py#L145-L153
null
# # fileutil.py # From the stagger project: http://code.google.com/p/stagger/ # # Copyright (c) 2009-2011 Karoly Lorentey <karoly@lorentey.hu> # All rights reserved. # # Redistribution and use in source and binary forms, with or without # modification, are permitted provided that the following conditions # are met: # # - Redistributions of source code must retain the above copyright # notice, this list of conditions and the following disclaimer. # # - Redistributions in binary form must reproduce the above copyright # notice, this list of conditions and the following disclaimer in # the documentation and/or other materials provided with the # distribution. # # THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS # "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT # LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS # FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE # COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, # INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, # BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; # LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER # CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT # LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN # ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE # POSSIBILITY OF SUCH DAMAGE. """File manipulation utilities.""" import io import os.path import shutil import tempfile import signal from contextlib import contextmanager def xread(file, length): "Read exactly length bytes from file; raise EOFError if file ends sooner." data = file.read(length) if len(data) != length: raise EOFError return data @contextmanager def opened(filename, mode): "Open filename, or do nothing if filename is already an open file object" if isinstance(filename, str): file = open(filename, mode) try: yield file finally: if not file.closed: file.close() else: yield filename @contextmanager def suppress_interrupt(): """Suppress KeyboardInterrupt exceptions while the context is active. The suppressed interrupt (if any) is raised when the context is exited. """ interrupted = False def sigint_handler(signum, frame): nonlocal interrupted interrupted = True s = signal.signal(signal.SIGINT, sigint_handler) try: yield None finally: signal.signal(signal.SIGINT, s) if interrupted: raise KeyboardInterrupt() def replace_chunk(filename, offset, length, chunk, in_place=True, max_mem=5): """Replace length bytes of data with chunk, starting at offset. Any KeyboardInterrupts arriving while replace_chunk is runnning are deferred until the operation is complete. If in_place is true, the operation works directly on the original file; this is fast and works on files that are already open, but an error or interrupt may lead to corrupt file contents. If in_place is false, the function prepares a copy first, then renames it back over the original file. This method is slower, but it prevents corruption on systems with atomic renames (UNIX), and reduces the window of vulnerability elsewhere (Windows). If there is no need to move data that is not being replaced, then we use the direct method irrespective of in_place. (In this case an interrupt may only corrupt the chunk being replaced.) """ with suppress_interrupt(): _replace_chunk(filename, offset, length, chunk, in_place, max_mem) def _replace_chunk(filename, offset, length, chunk, in_place, max_mem): assert isinstance(filename, str) or in_place with opened(filename, "rb+") as file: # If the sizes match, we can simply overwrite the original data. if length == len(chunk): file.seek(offset) file.write(chunk) return oldsize = file.seek(0, 2) newsize = oldsize - length + len(chunk) # If the orig chunk is exactly at the end of the file, we can # simply truncate the file and then append the new chunk. if offset + length == oldsize: file.seek(offset) file.truncate() file.write(chunk) return if in_place: _replace_chunk_in_place(file, offset, length, chunk, oldsize, newsize) else: # not in_place temp = tempfile.NamedTemporaryFile(dir=os.path.dirname(filename), prefix="stagger-", suffix=".tmp", delete=False) try: file.seek(0) _copy_chunk(file, temp, offset) temp.write(chunk) file.seek(offset + length) _copy_chunk(file, temp, oldsize - offset - length) finally: temp.close() file.close() shutil.copymode(filename, temp.name) shutil.move(temp.name, filename) return def _replace_chunk_in_place(file, offset, length, chunk, oldsize, newsize): if newsize > oldsize: file.seek(0, 2) file.write(b"\x00" * (len(chunk) - length)) file.seek(0) try: import mmap m = mmap.mmap(file.fileno(), max(oldsize, newsize)) try: m.move(offset + len(chunk), offset + length, oldsize - offset - length) m[offset:offset + len(chunk)] = chunk finally: m.close() except (ImportError, EnvironmentError, ValueError): # mmap didn't work. Let's load the tail into a tempfile # and construct the result from there. file.seek(offset + length) temp = tempfile.SpooledTemporaryFile( max_size=max_mem * (1<<20), prefix="stagger-", suffix=".tmp") try: _copy_chunk(file, temp, oldsize - offset - length) file.seek(offset) file.truncate() file.write(chunk) temp.seek(0) _copy_chunk(temp, file, oldsize - offset - length) finally: temp.close() return else: # mmap did work, we just need to truncate any leftover parts # at the end file.truncate(newsize) return
CameronLonsdale/lantern
lantern/analysis/frequency.py
frequency_to_probability
python
def frequency_to_probability(frequency_map, decorator=lambda f: f): total = sum(frequency_map.values()) return {k: decorator(v / total) for k, v in frequency_map.items()}
Transform a ``frequency_map`` into a map of probability using the sum of all frequencies as the total. Example: >>> frequency_to_probability({'a': 2, 'b': 2}) {'a': 0.5, 'b': 0.5} Args: frequency_map (dict): The dictionary to transform decorator (function): A function to manipulate the probability Returns: Dictionary of ngrams to probability
train
https://github.com/CameronLonsdale/lantern/blob/235e163e96bf0719d49c54204ee576b2ca93abb6/lantern/analysis/frequency.py#L34-L49
null
"""General purpose frequency analysis tools.""" import importlib import statistics from collections import Counter from lantern.structures import DynamicDict from lantern.util import iterate_ngrams def frequency_analyze(text, n=1): """Analyze the frequency of ngrams for a piece of text. Examples: >>> frequency_analyze("abb") {'a': 1, 'b': 2} >>> frequency_analyze("abb", 2) {'ab': 1, 'bb': 1} Args: text (str): The text to analyze n (int): The ngram size to use Returns: Dictionary of ngrams to frequency Raises: ValueError: If n is not a positive integer """ return Counter(iterate_ngrams(text, n)) def index_of_coincidence(*texts): """Calculate the index of coincidence for one or more ``texts``. The results are averaged over multiple texts to return the delta index of coincidence. Examples: >>> index_of_coincidence("aabbc") 0.2 >>> index_of_coincidence("aabbc", "abbcc") 0.2 Args: *texts (variable length argument list): The texts to analyze Returns: Decimal value of the index of coincidence Raises: ValueError: If texts is empty ValueError: If any text is less that 2 character long """ if not texts: raise ValueError("texts must not be empty") return statistics.mean(_calculate_index_of_coincidence(frequency_analyze(text), len(text)) for text in texts) def _calculate_index_of_coincidence(frequency_map, length): """A measure of how similar frequency_map is to the uniform distribution. Or the probability that two letters picked randomly are alike. """ if length <= 1: return 0 # We cannot error here as length can legitimiately be 1. # Imagine a ciphertext of length 3 and a key of length 2. # Spliting this text up and calculating the index of coincidence results in ['AC', 'B'] # IOC of B will be calcuated for the 2nd column of the key. We could represent the same # encryption with a key of length 3 but then we encounter the same problem. This is also # legitimiate encryption scheme we cannot ignore. Hence we have to deal with this fact here # A value of 0 will impact the overall mean, however it does make some sense when you ask the question # How many ways to choose 2 letters from the text, if theres only 1 letter then the answer is 0. # Mathemtical combination, number of ways to choose 2 letters, no replacement, order doesnt matter combination_of_letters = sum(freq * (freq - 1) for freq in frequency_map.values()) return combination_of_letters / (length * (length - 1)) def chi_squared(source_frequency, target_frequency): """Calculate the Chi Squared statistic by comparing ``source_frequency`` with ``target_frequency``. Example: >>> chi_squared({'a': 2, 'b': 3}, {'a': 1, 'b': 2}) 0.1 Args: source_frequency (dict): Frequency map of the text you are analyzing target_frequency (dict): Frequency map of the target language to compare with Returns: Decimal value of the chi-squared statistic """ # Ignore any symbols from source that are not in target. # TODO: raise Error if source_len is 0? target_prob = frequency_to_probability(target_frequency) source_len = sum(v for k, v in source_frequency.items() if k in target_frequency) result = 0 for symbol, prob in target_prob.items(): symbol_frequency = source_frequency.get(symbol, 0) # Frequecy is 0 if it doesnt appear in source result += _calculate_chi_squared(symbol_frequency, prob, source_len) return result def _calculate_chi_squared(source_freq, target_prob, source_len): """A measure of the observed frequency of the symbol versus the expected frequency. If the value is 0 then the texts are exactly alike for that symbol. """ expected = source_len * target_prob return (source_freq - expected)**2 / expected def _load_ngram(name): """Dynamically import the python module with the ngram defined as a dictionary. Since bigger ngrams are large files its wasteful to always statically import them if they're not used. """ module = importlib.import_module('lantern.analysis.english_ngrams.{}'.format(name)) return getattr(module, name) english = DynamicDict({ 'unigrams': lambda: _load_ngram('unigrams'), 'bigrams': lambda: _load_ngram('bigrams'), 'trigrams': lambda: _load_ngram('trigrams'), 'quadgrams': lambda: _load_ngram('quadgrams') }) """English ngram frequencies.""" ENGLISH_IC = _calculate_index_of_coincidence(english.unigrams, sum(english.unigrams.values())) """Index of coincidence for the English language."""
CameronLonsdale/lantern
lantern/analysis/frequency.py
index_of_coincidence
python
def index_of_coincidence(*texts): if not texts: raise ValueError("texts must not be empty") return statistics.mean(_calculate_index_of_coincidence(frequency_analyze(text), len(text)) for text in texts)
Calculate the index of coincidence for one or more ``texts``. The results are averaged over multiple texts to return the delta index of coincidence. Examples: >>> index_of_coincidence("aabbc") 0.2 >>> index_of_coincidence("aabbc", "abbcc") 0.2 Args: *texts (variable length argument list): The texts to analyze Returns: Decimal value of the index of coincidence Raises: ValueError: If texts is empty ValueError: If any text is less that 2 character long
train
https://github.com/CameronLonsdale/lantern/blob/235e163e96bf0719d49c54204ee576b2ca93abb6/lantern/analysis/frequency.py#L52-L76
null
"""General purpose frequency analysis tools.""" import importlib import statistics from collections import Counter from lantern.structures import DynamicDict from lantern.util import iterate_ngrams def frequency_analyze(text, n=1): """Analyze the frequency of ngrams for a piece of text. Examples: >>> frequency_analyze("abb") {'a': 1, 'b': 2} >>> frequency_analyze("abb", 2) {'ab': 1, 'bb': 1} Args: text (str): The text to analyze n (int): The ngram size to use Returns: Dictionary of ngrams to frequency Raises: ValueError: If n is not a positive integer """ return Counter(iterate_ngrams(text, n)) def frequency_to_probability(frequency_map, decorator=lambda f: f): """Transform a ``frequency_map`` into a map of probability using the sum of all frequencies as the total. Example: >>> frequency_to_probability({'a': 2, 'b': 2}) {'a': 0.5, 'b': 0.5} Args: frequency_map (dict): The dictionary to transform decorator (function): A function to manipulate the probability Returns: Dictionary of ngrams to probability """ total = sum(frequency_map.values()) return {k: decorator(v / total) for k, v in frequency_map.items()} def _calculate_index_of_coincidence(frequency_map, length): """A measure of how similar frequency_map is to the uniform distribution. Or the probability that two letters picked randomly are alike. """ if length <= 1: return 0 # We cannot error here as length can legitimiately be 1. # Imagine a ciphertext of length 3 and a key of length 2. # Spliting this text up and calculating the index of coincidence results in ['AC', 'B'] # IOC of B will be calcuated for the 2nd column of the key. We could represent the same # encryption with a key of length 3 but then we encounter the same problem. This is also # legitimiate encryption scheme we cannot ignore. Hence we have to deal with this fact here # A value of 0 will impact the overall mean, however it does make some sense when you ask the question # How many ways to choose 2 letters from the text, if theres only 1 letter then the answer is 0. # Mathemtical combination, number of ways to choose 2 letters, no replacement, order doesnt matter combination_of_letters = sum(freq * (freq - 1) for freq in frequency_map.values()) return combination_of_letters / (length * (length - 1)) def chi_squared(source_frequency, target_frequency): """Calculate the Chi Squared statistic by comparing ``source_frequency`` with ``target_frequency``. Example: >>> chi_squared({'a': 2, 'b': 3}, {'a': 1, 'b': 2}) 0.1 Args: source_frequency (dict): Frequency map of the text you are analyzing target_frequency (dict): Frequency map of the target language to compare with Returns: Decimal value of the chi-squared statistic """ # Ignore any symbols from source that are not in target. # TODO: raise Error if source_len is 0? target_prob = frequency_to_probability(target_frequency) source_len = sum(v for k, v in source_frequency.items() if k in target_frequency) result = 0 for symbol, prob in target_prob.items(): symbol_frequency = source_frequency.get(symbol, 0) # Frequecy is 0 if it doesnt appear in source result += _calculate_chi_squared(symbol_frequency, prob, source_len) return result def _calculate_chi_squared(source_freq, target_prob, source_len): """A measure of the observed frequency of the symbol versus the expected frequency. If the value is 0 then the texts are exactly alike for that symbol. """ expected = source_len * target_prob return (source_freq - expected)**2 / expected def _load_ngram(name): """Dynamically import the python module with the ngram defined as a dictionary. Since bigger ngrams are large files its wasteful to always statically import them if they're not used. """ module = importlib.import_module('lantern.analysis.english_ngrams.{}'.format(name)) return getattr(module, name) english = DynamicDict({ 'unigrams': lambda: _load_ngram('unigrams'), 'bigrams': lambda: _load_ngram('bigrams'), 'trigrams': lambda: _load_ngram('trigrams'), 'quadgrams': lambda: _load_ngram('quadgrams') }) """English ngram frequencies.""" ENGLISH_IC = _calculate_index_of_coincidence(english.unigrams, sum(english.unigrams.values())) """Index of coincidence for the English language."""
CameronLonsdale/lantern
lantern/analysis/frequency.py
_calculate_index_of_coincidence
python
def _calculate_index_of_coincidence(frequency_map, length): if length <= 1: return 0 # We cannot error here as length can legitimiately be 1. # Imagine a ciphertext of length 3 and a key of length 2. # Spliting this text up and calculating the index of coincidence results in ['AC', 'B'] # IOC of B will be calcuated for the 2nd column of the key. We could represent the same # encryption with a key of length 3 but then we encounter the same problem. This is also # legitimiate encryption scheme we cannot ignore. Hence we have to deal with this fact here # A value of 0 will impact the overall mean, however it does make some sense when you ask the question # How many ways to choose 2 letters from the text, if theres only 1 letter then the answer is 0. # Mathemtical combination, number of ways to choose 2 letters, no replacement, order doesnt matter combination_of_letters = sum(freq * (freq - 1) for freq in frequency_map.values()) return combination_of_letters / (length * (length - 1))
A measure of how similar frequency_map is to the uniform distribution. Or the probability that two letters picked randomly are alike.
train
https://github.com/CameronLonsdale/lantern/blob/235e163e96bf0719d49c54204ee576b2ca93abb6/lantern/analysis/frequency.py#L79-L96
null
"""General purpose frequency analysis tools.""" import importlib import statistics from collections import Counter from lantern.structures import DynamicDict from lantern.util import iterate_ngrams def frequency_analyze(text, n=1): """Analyze the frequency of ngrams for a piece of text. Examples: >>> frequency_analyze("abb") {'a': 1, 'b': 2} >>> frequency_analyze("abb", 2) {'ab': 1, 'bb': 1} Args: text (str): The text to analyze n (int): The ngram size to use Returns: Dictionary of ngrams to frequency Raises: ValueError: If n is not a positive integer """ return Counter(iterate_ngrams(text, n)) def frequency_to_probability(frequency_map, decorator=lambda f: f): """Transform a ``frequency_map`` into a map of probability using the sum of all frequencies as the total. Example: >>> frequency_to_probability({'a': 2, 'b': 2}) {'a': 0.5, 'b': 0.5} Args: frequency_map (dict): The dictionary to transform decorator (function): A function to manipulate the probability Returns: Dictionary of ngrams to probability """ total = sum(frequency_map.values()) return {k: decorator(v / total) for k, v in frequency_map.items()} def index_of_coincidence(*texts): """Calculate the index of coincidence for one or more ``texts``. The results are averaged over multiple texts to return the delta index of coincidence. Examples: >>> index_of_coincidence("aabbc") 0.2 >>> index_of_coincidence("aabbc", "abbcc") 0.2 Args: *texts (variable length argument list): The texts to analyze Returns: Decimal value of the index of coincidence Raises: ValueError: If texts is empty ValueError: If any text is less that 2 character long """ if not texts: raise ValueError("texts must not be empty") return statistics.mean(_calculate_index_of_coincidence(frequency_analyze(text), len(text)) for text in texts) def chi_squared(source_frequency, target_frequency): """Calculate the Chi Squared statistic by comparing ``source_frequency`` with ``target_frequency``. Example: >>> chi_squared({'a': 2, 'b': 3}, {'a': 1, 'b': 2}) 0.1 Args: source_frequency (dict): Frequency map of the text you are analyzing target_frequency (dict): Frequency map of the target language to compare with Returns: Decimal value of the chi-squared statistic """ # Ignore any symbols from source that are not in target. # TODO: raise Error if source_len is 0? target_prob = frequency_to_probability(target_frequency) source_len = sum(v for k, v in source_frequency.items() if k in target_frequency) result = 0 for symbol, prob in target_prob.items(): symbol_frequency = source_frequency.get(symbol, 0) # Frequecy is 0 if it doesnt appear in source result += _calculate_chi_squared(symbol_frequency, prob, source_len) return result def _calculate_chi_squared(source_freq, target_prob, source_len): """A measure of the observed frequency of the symbol versus the expected frequency. If the value is 0 then the texts are exactly alike for that symbol. """ expected = source_len * target_prob return (source_freq - expected)**2 / expected def _load_ngram(name): """Dynamically import the python module with the ngram defined as a dictionary. Since bigger ngrams are large files its wasteful to always statically import them if they're not used. """ module = importlib.import_module('lantern.analysis.english_ngrams.{}'.format(name)) return getattr(module, name) english = DynamicDict({ 'unigrams': lambda: _load_ngram('unigrams'), 'bigrams': lambda: _load_ngram('bigrams'), 'trigrams': lambda: _load_ngram('trigrams'), 'quadgrams': lambda: _load_ngram('quadgrams') }) """English ngram frequencies.""" ENGLISH_IC = _calculate_index_of_coincidence(english.unigrams, sum(english.unigrams.values())) """Index of coincidence for the English language."""
CameronLonsdale/lantern
lantern/analysis/frequency.py
chi_squared
python
def chi_squared(source_frequency, target_frequency): # Ignore any symbols from source that are not in target. # TODO: raise Error if source_len is 0? target_prob = frequency_to_probability(target_frequency) source_len = sum(v for k, v in source_frequency.items() if k in target_frequency) result = 0 for symbol, prob in target_prob.items(): symbol_frequency = source_frequency.get(symbol, 0) # Frequecy is 0 if it doesnt appear in source result += _calculate_chi_squared(symbol_frequency, prob, source_len) return result
Calculate the Chi Squared statistic by comparing ``source_frequency`` with ``target_frequency``. Example: >>> chi_squared({'a': 2, 'b': 3}, {'a': 1, 'b': 2}) 0.1 Args: source_frequency (dict): Frequency map of the text you are analyzing target_frequency (dict): Frequency map of the target language to compare with Returns: Decimal value of the chi-squared statistic
train
https://github.com/CameronLonsdale/lantern/blob/235e163e96bf0719d49c54204ee576b2ca93abb6/lantern/analysis/frequency.py#L99-L123
[ "def frequency_to_probability(frequency_map, decorator=lambda f: f):\n \"\"\"Transform a ``frequency_map`` into a map of probability using the sum of all frequencies as the total.\n\n Example:\n >>> frequency_to_probability({'a': 2, 'b': 2})\n {'a': 0.5, 'b': 0.5}\n\n Args:\n frequency...
"""General purpose frequency analysis tools.""" import importlib import statistics from collections import Counter from lantern.structures import DynamicDict from lantern.util import iterate_ngrams def frequency_analyze(text, n=1): """Analyze the frequency of ngrams for a piece of text. Examples: >>> frequency_analyze("abb") {'a': 1, 'b': 2} >>> frequency_analyze("abb", 2) {'ab': 1, 'bb': 1} Args: text (str): The text to analyze n (int): The ngram size to use Returns: Dictionary of ngrams to frequency Raises: ValueError: If n is not a positive integer """ return Counter(iterate_ngrams(text, n)) def frequency_to_probability(frequency_map, decorator=lambda f: f): """Transform a ``frequency_map`` into a map of probability using the sum of all frequencies as the total. Example: >>> frequency_to_probability({'a': 2, 'b': 2}) {'a': 0.5, 'b': 0.5} Args: frequency_map (dict): The dictionary to transform decorator (function): A function to manipulate the probability Returns: Dictionary of ngrams to probability """ total = sum(frequency_map.values()) return {k: decorator(v / total) for k, v in frequency_map.items()} def index_of_coincidence(*texts): """Calculate the index of coincidence for one or more ``texts``. The results are averaged over multiple texts to return the delta index of coincidence. Examples: >>> index_of_coincidence("aabbc") 0.2 >>> index_of_coincidence("aabbc", "abbcc") 0.2 Args: *texts (variable length argument list): The texts to analyze Returns: Decimal value of the index of coincidence Raises: ValueError: If texts is empty ValueError: If any text is less that 2 character long """ if not texts: raise ValueError("texts must not be empty") return statistics.mean(_calculate_index_of_coincidence(frequency_analyze(text), len(text)) for text in texts) def _calculate_index_of_coincidence(frequency_map, length): """A measure of how similar frequency_map is to the uniform distribution. Or the probability that two letters picked randomly are alike. """ if length <= 1: return 0 # We cannot error here as length can legitimiately be 1. # Imagine a ciphertext of length 3 and a key of length 2. # Spliting this text up and calculating the index of coincidence results in ['AC', 'B'] # IOC of B will be calcuated for the 2nd column of the key. We could represent the same # encryption with a key of length 3 but then we encounter the same problem. This is also # legitimiate encryption scheme we cannot ignore. Hence we have to deal with this fact here # A value of 0 will impact the overall mean, however it does make some sense when you ask the question # How many ways to choose 2 letters from the text, if theres only 1 letter then the answer is 0. # Mathemtical combination, number of ways to choose 2 letters, no replacement, order doesnt matter combination_of_letters = sum(freq * (freq - 1) for freq in frequency_map.values()) return combination_of_letters / (length * (length - 1)) def _calculate_chi_squared(source_freq, target_prob, source_len): """A measure of the observed frequency of the symbol versus the expected frequency. If the value is 0 then the texts are exactly alike for that symbol. """ expected = source_len * target_prob return (source_freq - expected)**2 / expected def _load_ngram(name): """Dynamically import the python module with the ngram defined as a dictionary. Since bigger ngrams are large files its wasteful to always statically import them if they're not used. """ module = importlib.import_module('lantern.analysis.english_ngrams.{}'.format(name)) return getattr(module, name) english = DynamicDict({ 'unigrams': lambda: _load_ngram('unigrams'), 'bigrams': lambda: _load_ngram('bigrams'), 'trigrams': lambda: _load_ngram('trigrams'), 'quadgrams': lambda: _load_ngram('quadgrams') }) """English ngram frequencies.""" ENGLISH_IC = _calculate_index_of_coincidence(english.unigrams, sum(english.unigrams.values())) """Index of coincidence for the English language."""
CameronLonsdale/lantern
lantern/analysis/frequency.py
_calculate_chi_squared
python
def _calculate_chi_squared(source_freq, target_prob, source_len): expected = source_len * target_prob return (source_freq - expected)**2 / expected
A measure of the observed frequency of the symbol versus the expected frequency. If the value is 0 then the texts are exactly alike for that symbol.
train
https://github.com/CameronLonsdale/lantern/blob/235e163e96bf0719d49c54204ee576b2ca93abb6/lantern/analysis/frequency.py#L126-L131
null
"""General purpose frequency analysis tools.""" import importlib import statistics from collections import Counter from lantern.structures import DynamicDict from lantern.util import iterate_ngrams def frequency_analyze(text, n=1): """Analyze the frequency of ngrams for a piece of text. Examples: >>> frequency_analyze("abb") {'a': 1, 'b': 2} >>> frequency_analyze("abb", 2) {'ab': 1, 'bb': 1} Args: text (str): The text to analyze n (int): The ngram size to use Returns: Dictionary of ngrams to frequency Raises: ValueError: If n is not a positive integer """ return Counter(iterate_ngrams(text, n)) def frequency_to_probability(frequency_map, decorator=lambda f: f): """Transform a ``frequency_map`` into a map of probability using the sum of all frequencies as the total. Example: >>> frequency_to_probability({'a': 2, 'b': 2}) {'a': 0.5, 'b': 0.5} Args: frequency_map (dict): The dictionary to transform decorator (function): A function to manipulate the probability Returns: Dictionary of ngrams to probability """ total = sum(frequency_map.values()) return {k: decorator(v / total) for k, v in frequency_map.items()} def index_of_coincidence(*texts): """Calculate the index of coincidence for one or more ``texts``. The results are averaged over multiple texts to return the delta index of coincidence. Examples: >>> index_of_coincidence("aabbc") 0.2 >>> index_of_coincidence("aabbc", "abbcc") 0.2 Args: *texts (variable length argument list): The texts to analyze Returns: Decimal value of the index of coincidence Raises: ValueError: If texts is empty ValueError: If any text is less that 2 character long """ if not texts: raise ValueError("texts must not be empty") return statistics.mean(_calculate_index_of_coincidence(frequency_analyze(text), len(text)) for text in texts) def _calculate_index_of_coincidence(frequency_map, length): """A measure of how similar frequency_map is to the uniform distribution. Or the probability that two letters picked randomly are alike. """ if length <= 1: return 0 # We cannot error here as length can legitimiately be 1. # Imagine a ciphertext of length 3 and a key of length 2. # Spliting this text up and calculating the index of coincidence results in ['AC', 'B'] # IOC of B will be calcuated for the 2nd column of the key. We could represent the same # encryption with a key of length 3 but then we encounter the same problem. This is also # legitimiate encryption scheme we cannot ignore. Hence we have to deal with this fact here # A value of 0 will impact the overall mean, however it does make some sense when you ask the question # How many ways to choose 2 letters from the text, if theres only 1 letter then the answer is 0. # Mathemtical combination, number of ways to choose 2 letters, no replacement, order doesnt matter combination_of_letters = sum(freq * (freq - 1) for freq in frequency_map.values()) return combination_of_letters / (length * (length - 1)) def chi_squared(source_frequency, target_frequency): """Calculate the Chi Squared statistic by comparing ``source_frequency`` with ``target_frequency``. Example: >>> chi_squared({'a': 2, 'b': 3}, {'a': 1, 'b': 2}) 0.1 Args: source_frequency (dict): Frequency map of the text you are analyzing target_frequency (dict): Frequency map of the target language to compare with Returns: Decimal value of the chi-squared statistic """ # Ignore any symbols from source that are not in target. # TODO: raise Error if source_len is 0? target_prob = frequency_to_probability(target_frequency) source_len = sum(v for k, v in source_frequency.items() if k in target_frequency) result = 0 for symbol, prob in target_prob.items(): symbol_frequency = source_frequency.get(symbol, 0) # Frequecy is 0 if it doesnt appear in source result += _calculate_chi_squared(symbol_frequency, prob, source_len) return result def _load_ngram(name): """Dynamically import the python module with the ngram defined as a dictionary. Since bigger ngrams are large files its wasteful to always statically import them if they're not used. """ module = importlib.import_module('lantern.analysis.english_ngrams.{}'.format(name)) return getattr(module, name) english = DynamicDict({ 'unigrams': lambda: _load_ngram('unigrams'), 'bigrams': lambda: _load_ngram('bigrams'), 'trigrams': lambda: _load_ngram('trigrams'), 'quadgrams': lambda: _load_ngram('quadgrams') }) """English ngram frequencies.""" ENGLISH_IC = _calculate_index_of_coincidence(english.unigrams, sum(english.unigrams.values())) """Index of coincidence for the English language."""
CameronLonsdale/lantern
lantern/analysis/frequency.py
_load_ngram
python
def _load_ngram(name): module = importlib.import_module('lantern.analysis.english_ngrams.{}'.format(name)) return getattr(module, name)
Dynamically import the python module with the ngram defined as a dictionary. Since bigger ngrams are large files its wasteful to always statically import them if they're not used.
train
https://github.com/CameronLonsdale/lantern/blob/235e163e96bf0719d49c54204ee576b2ca93abb6/lantern/analysis/frequency.py#L134-L139
null
"""General purpose frequency analysis tools.""" import importlib import statistics from collections import Counter from lantern.structures import DynamicDict from lantern.util import iterate_ngrams def frequency_analyze(text, n=1): """Analyze the frequency of ngrams for a piece of text. Examples: >>> frequency_analyze("abb") {'a': 1, 'b': 2} >>> frequency_analyze("abb", 2) {'ab': 1, 'bb': 1} Args: text (str): The text to analyze n (int): The ngram size to use Returns: Dictionary of ngrams to frequency Raises: ValueError: If n is not a positive integer """ return Counter(iterate_ngrams(text, n)) def frequency_to_probability(frequency_map, decorator=lambda f: f): """Transform a ``frequency_map`` into a map of probability using the sum of all frequencies as the total. Example: >>> frequency_to_probability({'a': 2, 'b': 2}) {'a': 0.5, 'b': 0.5} Args: frequency_map (dict): The dictionary to transform decorator (function): A function to manipulate the probability Returns: Dictionary of ngrams to probability """ total = sum(frequency_map.values()) return {k: decorator(v / total) for k, v in frequency_map.items()} def index_of_coincidence(*texts): """Calculate the index of coincidence for one or more ``texts``. The results are averaged over multiple texts to return the delta index of coincidence. Examples: >>> index_of_coincidence("aabbc") 0.2 >>> index_of_coincidence("aabbc", "abbcc") 0.2 Args: *texts (variable length argument list): The texts to analyze Returns: Decimal value of the index of coincidence Raises: ValueError: If texts is empty ValueError: If any text is less that 2 character long """ if not texts: raise ValueError("texts must not be empty") return statistics.mean(_calculate_index_of_coincidence(frequency_analyze(text), len(text)) for text in texts) def _calculate_index_of_coincidence(frequency_map, length): """A measure of how similar frequency_map is to the uniform distribution. Or the probability that two letters picked randomly are alike. """ if length <= 1: return 0 # We cannot error here as length can legitimiately be 1. # Imagine a ciphertext of length 3 and a key of length 2. # Spliting this text up and calculating the index of coincidence results in ['AC', 'B'] # IOC of B will be calcuated for the 2nd column of the key. We could represent the same # encryption with a key of length 3 but then we encounter the same problem. This is also # legitimiate encryption scheme we cannot ignore. Hence we have to deal with this fact here # A value of 0 will impact the overall mean, however it does make some sense when you ask the question # How many ways to choose 2 letters from the text, if theres only 1 letter then the answer is 0. # Mathemtical combination, number of ways to choose 2 letters, no replacement, order doesnt matter combination_of_letters = sum(freq * (freq - 1) for freq in frequency_map.values()) return combination_of_letters / (length * (length - 1)) def chi_squared(source_frequency, target_frequency): """Calculate the Chi Squared statistic by comparing ``source_frequency`` with ``target_frequency``. Example: >>> chi_squared({'a': 2, 'b': 3}, {'a': 1, 'b': 2}) 0.1 Args: source_frequency (dict): Frequency map of the text you are analyzing target_frequency (dict): Frequency map of the target language to compare with Returns: Decimal value of the chi-squared statistic """ # Ignore any symbols from source that are not in target. # TODO: raise Error if source_len is 0? target_prob = frequency_to_probability(target_frequency) source_len = sum(v for k, v in source_frequency.items() if k in target_frequency) result = 0 for symbol, prob in target_prob.items(): symbol_frequency = source_frequency.get(symbol, 0) # Frequecy is 0 if it doesnt appear in source result += _calculate_chi_squared(symbol_frequency, prob, source_len) return result def _calculate_chi_squared(source_freq, target_prob, source_len): """A measure of the observed frequency of the symbol versus the expected frequency. If the value is 0 then the texts are exactly alike for that symbol. """ expected = source_len * target_prob return (source_freq - expected)**2 / expected english = DynamicDict({ 'unigrams': lambda: _load_ngram('unigrams'), 'bigrams': lambda: _load_ngram('bigrams'), 'trigrams': lambda: _load_ngram('trigrams'), 'quadgrams': lambda: _load_ngram('quadgrams') }) """English ngram frequencies.""" ENGLISH_IC = _calculate_index_of_coincidence(english.unigrams, sum(english.unigrams.values())) """Index of coincidence for the English language."""
CameronLonsdale/lantern
lantern/score.py
score
python
def score(text, *score_functions): if not score_functions: raise ValueError("score_functions must not be empty") return statistics.mean(func(text) for func in score_functions)
Score ``text`` using ``score_functions``. Examples: >>> score("abc", function_a) >>> score("abc", function_a, function_b) Args: text (str): The text to score *score_functions (variable length argument list): functions to score with Returns: Arithmetic mean of scores Raises: ValueError: If score_functions is empty
train
https://github.com/CameronLonsdale/lantern/blob/235e163e96bf0719d49c54204ee576b2ca93abb6/lantern/score.py#L9-L29
null
""" Scoring algorithm to return probability of correct decryption. Output range depends on the score functions used. """ import statistics
CameronLonsdale/lantern
lantern/fitness/patternmatch.py
PatternMatch
python
def PatternMatch(regex): pattern = re.compile(regex) return lambda text: -1 if pattern.search(text) is None else 0
Compute the score of a text by determing if a pattern matches. Example: >>> fitness = PatternMatch("flag{.*}") >>> fitness("flag{example}") 0 >>> fitness("junk") -1 Args: regex (str): regular expression string to use as a pattern
train
https://github.com/CameronLonsdale/lantern/blob/235e163e96bf0719d49c54204ee576b2ca93abb6/lantern/fitness/patternmatch.py#L6-L21
null
"""Fitness scoring using pattern matching.""" import re
CameronLonsdale/lantern
lantern/util.py
remove
python
def remove(text, exclude): exclude = ''.join(str(symbol) for symbol in exclude) return text.translate(str.maketrans('', '', exclude))
Remove ``exclude`` symbols from ``text``. Example: >>> remove("example text", string.whitespace) 'exampletext' Args: text (str): The text to modify exclude (iterable): The symbols to exclude Returns: ``text`` with ``exclude`` symbols removed
train
https://github.com/CameronLonsdale/lantern/blob/235e163e96bf0719d49c54204ee576b2ca93abb6/lantern/util.py#L6-L21
null
"""Utility functions to format and marshal data.""" import itertools def split_columns(text, n_columns): """Split ``text`` into ``n_columns`` many columns. Example: >>> split_columns("example", 2) ['eape', 'xml'] Args: text (str): The text to split n_columns (int): The number of columns to create Returns: List of columns Raises: ValueError: If n_cols is <= 0 or >= len(text) """ if n_columns <= 0 or n_columns > len(text): raise ValueError("n_columns must be within the bounds of 1 and text length") return [text[i::n_columns] for i in range(n_columns)] def combine_columns(columns): """Combine ``columns`` into a single string. Example: >>> combine_columns(['eape', 'xml']) 'example' Args: columns (iterable): ordered columns to combine Returns: String of combined columns """ columns_zipped = itertools.zip_longest(*columns) return ''.join(x for zipped in columns_zipped for x in zipped if x) def iterate_ngrams(text, n): """Generator to yield ngrams in ``text``. Example: >>> for ngram in iterate_ngrams("example", 4): ... print(ngram) exam xamp ampl mple Args: text (str): text to iterate over n (int): size of window for iteration Returns: Generator expression to yield the next ngram in the text Raises: ValueError: If n is non positive """ if n <= 0: raise ValueError("n must be a positive integer") return [text[i: i + n] for i in range(len(text) - n + 1)] def group(text, size): """Group ``text`` into blocks of ``size``. Example: >>> group("test", 2) ['te', 'st'] Args: text (str): text to separate size (int): size of groups to split the text into Returns: List of n-sized groups of text Raises: ValueError: If n is non positive """ if size <= 0: raise ValueError("n must be a positive integer") return [text[i:i + size] for i in range(0, len(text), size)]
CameronLonsdale/lantern
lantern/util.py
split_columns
python
def split_columns(text, n_columns): if n_columns <= 0 or n_columns > len(text): raise ValueError("n_columns must be within the bounds of 1 and text length") return [text[i::n_columns] for i in range(n_columns)]
Split ``text`` into ``n_columns`` many columns. Example: >>> split_columns("example", 2) ['eape', 'xml'] Args: text (str): The text to split n_columns (int): The number of columns to create Returns: List of columns Raises: ValueError: If n_cols is <= 0 or >= len(text)
train
https://github.com/CameronLonsdale/lantern/blob/235e163e96bf0719d49c54204ee576b2ca93abb6/lantern/util.py#L24-L44
null
"""Utility functions to format and marshal data.""" import itertools def remove(text, exclude): """Remove ``exclude`` symbols from ``text``. Example: >>> remove("example text", string.whitespace) 'exampletext' Args: text (str): The text to modify exclude (iterable): The symbols to exclude Returns: ``text`` with ``exclude`` symbols removed """ exclude = ''.join(str(symbol) for symbol in exclude) return text.translate(str.maketrans('', '', exclude)) def combine_columns(columns): """Combine ``columns`` into a single string. Example: >>> combine_columns(['eape', 'xml']) 'example' Args: columns (iterable): ordered columns to combine Returns: String of combined columns """ columns_zipped = itertools.zip_longest(*columns) return ''.join(x for zipped in columns_zipped for x in zipped if x) def iterate_ngrams(text, n): """Generator to yield ngrams in ``text``. Example: >>> for ngram in iterate_ngrams("example", 4): ... print(ngram) exam xamp ampl mple Args: text (str): text to iterate over n (int): size of window for iteration Returns: Generator expression to yield the next ngram in the text Raises: ValueError: If n is non positive """ if n <= 0: raise ValueError("n must be a positive integer") return [text[i: i + n] for i in range(len(text) - n + 1)] def group(text, size): """Group ``text`` into blocks of ``size``. Example: >>> group("test", 2) ['te', 'st'] Args: text (str): text to separate size (int): size of groups to split the text into Returns: List of n-sized groups of text Raises: ValueError: If n is non positive """ if size <= 0: raise ValueError("n must be a positive integer") return [text[i:i + size] for i in range(0, len(text), size)]
CameronLonsdale/lantern
lantern/util.py
combine_columns
python
def combine_columns(columns): columns_zipped = itertools.zip_longest(*columns) return ''.join(x for zipped in columns_zipped for x in zipped if x)
Combine ``columns`` into a single string. Example: >>> combine_columns(['eape', 'xml']) 'example' Args: columns (iterable): ordered columns to combine Returns: String of combined columns
train
https://github.com/CameronLonsdale/lantern/blob/235e163e96bf0719d49c54204ee576b2ca93abb6/lantern/util.py#L47-L61
null
"""Utility functions to format and marshal data.""" import itertools def remove(text, exclude): """Remove ``exclude`` symbols from ``text``. Example: >>> remove("example text", string.whitespace) 'exampletext' Args: text (str): The text to modify exclude (iterable): The symbols to exclude Returns: ``text`` with ``exclude`` symbols removed """ exclude = ''.join(str(symbol) for symbol in exclude) return text.translate(str.maketrans('', '', exclude)) def split_columns(text, n_columns): """Split ``text`` into ``n_columns`` many columns. Example: >>> split_columns("example", 2) ['eape', 'xml'] Args: text (str): The text to split n_columns (int): The number of columns to create Returns: List of columns Raises: ValueError: If n_cols is <= 0 or >= len(text) """ if n_columns <= 0 or n_columns > len(text): raise ValueError("n_columns must be within the bounds of 1 and text length") return [text[i::n_columns] for i in range(n_columns)] def iterate_ngrams(text, n): """Generator to yield ngrams in ``text``. Example: >>> for ngram in iterate_ngrams("example", 4): ... print(ngram) exam xamp ampl mple Args: text (str): text to iterate over n (int): size of window for iteration Returns: Generator expression to yield the next ngram in the text Raises: ValueError: If n is non positive """ if n <= 0: raise ValueError("n must be a positive integer") return [text[i: i + n] for i in range(len(text) - n + 1)] def group(text, size): """Group ``text`` into blocks of ``size``. Example: >>> group("test", 2) ['te', 'st'] Args: text (str): text to separate size (int): size of groups to split the text into Returns: List of n-sized groups of text Raises: ValueError: If n is non positive """ if size <= 0: raise ValueError("n must be a positive integer") return [text[i:i + size] for i in range(0, len(text), size)]
CameronLonsdale/lantern
lantern/util.py
iterate_ngrams
python
def iterate_ngrams(text, n): if n <= 0: raise ValueError("n must be a positive integer") return [text[i: i + n] for i in range(len(text) - n + 1)]
Generator to yield ngrams in ``text``. Example: >>> for ngram in iterate_ngrams("example", 4): ... print(ngram) exam xamp ampl mple Args: text (str): text to iterate over n (int): size of window for iteration Returns: Generator expression to yield the next ngram in the text Raises: ValueError: If n is non positive
train
https://github.com/CameronLonsdale/lantern/blob/235e163e96bf0719d49c54204ee576b2ca93abb6/lantern/util.py#L64-L88
null
"""Utility functions to format and marshal data.""" import itertools def remove(text, exclude): """Remove ``exclude`` symbols from ``text``. Example: >>> remove("example text", string.whitespace) 'exampletext' Args: text (str): The text to modify exclude (iterable): The symbols to exclude Returns: ``text`` with ``exclude`` symbols removed """ exclude = ''.join(str(symbol) for symbol in exclude) return text.translate(str.maketrans('', '', exclude)) def split_columns(text, n_columns): """Split ``text`` into ``n_columns`` many columns. Example: >>> split_columns("example", 2) ['eape', 'xml'] Args: text (str): The text to split n_columns (int): The number of columns to create Returns: List of columns Raises: ValueError: If n_cols is <= 0 or >= len(text) """ if n_columns <= 0 or n_columns > len(text): raise ValueError("n_columns must be within the bounds of 1 and text length") return [text[i::n_columns] for i in range(n_columns)] def combine_columns(columns): """Combine ``columns`` into a single string. Example: >>> combine_columns(['eape', 'xml']) 'example' Args: columns (iterable): ordered columns to combine Returns: String of combined columns """ columns_zipped = itertools.zip_longest(*columns) return ''.join(x for zipped in columns_zipped for x in zipped if x) def group(text, size): """Group ``text`` into blocks of ``size``. Example: >>> group("test", 2) ['te', 'st'] Args: text (str): text to separate size (int): size of groups to split the text into Returns: List of n-sized groups of text Raises: ValueError: If n is non positive """ if size <= 0: raise ValueError("n must be a positive integer") return [text[i:i + size] for i in range(0, len(text), size)]
CameronLonsdale/lantern
lantern/util.py
group
python
def group(text, size): if size <= 0: raise ValueError("n must be a positive integer") return [text[i:i + size] for i in range(0, len(text), size)]
Group ``text`` into blocks of ``size``. Example: >>> group("test", 2) ['te', 'st'] Args: text (str): text to separate size (int): size of groups to split the text into Returns: List of n-sized groups of text Raises: ValueError: If n is non positive
train
https://github.com/CameronLonsdale/lantern/blob/235e163e96bf0719d49c54204ee576b2ca93abb6/lantern/util.py#L91-L111
null
"""Utility functions to format and marshal data.""" import itertools def remove(text, exclude): """Remove ``exclude`` symbols from ``text``. Example: >>> remove("example text", string.whitespace) 'exampletext' Args: text (str): The text to modify exclude (iterable): The symbols to exclude Returns: ``text`` with ``exclude`` symbols removed """ exclude = ''.join(str(symbol) for symbol in exclude) return text.translate(str.maketrans('', '', exclude)) def split_columns(text, n_columns): """Split ``text`` into ``n_columns`` many columns. Example: >>> split_columns("example", 2) ['eape', 'xml'] Args: text (str): The text to split n_columns (int): The number of columns to create Returns: List of columns Raises: ValueError: If n_cols is <= 0 or >= len(text) """ if n_columns <= 0 or n_columns > len(text): raise ValueError("n_columns must be within the bounds of 1 and text length") return [text[i::n_columns] for i in range(n_columns)] def combine_columns(columns): """Combine ``columns`` into a single string. Example: >>> combine_columns(['eape', 'xml']) 'example' Args: columns (iterable): ordered columns to combine Returns: String of combined columns """ columns_zipped = itertools.zip_longest(*columns) return ''.join(x for zipped in columns_zipped for x in zipped if x) def iterate_ngrams(text, n): """Generator to yield ngrams in ``text``. Example: >>> for ngram in iterate_ngrams("example", 4): ... print(ngram) exam xamp ampl mple Args: text (str): text to iterate over n (int): size of window for iteration Returns: Generator expression to yield the next ngram in the text Raises: ValueError: If n is non positive """ if n <= 0: raise ValueError("n must be a positive integer") return [text[i: i + n] for i in range(len(text) - n + 1)]
CameronLonsdale/lantern
lantern/fitness/chisquared.py
ChiSquared
python
def ChiSquared(target_frequency): def inner(text): text = ''.join(text) return -chi_squared(frequency_analyze(text), target_frequency) return inner
Score a text by comparing its frequency distribution against another. Note: It is easy to be penalised without knowing it when using this scorer. English frequency ngrams are capital letters, meaning when using it any text you score against must be all capitals for it to give correct results. I am aware of the issue and will work on a fix. Todo: Maybe include paramter for ngram size. Havent had a use case for this yet. Once there is evidence it is needed, I will add it. Example: >>> fitness = ChiSquared(english.unigrams) >>> fitness("ABC") -32.2 Args: target_frequency (dict): symbol to frequency mapping of the distribution to compare with
train
https://github.com/CameronLonsdale/lantern/blob/235e163e96bf0719d49c54204ee576b2ca93abb6/lantern/fitness/chisquared.py#L6-L31
null
"""Chi Squared Scoring function.""" from lantern.analysis.frequency import frequency_analyze, chi_squared
CameronLonsdale/lantern
lantern/modules/simplesubstitution.py
crack
python
def crack(ciphertext, *fitness_functions, ntrials=30, nswaps=3000): if ntrials <= 0 or nswaps <= 0: raise ValueError("ntrials and nswaps must be positive integers") # Find a local maximum by swapping two letters and scoring the decryption def next_node_inner_climb(node): # Swap 2 characters in the key a, b = random.sample(range(len(node)), 2) node[a], node[b] = node[b], node[a] plaintext = decrypt(node, ciphertext) node_score = score(plaintext, *fitness_functions) return node, node_score, Decryption(plaintext, ''.join(node), node_score) # Outer climb rereuns hill climb ntrials number of times each time at a different start location def next_node_outer_climb(node): random.shuffle(node) key, best_score, outputs = hill_climb(nswaps, node[:], next_node_inner_climb) return key, best_score, outputs[-1] # The last item in this list is the item with the highest score _, _, decryptions = hill_climb(ntrials, list(string.ascii_uppercase), next_node_outer_climb) return sorted(decryptions, reverse=True)
Break ``ciphertext`` using hill climbing. Note: Currently ntrails and nswaps default to magic numbers. Generally the trend is, the longer the text, the lower the number of trials you need to run, because the hill climbing will lead to the best answer faster. Because randomness is involved, there is the possibility of the correct decryption not being found. In this circumstance you just need to run the code again. Example: >>> decryptions = crack("XUOOB", fitness.english.quadgrams) >>> print(decryptions[0]) HELLO Args: ciphertext (str): The text to decrypt *fitness_functions (variable length argument list): Functions to score decryption with Keyword Args: ntrials (int): The number of times to run the hill climbing algorithm nswaps (int): The number of rounds to find a local maximum Returns: Sorted list of decryptions Raises: ValueError: If nswaps or ntrails are not positive integers ValueError: If no fitness_functions are given
train
https://github.com/CameronLonsdale/lantern/blob/235e163e96bf0719d49c54204ee576b2ca93abb6/lantern/modules/simplesubstitution.py#L11-L60
[ "def hill_climb(nsteps, start_node, get_next_node):\n \"\"\"Modular hill climbing algorithm.\n\n Example:\n >>> def get_next_node(node):\n ... a, b = random.sample(range(len(node)), 2)\n ... node[a], node[b] = node[b], node[a]\n ... plaintext = decrypt(node, ciphertext)...
"""Automated breaking of the Simple Substitution Cipher.""" import random import string from lantern import score from lantern.analysis.search import hill_climb from lantern.structures import Decryption # We sort the list to ensure the best results are at the front of the list def decrypt(key, ciphertext): """Decrypt Simple Substitution enciphered ``ciphertext`` using ``key``. Example: >>> decrypt("PQSTUVWXYZCODEBRAKINGFHJLM", "XUOOB") HELLO Args: key (iterable): The key to use ciphertext (str): The text to decrypt Returns: Decrypted ciphertext """ # TODO: Is it worth keeping this here I should I only accept strings? key = ''.join(key) alphabet = string.ascii_letters cipher_alphabet = key.lower() + key.upper() return ciphertext.translate(str.maketrans(cipher_alphabet, alphabet))
CameronLonsdale/lantern
lantern/modules/simplesubstitution.py
decrypt
python
def decrypt(key, ciphertext): # TODO: Is it worth keeping this here I should I only accept strings? key = ''.join(key) alphabet = string.ascii_letters cipher_alphabet = key.lower() + key.upper() return ciphertext.translate(str.maketrans(cipher_alphabet, alphabet))
Decrypt Simple Substitution enciphered ``ciphertext`` using ``key``. Example: >>> decrypt("PQSTUVWXYZCODEBRAKINGFHJLM", "XUOOB") HELLO Args: key (iterable): The key to use ciphertext (str): The text to decrypt Returns: Decrypted ciphertext
train
https://github.com/CameronLonsdale/lantern/blob/235e163e96bf0719d49c54204ee576b2ca93abb6/lantern/modules/simplesubstitution.py#L63-L81
null
"""Automated breaking of the Simple Substitution Cipher.""" import random import string from lantern import score from lantern.analysis.search import hill_climb from lantern.structures import Decryption def crack(ciphertext, *fitness_functions, ntrials=30, nswaps=3000): """Break ``ciphertext`` using hill climbing. Note: Currently ntrails and nswaps default to magic numbers. Generally the trend is, the longer the text, the lower the number of trials you need to run, because the hill climbing will lead to the best answer faster. Because randomness is involved, there is the possibility of the correct decryption not being found. In this circumstance you just need to run the code again. Example: >>> decryptions = crack("XUOOB", fitness.english.quadgrams) >>> print(decryptions[0]) HELLO Args: ciphertext (str): The text to decrypt *fitness_functions (variable length argument list): Functions to score decryption with Keyword Args: ntrials (int): The number of times to run the hill climbing algorithm nswaps (int): The number of rounds to find a local maximum Returns: Sorted list of decryptions Raises: ValueError: If nswaps or ntrails are not positive integers ValueError: If no fitness_functions are given """ if ntrials <= 0 or nswaps <= 0: raise ValueError("ntrials and nswaps must be positive integers") # Find a local maximum by swapping two letters and scoring the decryption def next_node_inner_climb(node): # Swap 2 characters in the key a, b = random.sample(range(len(node)), 2) node[a], node[b] = node[b], node[a] plaintext = decrypt(node, ciphertext) node_score = score(plaintext, *fitness_functions) return node, node_score, Decryption(plaintext, ''.join(node), node_score) # Outer climb rereuns hill climb ntrials number of times each time at a different start location def next_node_outer_climb(node): random.shuffle(node) key, best_score, outputs = hill_climb(nswaps, node[:], next_node_inner_climb) return key, best_score, outputs[-1] # The last item in this list is the item with the highest score _, _, decryptions = hill_climb(ntrials, list(string.ascii_uppercase), next_node_outer_climb) return sorted(decryptions, reverse=True) # We sort the list to ensure the best results are at the front of the list
CameronLonsdale/lantern
lantern/fitness/ngram.py
NgramScorer
python
def NgramScorer(frequency_map): # Calculate the log probability length = len(next(iter(frequency_map))) # TODO: 0.01 is a magic number. Needs to be better than that. floor = math.log10(0.01 / sum(frequency_map.values())) ngrams = frequency.frequency_to_probability(frequency_map, decorator=math.log10) def inner(text): # I dont like this, it is only for the .upper() to work, # But I feel as though this can be removed in later refactoring text = ''.join(text) text = remove(text.upper(), string.whitespace + string.punctuation) return sum(ngrams.get(ngram, floor) for ngram in iterate_ngrams(text, length)) return inner
Compute the score of a text by using the frequencies of ngrams. Example: >>> fitness = NgramScorer(english.unigrams) >>> fitness("ABC") -4.3622319742618245 Args: frequency_map (dict): ngram to frequency mapping
train
https://github.com/CameronLonsdale/lantern/blob/235e163e96bf0719d49c54204ee576b2ca93abb6/lantern/fitness/ngram.py#L11-L35
[ "def frequency_to_probability(frequency_map, decorator=lambda f: f):\n \"\"\"Transform a ``frequency_map`` into a map of probability using the sum of all frequencies as the total.\n\n Example:\n >>> frequency_to_probability({'a': 2, 'b': 2})\n {'a': 0.5, 'b': 0.5}\n\n Args:\n frequency...
"""Fitness scoring using ngram frequency.""" import math import string from lantern.analysis import frequency from lantern.structures import DynamicDict from lantern.util import remove, iterate_ngrams english = DynamicDict({ 'unigrams': lambda: NgramScorer(frequency.english.unigrams), 'bigrams': lambda: NgramScorer(frequency.english.bigrams), 'trigrams': lambda: NgramScorer(frequency.english.trigrams), 'quadgrams': lambda: NgramScorer(frequency.english.quadgrams) }) """English ngram scorers."""
CameronLonsdale/lantern
lantern/modules/vigenere.py
crack
python
def crack(ciphertext, *fitness_functions, key_period=None, max_key_period=30): if max_key_period <= 0 or (key_period is not None and key_period <= 0): raise ValueError("Period values must be positive integers") original_text = ciphertext # Make the assumption that non alphabet characters have not been encrypted # TODO: This is fairly poor code. Once languages are a thing, there should be some nice abstractions for this stuff ciphertext = remove(ciphertext, string.punctuation + string.whitespace + string.digits) periods = [int(key_period)] if key_period else key_periods(ciphertext, max_key_period) # Decrypt for every valid period period_decryptions = [] for period in periods: if period >= len(ciphertext): continue # Collect the best decryptions for every column column_decryptions = [shift.crack(col, *fitness_functions)[0] for col in split_columns(ciphertext, period)] key = _build_key(decrypt.key for decrypt in column_decryptions) plaintext = decrypt(key, original_text) period_decryptions.append(Decryption(plaintext, key, score(plaintext, *fitness_functions))) return sorted(period_decryptions, reverse=True)
Break ``ciphertext`` by finding (or using the given) key_period then breaking ``key_period`` many Caesar ciphers. Example: >>> decryptions = crack("OMSTV", fitness.ChiSquared(analysis.frequency.english.unigrams)) >>> print(decryptions[0]) HELLO Args: ciphertext (str): The text to decrypt *fitness_functions (variable length argument list): Functions to score decryption with Keyword Args: key_period (int): The period of the key max_key_period (int): The maximum period the key could be Returns: Sorted list of decryptions Raises: ValueError: If key_period or max_key_period are less than or equal to 0 ValueError: If no fitness_functions are given
train
https://github.com/CameronLonsdale/lantern/blob/235e163e96bf0719d49c54204ee576b2ca93abb6/lantern/modules/vigenere.py#L15-L60
[ "def remove(text, exclude):\n \"\"\"Remove ``exclude`` symbols from ``text``.\n\n Example:\n >>> remove(\"example text\", string.whitespace)\n 'exampletext'\n\n Args:\n text (str): The text to modify\n exclude (iterable): The symbols to exclude\n\n Returns:\n ``text`` ...
"""Automated breaking of the Vigenere Cipher.""" import string from lantern import score from lantern.modules import shift from lantern.structures import Decryption from lantern.analysis.frequency import index_of_coincidence, ENGLISH_IC from lantern.util import split_columns, remove # TODO: maybe add finding keyperiods as a parameter because people might want to use kasiski # TODO: Maybe consider splitting the scoring functions for columns and the whole # Name should be different?, say youre finding key periods through IC. def key_periods(ciphertext, max_key_period): """Rank all key periods for ``ciphertext`` up to and including ``max_key_period`` Example: >>> key_periods(ciphertext, 30) [2, 4, 8, 3, ...] Args: ciphertext (str): The text to analyze max_key_period (int): The maximum period the key could be Returns: Sorted list of keys Raises: ValueError: If max_key_period is less than or equal to 0 """ if max_key_period <= 0: raise ValueError("max_key_period must be a positive integer") key_scores = [] for period in range(1, min(max_key_period, len(ciphertext)) + 1): score = abs(ENGLISH_IC - index_of_coincidence(*split_columns(ciphertext, period))) key_scores.append((period, score)) return [p[0] for p in sorted(key_scores, key=lambda x: x[1])] def _build_key(keys): num_letters = len(string.ascii_uppercase) return ''.join(string.ascii_uppercase[(key) % num_letters] for key in keys) def decrypt(key, ciphertext): """Decrypt Vigenere encrypted ``ciphertext`` using ``key``. Example: >>> decrypt("KEY", "RIJVS") HELLO Args: key (iterable): The key to use ciphertext (str): The text to decrypt Returns: Decrypted ciphertext """ index = 0 decrypted = "" for char in ciphertext: if char in string.punctuation + string.whitespace + string.digits: decrypted += char continue # Not part of the decryption # Rotate character by the alphabet position of the letter in the key alphabet = string.ascii_uppercase if key[index].isupper() else string.ascii_lowercase decrypted += ''.join(shift.decrypt(int(alphabet.index(key[index])), char)) index = (index + 1) % len(key) return decrypted
CameronLonsdale/lantern
lantern/modules/vigenere.py
key_periods
python
def key_periods(ciphertext, max_key_period): if max_key_period <= 0: raise ValueError("max_key_period must be a positive integer") key_scores = [] for period in range(1, min(max_key_period, len(ciphertext)) + 1): score = abs(ENGLISH_IC - index_of_coincidence(*split_columns(ciphertext, period))) key_scores.append((period, score)) return [p[0] for p in sorted(key_scores, key=lambda x: x[1])]
Rank all key periods for ``ciphertext`` up to and including ``max_key_period`` Example: >>> key_periods(ciphertext, 30) [2, 4, 8, 3, ...] Args: ciphertext (str): The text to analyze max_key_period (int): The maximum period the key could be Returns: Sorted list of keys Raises: ValueError: If max_key_period is less than or equal to 0
train
https://github.com/CameronLonsdale/lantern/blob/235e163e96bf0719d49c54204ee576b2ca93abb6/lantern/modules/vigenere.py#L64-L89
[ "def index_of_coincidence(*texts):\n \"\"\"Calculate the index of coincidence for one or more ``texts``.\n The results are averaged over multiple texts to return the delta index of coincidence.\n\n Examples:\n >>> index_of_coincidence(\"aabbc\")\n 0.2\n\n >>> index_of_coincidence(\"aab...
"""Automated breaking of the Vigenere Cipher.""" import string from lantern import score from lantern.modules import shift from lantern.structures import Decryption from lantern.analysis.frequency import index_of_coincidence, ENGLISH_IC from lantern.util import split_columns, remove # TODO: maybe add finding keyperiods as a parameter because people might want to use kasiski # TODO: Maybe consider splitting the scoring functions for columns and the whole def crack(ciphertext, *fitness_functions, key_period=None, max_key_period=30): """Break ``ciphertext`` by finding (or using the given) key_period then breaking ``key_period`` many Caesar ciphers. Example: >>> decryptions = crack("OMSTV", fitness.ChiSquared(analysis.frequency.english.unigrams)) >>> print(decryptions[0]) HELLO Args: ciphertext (str): The text to decrypt *fitness_functions (variable length argument list): Functions to score decryption with Keyword Args: key_period (int): The period of the key max_key_period (int): The maximum period the key could be Returns: Sorted list of decryptions Raises: ValueError: If key_period or max_key_period are less than or equal to 0 ValueError: If no fitness_functions are given """ if max_key_period <= 0 or (key_period is not None and key_period <= 0): raise ValueError("Period values must be positive integers") original_text = ciphertext # Make the assumption that non alphabet characters have not been encrypted # TODO: This is fairly poor code. Once languages are a thing, there should be some nice abstractions for this stuff ciphertext = remove(ciphertext, string.punctuation + string.whitespace + string.digits) periods = [int(key_period)] if key_period else key_periods(ciphertext, max_key_period) # Decrypt for every valid period period_decryptions = [] for period in periods: if period >= len(ciphertext): continue # Collect the best decryptions for every column column_decryptions = [shift.crack(col, *fitness_functions)[0] for col in split_columns(ciphertext, period)] key = _build_key(decrypt.key for decrypt in column_decryptions) plaintext = decrypt(key, original_text) period_decryptions.append(Decryption(plaintext, key, score(plaintext, *fitness_functions))) return sorted(period_decryptions, reverse=True) # Name should be different?, say youre finding key periods through IC. def _build_key(keys): num_letters = len(string.ascii_uppercase) return ''.join(string.ascii_uppercase[(key) % num_letters] for key in keys) def decrypt(key, ciphertext): """Decrypt Vigenere encrypted ``ciphertext`` using ``key``. Example: >>> decrypt("KEY", "RIJVS") HELLO Args: key (iterable): The key to use ciphertext (str): The text to decrypt Returns: Decrypted ciphertext """ index = 0 decrypted = "" for char in ciphertext: if char in string.punctuation + string.whitespace + string.digits: decrypted += char continue # Not part of the decryption # Rotate character by the alphabet position of the letter in the key alphabet = string.ascii_uppercase if key[index].isupper() else string.ascii_lowercase decrypted += ''.join(shift.decrypt(int(alphabet.index(key[index])), char)) index = (index + 1) % len(key) return decrypted
CameronLonsdale/lantern
lantern/modules/vigenere.py
decrypt
python
def decrypt(key, ciphertext): index = 0 decrypted = "" for char in ciphertext: if char in string.punctuation + string.whitespace + string.digits: decrypted += char continue # Not part of the decryption # Rotate character by the alphabet position of the letter in the key alphabet = string.ascii_uppercase if key[index].isupper() else string.ascii_lowercase decrypted += ''.join(shift.decrypt(int(alphabet.index(key[index])), char)) index = (index + 1) % len(key) return decrypted
Decrypt Vigenere encrypted ``ciphertext`` using ``key``. Example: >>> decrypt("KEY", "RIJVS") HELLO Args: key (iterable): The key to use ciphertext (str): The text to decrypt Returns: Decrypted ciphertext
train
https://github.com/CameronLonsdale/lantern/blob/235e163e96bf0719d49c54204ee576b2ca93abb6/lantern/modules/vigenere.py#L97-L123
[ "def decrypt(key, ciphertext, shift_function=shift_case_english):\n \"\"\"Decrypt Shift enciphered ``ciphertext`` using ``key``.\n\n Examples:\n >>> ''.join(decrypt(3, \"KHOOR\"))\n HELLO\n\n >> decrypt(15, [0xcf, 0x9e, 0xaf, 0xe0], shift_bytes)\n [0xde, 0xad, 0xbe, 0xef]\n\n Ar...
"""Automated breaking of the Vigenere Cipher.""" import string from lantern import score from lantern.modules import shift from lantern.structures import Decryption from lantern.analysis.frequency import index_of_coincidence, ENGLISH_IC from lantern.util import split_columns, remove # TODO: maybe add finding keyperiods as a parameter because people might want to use kasiski # TODO: Maybe consider splitting the scoring functions for columns and the whole def crack(ciphertext, *fitness_functions, key_period=None, max_key_period=30): """Break ``ciphertext`` by finding (or using the given) key_period then breaking ``key_period`` many Caesar ciphers. Example: >>> decryptions = crack("OMSTV", fitness.ChiSquared(analysis.frequency.english.unigrams)) >>> print(decryptions[0]) HELLO Args: ciphertext (str): The text to decrypt *fitness_functions (variable length argument list): Functions to score decryption with Keyword Args: key_period (int): The period of the key max_key_period (int): The maximum period the key could be Returns: Sorted list of decryptions Raises: ValueError: If key_period or max_key_period are less than or equal to 0 ValueError: If no fitness_functions are given """ if max_key_period <= 0 or (key_period is not None and key_period <= 0): raise ValueError("Period values must be positive integers") original_text = ciphertext # Make the assumption that non alphabet characters have not been encrypted # TODO: This is fairly poor code. Once languages are a thing, there should be some nice abstractions for this stuff ciphertext = remove(ciphertext, string.punctuation + string.whitespace + string.digits) periods = [int(key_period)] if key_period else key_periods(ciphertext, max_key_period) # Decrypt for every valid period period_decryptions = [] for period in periods: if period >= len(ciphertext): continue # Collect the best decryptions for every column column_decryptions = [shift.crack(col, *fitness_functions)[0] for col in split_columns(ciphertext, period)] key = _build_key(decrypt.key for decrypt in column_decryptions) plaintext = decrypt(key, original_text) period_decryptions.append(Decryption(plaintext, key, score(plaintext, *fitness_functions))) return sorted(period_decryptions, reverse=True) # Name should be different?, say youre finding key periods through IC. def key_periods(ciphertext, max_key_period): """Rank all key periods for ``ciphertext`` up to and including ``max_key_period`` Example: >>> key_periods(ciphertext, 30) [2, 4, 8, 3, ...] Args: ciphertext (str): The text to analyze max_key_period (int): The maximum period the key could be Returns: Sorted list of keys Raises: ValueError: If max_key_period is less than or equal to 0 """ if max_key_period <= 0: raise ValueError("max_key_period must be a positive integer") key_scores = [] for period in range(1, min(max_key_period, len(ciphertext)) + 1): score = abs(ENGLISH_IC - index_of_coincidence(*split_columns(ciphertext, period))) key_scores.append((period, score)) return [p[0] for p in sorted(key_scores, key=lambda x: x[1])] def _build_key(keys): num_letters = len(string.ascii_uppercase) return ''.join(string.ascii_uppercase[(key) % num_letters] for key in keys)
CameronLonsdale/lantern
lantern/analysis/search.py
hill_climb
python
def hill_climb(nsteps, start_node, get_next_node): outputs = [] best_score = -float('inf') for step in range(nsteps): next_node, score, output = get_next_node(copy.deepcopy(start_node)) # Keep track of best score and the start node becomes finish node if score > best_score: start_node = copy.deepcopy(next_node) best_score = score outputs.append(output) return start_node, best_score, outputs
Modular hill climbing algorithm. Example: >>> def get_next_node(node): ... a, b = random.sample(range(len(node)), 2) ... node[a], node[b] = node[b], node[a] ... plaintext = decrypt(node, ciphertext) ... score = lantern.score(plaintext, *fitness_functions) ... return node, score, Decryption(plaintext, ''.join(node), score) >>> final_node, best_score, outputs = hill_climb(10, "ABC", get_next_node) Args: nsteps (int): The number of neighbours to visit start_node: The starting node get_next_node (function): Function to return the next node the score of the current node and any optional output from the current node Returns: The highest node found, the score of this node and the outputs from the best nodes along the way
train
https://github.com/CameronLonsdale/lantern/blob/235e163e96bf0719d49c54204ee576b2ca93abb6/lantern/analysis/search.py#L6-L39
[ "def next_node_outer_climb(node):\n random.shuffle(node)\n key, best_score, outputs = hill_climb(nswaps, node[:], next_node_inner_climb)\n return key, best_score, outputs[-1] # The last item in this list is the item with the highest score\n", "def get_next_node(node):\n a, b = random.sample(range(len...
"""Algorithms for searching and optimisation.""" import copy
CameronLonsdale/lantern
lantern/modules/shift.py
make_shift_function
python
def make_shift_function(alphabet): def shift_case_sensitive(shift, symbol): case = [case for case in alphabet if symbol in case] if not case: return symbol case = case[0] index = case.index(symbol) return case[(index - shift) % len(case)] return shift_case_sensitive
Construct a shift function from an alphabet. Examples: Shift cases independently >>> make_shift_function([string.ascii_uppercase, string.ascii_lowercase]) <function make_shift_function.<locals>.shift_case_sensitive> Additionally shift punctuation characters >>> make_shift_function([string.ascii_uppercase, string.ascii_lowercase, string.punctuation]) <function make_shift_function.<locals>.shift_case_sensitive> Shift entire ASCII range, overflowing cases >>> make_shift_function([''.join(chr(x) for x in range(32, 127))]) <function make_shift_function.<locals>.shift_case_sensitive> Args: alphabet (iterable): Ordered iterable of strings representing separate cases of an alphabet Returns: Function (shift, symbol)
train
https://github.com/CameronLonsdale/lantern/blob/235e163e96bf0719d49c54204ee576b2ca93abb6/lantern/modules/shift.py#L9-L43
null
"""Automated breaking of the Shift Cipher.""" import string from lantern import score from lantern.structures import Decryption shift_case_english = make_shift_function([string.ascii_uppercase, string.ascii_lowercase]) def crack(ciphertext, *fitness_functions, min_key=0, max_key=26, shift_function=shift_case_english): """Break ``ciphertext`` by enumerating keys between ``min_key`` and ``max_key``. Example: >>> decryptions = crack("KHOOR", fitness.english.quadgrams) >>> print(''.join(decryptions[0].plaintext)) HELLO Args: ciphertext (iterable): The symbols to decrypt *fitness_functions (variable length argument list): Functions to score decryption with Keyword Args: min_key (int): Key to start with max_key (int): Key to stop at (exclusive) shift_function (function(shift, symbol)): Shift function to use Returns: Sorted list of decryptions Raises: ValueError: If min_key exceeds max_key ValueError: If no fitness_functions are given """ if min_key >= max_key: raise ValueError("min_key cannot exceed max_key") decryptions = [] for key in range(min_key, max_key): plaintext = decrypt(key, ciphertext, shift_function=shift_function) decryptions.append(Decryption(plaintext, key, score(plaintext, *fitness_functions))) return sorted(decryptions, reverse=True) def decrypt(key, ciphertext, shift_function=shift_case_english): """Decrypt Shift enciphered ``ciphertext`` using ``key``. Examples: >>> ''.join(decrypt(3, "KHOOR")) HELLO >> decrypt(15, [0xcf, 0x9e, 0xaf, 0xe0], shift_bytes) [0xde, 0xad, 0xbe, 0xef] Args: key (int): The shift to use ciphertext (iterable): The symbols to decrypt shift_function (function (shift, symbol)): Shift function to apply to symbols in the ciphertext Returns: Decrypted ciphertext, list of plaintext symbols """ return [shift_function(key, symbol) for symbol in ciphertext]
CameronLonsdale/lantern
lantern/modules/shift.py
crack
python
def crack(ciphertext, *fitness_functions, min_key=0, max_key=26, shift_function=shift_case_english): if min_key >= max_key: raise ValueError("min_key cannot exceed max_key") decryptions = [] for key in range(min_key, max_key): plaintext = decrypt(key, ciphertext, shift_function=shift_function) decryptions.append(Decryption(plaintext, key, score(plaintext, *fitness_functions))) return sorted(decryptions, reverse=True)
Break ``ciphertext`` by enumerating keys between ``min_key`` and ``max_key``. Example: >>> decryptions = crack("KHOOR", fitness.english.quadgrams) >>> print(''.join(decryptions[0].plaintext)) HELLO Args: ciphertext (iterable): The symbols to decrypt *fitness_functions (variable length argument list): Functions to score decryption with Keyword Args: min_key (int): Key to start with max_key (int): Key to stop at (exclusive) shift_function (function(shift, symbol)): Shift function to use Returns: Sorted list of decryptions Raises: ValueError: If min_key exceeds max_key ValueError: If no fitness_functions are given
train
https://github.com/CameronLonsdale/lantern/blob/235e163e96bf0719d49c54204ee576b2ca93abb6/lantern/modules/shift.py#L48-L80
[ "def score(text, *score_functions):\n \"\"\"Score ``text`` using ``score_functions``.\n\n Examples:\n >>> score(\"abc\", function_a)\n >>> score(\"abc\", function_a, function_b)\n\n Args:\n text (str): The text to score\n *score_functions (variable length argument list): functio...
"""Automated breaking of the Shift Cipher.""" import string from lantern import score from lantern.structures import Decryption def make_shift_function(alphabet): """Construct a shift function from an alphabet. Examples: Shift cases independently >>> make_shift_function([string.ascii_uppercase, string.ascii_lowercase]) <function make_shift_function.<locals>.shift_case_sensitive> Additionally shift punctuation characters >>> make_shift_function([string.ascii_uppercase, string.ascii_lowercase, string.punctuation]) <function make_shift_function.<locals>.shift_case_sensitive> Shift entire ASCII range, overflowing cases >>> make_shift_function([''.join(chr(x) for x in range(32, 127))]) <function make_shift_function.<locals>.shift_case_sensitive> Args: alphabet (iterable): Ordered iterable of strings representing separate cases of an alphabet Returns: Function (shift, symbol) """ def shift_case_sensitive(shift, symbol): case = [case for case in alphabet if symbol in case] if not case: return symbol case = case[0] index = case.index(symbol) return case[(index - shift) % len(case)] return shift_case_sensitive shift_case_english = make_shift_function([string.ascii_uppercase, string.ascii_lowercase]) def decrypt(key, ciphertext, shift_function=shift_case_english): """Decrypt Shift enciphered ``ciphertext`` using ``key``. Examples: >>> ''.join(decrypt(3, "KHOOR")) HELLO >> decrypt(15, [0xcf, 0x9e, 0xaf, 0xe0], shift_bytes) [0xde, 0xad, 0xbe, 0xef] Args: key (int): The shift to use ciphertext (iterable): The symbols to decrypt shift_function (function (shift, symbol)): Shift function to apply to symbols in the ciphertext Returns: Decrypted ciphertext, list of plaintext symbols """ return [shift_function(key, symbol) for symbol in ciphertext]
CameronLonsdale/lantern
lantern/modules/shift.py
decrypt
python
def decrypt(key, ciphertext, shift_function=shift_case_english): return [shift_function(key, symbol) for symbol in ciphertext]
Decrypt Shift enciphered ``ciphertext`` using ``key``. Examples: >>> ''.join(decrypt(3, "KHOOR")) HELLO >> decrypt(15, [0xcf, 0x9e, 0xaf, 0xe0], shift_bytes) [0xde, 0xad, 0xbe, 0xef] Args: key (int): The shift to use ciphertext (iterable): The symbols to decrypt shift_function (function (shift, symbol)): Shift function to apply to symbols in the ciphertext Returns: Decrypted ciphertext, list of plaintext symbols
train
https://github.com/CameronLonsdale/lantern/blob/235e163e96bf0719d49c54204ee576b2ca93abb6/lantern/modules/shift.py#L83-L101
null
"""Automated breaking of the Shift Cipher.""" import string from lantern import score from lantern.structures import Decryption def make_shift_function(alphabet): """Construct a shift function from an alphabet. Examples: Shift cases independently >>> make_shift_function([string.ascii_uppercase, string.ascii_lowercase]) <function make_shift_function.<locals>.shift_case_sensitive> Additionally shift punctuation characters >>> make_shift_function([string.ascii_uppercase, string.ascii_lowercase, string.punctuation]) <function make_shift_function.<locals>.shift_case_sensitive> Shift entire ASCII range, overflowing cases >>> make_shift_function([''.join(chr(x) for x in range(32, 127))]) <function make_shift_function.<locals>.shift_case_sensitive> Args: alphabet (iterable): Ordered iterable of strings representing separate cases of an alphabet Returns: Function (shift, symbol) """ def shift_case_sensitive(shift, symbol): case = [case for case in alphabet if symbol in case] if not case: return symbol case = case[0] index = case.index(symbol) return case[(index - shift) % len(case)] return shift_case_sensitive shift_case_english = make_shift_function([string.ascii_uppercase, string.ascii_lowercase]) def crack(ciphertext, *fitness_functions, min_key=0, max_key=26, shift_function=shift_case_english): """Break ``ciphertext`` by enumerating keys between ``min_key`` and ``max_key``. Example: >>> decryptions = crack("KHOOR", fitness.english.quadgrams) >>> print(''.join(decryptions[0].plaintext)) HELLO Args: ciphertext (iterable): The symbols to decrypt *fitness_functions (variable length argument list): Functions to score decryption with Keyword Args: min_key (int): Key to start with max_key (int): Key to stop at (exclusive) shift_function (function(shift, symbol)): Shift function to use Returns: Sorted list of decryptions Raises: ValueError: If min_key exceeds max_key ValueError: If no fitness_functions are given """ if min_key >= max_key: raise ValueError("min_key cannot exceed max_key") decryptions = [] for key in range(min_key, max_key): plaintext = decrypt(key, ciphertext, shift_function=shift_function) decryptions.append(Decryption(plaintext, key, score(plaintext, *fitness_functions))) return sorted(decryptions, reverse=True)
tjguk/networkzero
misc/pyconuk2017/robotics/robot/robot.py
Robot.get_command
python
def get_command(self): try: message_bytes = self.socket.recv(zmq.NOBLOCK) log.debug("Received message: %r", message_bytes) except zmq.ZMQError as exc: if exc.errno == zmq.EAGAIN: return None else: raise else: return message_bytes.decode(config.CODEC)
Attempt to return a unicode object from the command socket If no message is available without blocking (as opposed to a blank message), return None
train
https://github.com/tjguk/networkzero/blob/0e3e81d2e9200b25a83ac07741612283599486d7/misc/pyconuk2017/robotics/robot/robot.py#L39-L54
null
class Robot(object): def __init__( self, output, stop_event=None, listen_on_ip=config.LISTEN_ON_IP, listen_on_port=config.LISTEN_ON_PORT ): log.info("Setting up Robot on %s:%s", listen_on_ip, listen_on_port) log.info("Outputting to %s", output) self.stop_event = stop_event or threading.Event() self._init_socket(listen_on_ip, listen_on_port) self.output = output self.output._init() def _init_socket(self, listen_on_ip, listen_on_port): context = zmq.Context() self.socket = context.socket(zmq.REP) self.socket.bind("tcp://%s:%s" % (listen_on_ip, listen_on_port)) def get_command(self): """Attempt to return a unicode object from the command socket If no message is available without blocking (as opposed to a blank message), return None """ try: message_bytes = self.socket.recv(zmq.NOBLOCK) log.debug("Received message: %r", message_bytes) except zmq.ZMQError as exc: if exc.errno == zmq.EAGAIN: return None else: raise else: return message_bytes.decode(config.CODEC) def send_response(self, response): """Send a unicode object as reply to the most recently-issued command """ response_bytes = response.encode(config.CODEC) log.debug("About to send reponse: %r", response_bytes) self.socket.send(response_bytes) def parse_command(self, command): """Break a multi word command up into an action and its parameters """ words = shlex.split(command.lower()) return words[0], words[1:] def dispatch(self, command): """Pass a command along with its params to a suitable handler If the command is blank, succeed silently If the command has no handler, succeed silently If the handler raises an exception, fail with the exception message """ log.info("Dispatch on %s", command) if not command: return "OK" action, params = self.parse_command(command) log.debug("Action = %s, Params = %s", action, params) try: function = getattr(self, "do_" + action, None) if function: function(*params) return "OK" except KeyboardInterrupt: raise except Exception as exc: log.exception("Problem executing action %s", action) return "ERROR: %s" % exc def do_output(self, *args): """Pass a command directly to the current output processor """ if args: action, params = args[0], args[1:] log.debug("Pass %s directly to output with %s", action, params) function = getattr(self.output, "do_" + action, None) if function: function(*params) def do_finish(self): self.stop_event.set() # # Main loop # def start(self): while not self.stop_event.is_set(): try: command = self.get_command() if command is not None: response = self.dispatch(command.strip()) self.send_response(response) except KeyboardInterrupt: log.warn("Closing gracefully...") self.stop_event.set() break except: log.exception("Problem in main loop") self.stop_event.set() raise
tjguk/networkzero
misc/pyconuk2017/robotics/robot/robot.py
Robot.send_response
python
def send_response(self, response): response_bytes = response.encode(config.CODEC) log.debug("About to send reponse: %r", response_bytes) self.socket.send(response_bytes)
Send a unicode object as reply to the most recently-issued command
train
https://github.com/tjguk/networkzero/blob/0e3e81d2e9200b25a83ac07741612283599486d7/misc/pyconuk2017/robotics/robot/robot.py#L56-L61
null
class Robot(object): def __init__( self, output, stop_event=None, listen_on_ip=config.LISTEN_ON_IP, listen_on_port=config.LISTEN_ON_PORT ): log.info("Setting up Robot on %s:%s", listen_on_ip, listen_on_port) log.info("Outputting to %s", output) self.stop_event = stop_event or threading.Event() self._init_socket(listen_on_ip, listen_on_port) self.output = output self.output._init() def _init_socket(self, listen_on_ip, listen_on_port): context = zmq.Context() self.socket = context.socket(zmq.REP) self.socket.bind("tcp://%s:%s" % (listen_on_ip, listen_on_port)) def get_command(self): """Attempt to return a unicode object from the command socket If no message is available without blocking (as opposed to a blank message), return None """ try: message_bytes = self.socket.recv(zmq.NOBLOCK) log.debug("Received message: %r", message_bytes) except zmq.ZMQError as exc: if exc.errno == zmq.EAGAIN: return None else: raise else: return message_bytes.decode(config.CODEC) def parse_command(self, command): """Break a multi word command up into an action and its parameters """ words = shlex.split(command.lower()) return words[0], words[1:] def dispatch(self, command): """Pass a command along with its params to a suitable handler If the command is blank, succeed silently If the command has no handler, succeed silently If the handler raises an exception, fail with the exception message """ log.info("Dispatch on %s", command) if not command: return "OK" action, params = self.parse_command(command) log.debug("Action = %s, Params = %s", action, params) try: function = getattr(self, "do_" + action, None) if function: function(*params) return "OK" except KeyboardInterrupt: raise except Exception as exc: log.exception("Problem executing action %s", action) return "ERROR: %s" % exc def do_output(self, *args): """Pass a command directly to the current output processor """ if args: action, params = args[0], args[1:] log.debug("Pass %s directly to output with %s", action, params) function = getattr(self.output, "do_" + action, None) if function: function(*params) def do_finish(self): self.stop_event.set() # # Main loop # def start(self): while not self.stop_event.is_set(): try: command = self.get_command() if command is not None: response = self.dispatch(command.strip()) self.send_response(response) except KeyboardInterrupt: log.warn("Closing gracefully...") self.stop_event.set() break except: log.exception("Problem in main loop") self.stop_event.set() raise
tjguk/networkzero
misc/pyconuk2017/robotics/robot/robot.py
Robot.parse_command
python
def parse_command(self, command): words = shlex.split(command.lower()) return words[0], words[1:]
Break a multi word command up into an action and its parameters
train
https://github.com/tjguk/networkzero/blob/0e3e81d2e9200b25a83ac07741612283599486d7/misc/pyconuk2017/robotics/robot/robot.py#L63-L67
null
class Robot(object): def __init__( self, output, stop_event=None, listen_on_ip=config.LISTEN_ON_IP, listen_on_port=config.LISTEN_ON_PORT ): log.info("Setting up Robot on %s:%s", listen_on_ip, listen_on_port) log.info("Outputting to %s", output) self.stop_event = stop_event or threading.Event() self._init_socket(listen_on_ip, listen_on_port) self.output = output self.output._init() def _init_socket(self, listen_on_ip, listen_on_port): context = zmq.Context() self.socket = context.socket(zmq.REP) self.socket.bind("tcp://%s:%s" % (listen_on_ip, listen_on_port)) def get_command(self): """Attempt to return a unicode object from the command socket If no message is available without blocking (as opposed to a blank message), return None """ try: message_bytes = self.socket.recv(zmq.NOBLOCK) log.debug("Received message: %r", message_bytes) except zmq.ZMQError as exc: if exc.errno == zmq.EAGAIN: return None else: raise else: return message_bytes.decode(config.CODEC) def send_response(self, response): """Send a unicode object as reply to the most recently-issued command """ response_bytes = response.encode(config.CODEC) log.debug("About to send reponse: %r", response_bytes) self.socket.send(response_bytes) def dispatch(self, command): """Pass a command along with its params to a suitable handler If the command is blank, succeed silently If the command has no handler, succeed silently If the handler raises an exception, fail with the exception message """ log.info("Dispatch on %s", command) if not command: return "OK" action, params = self.parse_command(command) log.debug("Action = %s, Params = %s", action, params) try: function = getattr(self, "do_" + action, None) if function: function(*params) return "OK" except KeyboardInterrupt: raise except Exception as exc: log.exception("Problem executing action %s", action) return "ERROR: %s" % exc def do_output(self, *args): """Pass a command directly to the current output processor """ if args: action, params = args[0], args[1:] log.debug("Pass %s directly to output with %s", action, params) function = getattr(self.output, "do_" + action, None) if function: function(*params) def do_finish(self): self.stop_event.set() # # Main loop # def start(self): while not self.stop_event.is_set(): try: command = self.get_command() if command is not None: response = self.dispatch(command.strip()) self.send_response(response) except KeyboardInterrupt: log.warn("Closing gracefully...") self.stop_event.set() break except: log.exception("Problem in main loop") self.stop_event.set() raise
tjguk/networkzero
misc/pyconuk2017/robotics/robot/robot.py
Robot.dispatch
python
def dispatch(self, command): log.info("Dispatch on %s", command) if not command: return "OK" action, params = self.parse_command(command) log.debug("Action = %s, Params = %s", action, params) try: function = getattr(self, "do_" + action, None) if function: function(*params) return "OK" except KeyboardInterrupt: raise except Exception as exc: log.exception("Problem executing action %s", action) return "ERROR: %s" % exc
Pass a command along with its params to a suitable handler If the command is blank, succeed silently If the command has no handler, succeed silently If the handler raises an exception, fail with the exception message
train
https://github.com/tjguk/networkzero/blob/0e3e81d2e9200b25a83ac07741612283599486d7/misc/pyconuk2017/robotics/robot/robot.py#L69-L91
[ "def parse_command(self, command):\n \"\"\"Break a multi word command up into an action and its parameters\n \"\"\"\n words = shlex.split(command.lower())\n return words[0], words[1:]\n" ]
class Robot(object): def __init__( self, output, stop_event=None, listen_on_ip=config.LISTEN_ON_IP, listen_on_port=config.LISTEN_ON_PORT ): log.info("Setting up Robot on %s:%s", listen_on_ip, listen_on_port) log.info("Outputting to %s", output) self.stop_event = stop_event or threading.Event() self._init_socket(listen_on_ip, listen_on_port) self.output = output self.output._init() def _init_socket(self, listen_on_ip, listen_on_port): context = zmq.Context() self.socket = context.socket(zmq.REP) self.socket.bind("tcp://%s:%s" % (listen_on_ip, listen_on_port)) def get_command(self): """Attempt to return a unicode object from the command socket If no message is available without blocking (as opposed to a blank message), return None """ try: message_bytes = self.socket.recv(zmq.NOBLOCK) log.debug("Received message: %r", message_bytes) except zmq.ZMQError as exc: if exc.errno == zmq.EAGAIN: return None else: raise else: return message_bytes.decode(config.CODEC) def send_response(self, response): """Send a unicode object as reply to the most recently-issued command """ response_bytes = response.encode(config.CODEC) log.debug("About to send reponse: %r", response_bytes) self.socket.send(response_bytes) def parse_command(self, command): """Break a multi word command up into an action and its parameters """ words = shlex.split(command.lower()) return words[0], words[1:] def dispatch(self, command): """Pass a command along with its params to a suitable handler If the command is blank, succeed silently If the command has no handler, succeed silently If the handler raises an exception, fail with the exception message """ log.info("Dispatch on %s", command) if not command: return "OK" action, params = self.parse_command(command) log.debug("Action = %s, Params = %s", action, params) try: function = getattr(self, "do_" + action, None) if function: function(*params) return "OK" except KeyboardInterrupt: raise except Exception as exc: log.exception("Problem executing action %s", action) return "ERROR: %s" % exc def do_output(self, *args): """Pass a command directly to the current output processor """ if args: action, params = args[0], args[1:] log.debug("Pass %s directly to output with %s", action, params) function = getattr(self.output, "do_" + action, None) if function: function(*params) def do_finish(self): self.stop_event.set() # # Main loop # def start(self): while not self.stop_event.is_set(): try: command = self.get_command() if command is not None: response = self.dispatch(command.strip()) self.send_response(response) except KeyboardInterrupt: log.warn("Closing gracefully...") self.stop_event.set() break except: log.exception("Problem in main loop") self.stop_event.set() raise
tjguk/networkzero
misc/pyconuk2017/robotics/robot/robot.py
Robot.do_output
python
def do_output(self, *args): if args: action, params = args[0], args[1:] log.debug("Pass %s directly to output with %s", action, params) function = getattr(self.output, "do_" + action, None) if function: function(*params)
Pass a command directly to the current output processor
train
https://github.com/tjguk/networkzero/blob/0e3e81d2e9200b25a83ac07741612283599486d7/misc/pyconuk2017/robotics/robot/robot.py#L93-L101
null
class Robot(object): def __init__( self, output, stop_event=None, listen_on_ip=config.LISTEN_ON_IP, listen_on_port=config.LISTEN_ON_PORT ): log.info("Setting up Robot on %s:%s", listen_on_ip, listen_on_port) log.info("Outputting to %s", output) self.stop_event = stop_event or threading.Event() self._init_socket(listen_on_ip, listen_on_port) self.output = output self.output._init() def _init_socket(self, listen_on_ip, listen_on_port): context = zmq.Context() self.socket = context.socket(zmq.REP) self.socket.bind("tcp://%s:%s" % (listen_on_ip, listen_on_port)) def get_command(self): """Attempt to return a unicode object from the command socket If no message is available without blocking (as opposed to a blank message), return None """ try: message_bytes = self.socket.recv(zmq.NOBLOCK) log.debug("Received message: %r", message_bytes) except zmq.ZMQError as exc: if exc.errno == zmq.EAGAIN: return None else: raise else: return message_bytes.decode(config.CODEC) def send_response(self, response): """Send a unicode object as reply to the most recently-issued command """ response_bytes = response.encode(config.CODEC) log.debug("About to send reponse: %r", response_bytes) self.socket.send(response_bytes) def parse_command(self, command): """Break a multi word command up into an action and its parameters """ words = shlex.split(command.lower()) return words[0], words[1:] def dispatch(self, command): """Pass a command along with its params to a suitable handler If the command is blank, succeed silently If the command has no handler, succeed silently If the handler raises an exception, fail with the exception message """ log.info("Dispatch on %s", command) if not command: return "OK" action, params = self.parse_command(command) log.debug("Action = %s, Params = %s", action, params) try: function = getattr(self, "do_" + action, None) if function: function(*params) return "OK" except KeyboardInterrupt: raise except Exception as exc: log.exception("Problem executing action %s", action) return "ERROR: %s" % exc def do_finish(self): self.stop_event.set() # # Main loop # def start(self): while not self.stop_event.is_set(): try: command = self.get_command() if command is not None: response = self.dispatch(command.strip()) self.send_response(response) except KeyboardInterrupt: log.warn("Closing gracefully...") self.stop_event.set() break except: log.exception("Problem in main loop") self.stop_event.set() raise
tjguk/networkzero
networkzero/sockets.py
Sockets.get_socket
python
def get_socket(self, address, role): # # If this thread doesn't yet have a sockets cache # in its local storage, create one here. # try: self._tls.sockets except AttributeError: self._tls.sockets = {} # Convert the address to a single canonical string. # # If a list of addresses is passed, turn it into a tuple # of canonical addresses for use as a dictionary key. # Otherwise convert it to a single canonical string. # if isinstance(address, list): caddress = tuple(core.address(a) for a in address) else: caddress = core.address(address) # # Each socket is identified for this thread by its address(es) # and the role the socket is playing (listener, publisher, etc.) # That is, within one thread, we are cacheing a read or a write # socket to the same address(es). # # The slight corner case from this is that if you attempt to # send to [addressA, addressB] and then to addressA and then # to [addressB, addressA], three separate sockets will be # created and used. # identifier = caddress if identifier not in self._tls.sockets: _logger.debug("%s does not exist in local sockets", identifier) # # If this is a listening / subscribing socket, it can only # be bound once, regardless of thread. Therefore keep a # threads-global list of addresses used and make sure this # one hasn't been used elsewhere. # if role in Socket.binding_roles: with self._lock: if identifier in self._sockets: raise core.SocketAlreadyExistsError("You cannot create a listening socket in more than one thread") else: self._sockets.add(identifier) type = self.roles[role] socket = context.socket(type) socket.role = role socket.address = caddress # # Do this last so that an exception earlier will result # in the socket not being cached # self._tls.sockets[identifier] = socket else: _logger.debug("%s already not exist in local sockets", identifier) # # Only return sockets created in this thread # socket = self._tls.sockets[identifier] return socket
Create or retrieve a socket of the right type, already connected to the address. Address (ip:port) must be fully specified at this point. core.address can be used to generate an address.
train
https://github.com/tjguk/networkzero/blob/0e3e81d2e9200b25a83ac07741612283599486d7/networkzero/sockets.py#L119-L188
[ "def address(address=None):\n \"\"\"Convert one of a number of inputs into a valid ip:port string.\n\n Elements which are not provided are filled in as follows:\n\n * IP Address: the system is asked for the set of IP addresses associated\n with the machine and the first one is used, preferring...
class Sockets: try_length_ms = 500 # wait for .5 second at a time roles = { "listener" : zmq.REP, "speaker" : zmq.REQ, "publisher" : zmq.PUB, "subscriber" : zmq.SUB } def __init__(self): self._tls = threading.local() self._lock = threading.Lock() with self._lock: self._sockets = set() def get_socket(self, address, role): """Create or retrieve a socket of the right type, already connected to the address. Address (ip:port) must be fully specified at this point. core.address can be used to generate an address. """ # # If this thread doesn't yet have a sockets cache # in its local storage, create one here. # try: self._tls.sockets except AttributeError: self._tls.sockets = {} # Convert the address to a single canonical string. # # If a list of addresses is passed, turn it into a tuple # of canonical addresses for use as a dictionary key. # Otherwise convert it to a single canonical string. # if isinstance(address, list): caddress = tuple(core.address(a) for a in address) else: caddress = core.address(address) # # Each socket is identified for this thread by its address(es) # and the role the socket is playing (listener, publisher, etc.) # That is, within one thread, we are cacheing a read or a write # socket to the same address(es). # # The slight corner case from this is that if you attempt to # send to [addressA, addressB] and then to addressA and then # to [addressB, addressA], three separate sockets will be # created and used. # identifier = caddress if identifier not in self._tls.sockets: _logger.debug("%s does not exist in local sockets", identifier) # # If this is a listening / subscribing socket, it can only # be bound once, regardless of thread. Therefore keep a # threads-global list of addresses used and make sure this # one hasn't been used elsewhere. # if role in Socket.binding_roles: with self._lock: if identifier in self._sockets: raise core.SocketAlreadyExistsError("You cannot create a listening socket in more than one thread") else: self._sockets.add(identifier) type = self.roles[role] socket = context.socket(type) socket.role = role socket.address = caddress # # Do this last so that an exception earlier will result # in the socket not being cached # self._tls.sockets[identifier] = socket else: _logger.debug("%s already not exist in local sockets", identifier) # # Only return sockets created in this thread # socket = self._tls.sockets[identifier] return socket def intervals_ms(self, timeout_ms): """Generate a series of interval lengths, in ms, which will add up to the number of ms in timeout_ms. If timeout_ms is None, keep returning intervals forever. """ if timeout_ms is config.FOREVER: while True: yield self.try_length_ms else: whole_intervals, part_interval = divmod(timeout_ms, self.try_length_ms) for _ in range(whole_intervals): yield self.try_length_ms yield part_interval def _receive_with_timeout(self, socket, timeout_s, use_multipart=False): """Check for socket activity and either return what's received on the socket or time out if timeout_s expires without anything on the socket. This is implemented in loops of self.try_length_ms milliseconds to allow Ctrl-C handling to take place. """ if timeout_s is config.FOREVER: timeout_ms = config.FOREVER else: timeout_ms = int(1000 * timeout_s) poller = zmq.Poller() poller.register(socket, zmq.POLLIN) ms_so_far = 0 try: for interval_ms in self.intervals_ms(timeout_ms): sockets = dict(poller.poll(interval_ms)) ms_so_far += interval_ms if socket in sockets: if use_multipart: return socket.recv_multipart() else: return socket.recv() else: raise core.SocketTimedOutError(timeout_s) except KeyboardInterrupt: raise core.SocketInterruptedError(ms_so_far / 1000.0) def wait_for_message_from(self, address, wait_for_s): socket = self.get_socket(address, "listener") try: message = self._receive_with_timeout(socket, wait_for_s) except (core.SocketTimedOutError): return None else: return _unserialise(message) def send_message_to(self, address, message, wait_for_reply_s): socket = self.get_socket(address, "speaker") serialised_message = _serialise(message) socket.send(serialised_message) return _unserialise(self._receive_with_timeout(socket, wait_for_reply_s)) def send_reply_to(self, address, reply): socket = self.get_socket(address, "listener") reply = _serialise(reply) return socket.send(reply) def send_news_to(self, address, topic, data): socket = self.get_socket(address, "publisher") return socket.send_multipart(_serialise_for_pubsub(topic, data)) def wait_for_news_from(self, address, topic, wait_for_s, is_raw=False): if isinstance(address, list): addresses = address else: addresses = [address] socket = self.get_socket(addresses, "subscriber") if isinstance(topic, str): topics = [topic] else: topics = topic for t in topics: socket.set(zmq.SUBSCRIBE, t.encode(config.ENCODING)) try: result = self._receive_with_timeout(socket, wait_for_s, use_multipart=True) unserialised_result = _unserialise_for_pubsub(result, is_raw) return unserialised_result except (core.SocketTimedOutError, core.SocketInterruptedError): return None, None
tjguk/networkzero
networkzero/sockets.py
Sockets.intervals_ms
python
def intervals_ms(self, timeout_ms): if timeout_ms is config.FOREVER: while True: yield self.try_length_ms else: whole_intervals, part_interval = divmod(timeout_ms, self.try_length_ms) for _ in range(whole_intervals): yield self.try_length_ms yield part_interval
Generate a series of interval lengths, in ms, which will add up to the number of ms in timeout_ms. If timeout_ms is None, keep returning intervals forever.
train
https://github.com/tjguk/networkzero/blob/0e3e81d2e9200b25a83ac07741612283599486d7/networkzero/sockets.py#L190-L202
null
class Sockets: try_length_ms = 500 # wait for .5 second at a time roles = { "listener" : zmq.REP, "speaker" : zmq.REQ, "publisher" : zmq.PUB, "subscriber" : zmq.SUB } def __init__(self): self._tls = threading.local() self._lock = threading.Lock() with self._lock: self._sockets = set() def get_socket(self, address, role): """Create or retrieve a socket of the right type, already connected to the address. Address (ip:port) must be fully specified at this point. core.address can be used to generate an address. """ # # If this thread doesn't yet have a sockets cache # in its local storage, create one here. # try: self._tls.sockets except AttributeError: self._tls.sockets = {} # Convert the address to a single canonical string. # # If a list of addresses is passed, turn it into a tuple # of canonical addresses for use as a dictionary key. # Otherwise convert it to a single canonical string. # if isinstance(address, list): caddress = tuple(core.address(a) for a in address) else: caddress = core.address(address) # # Each socket is identified for this thread by its address(es) # and the role the socket is playing (listener, publisher, etc.) # That is, within one thread, we are cacheing a read or a write # socket to the same address(es). # # The slight corner case from this is that if you attempt to # send to [addressA, addressB] and then to addressA and then # to [addressB, addressA], three separate sockets will be # created and used. # identifier = caddress if identifier not in self._tls.sockets: _logger.debug("%s does not exist in local sockets", identifier) # # If this is a listening / subscribing socket, it can only # be bound once, regardless of thread. Therefore keep a # threads-global list of addresses used and make sure this # one hasn't been used elsewhere. # if role in Socket.binding_roles: with self._lock: if identifier in self._sockets: raise core.SocketAlreadyExistsError("You cannot create a listening socket in more than one thread") else: self._sockets.add(identifier) type = self.roles[role] socket = context.socket(type) socket.role = role socket.address = caddress # # Do this last so that an exception earlier will result # in the socket not being cached # self._tls.sockets[identifier] = socket else: _logger.debug("%s already not exist in local sockets", identifier) # # Only return sockets created in this thread # socket = self._tls.sockets[identifier] return socket def _receive_with_timeout(self, socket, timeout_s, use_multipart=False): """Check for socket activity and either return what's received on the socket or time out if timeout_s expires without anything on the socket. This is implemented in loops of self.try_length_ms milliseconds to allow Ctrl-C handling to take place. """ if timeout_s is config.FOREVER: timeout_ms = config.FOREVER else: timeout_ms = int(1000 * timeout_s) poller = zmq.Poller() poller.register(socket, zmq.POLLIN) ms_so_far = 0 try: for interval_ms in self.intervals_ms(timeout_ms): sockets = dict(poller.poll(interval_ms)) ms_so_far += interval_ms if socket in sockets: if use_multipart: return socket.recv_multipart() else: return socket.recv() else: raise core.SocketTimedOutError(timeout_s) except KeyboardInterrupt: raise core.SocketInterruptedError(ms_so_far / 1000.0) def wait_for_message_from(self, address, wait_for_s): socket = self.get_socket(address, "listener") try: message = self._receive_with_timeout(socket, wait_for_s) except (core.SocketTimedOutError): return None else: return _unserialise(message) def send_message_to(self, address, message, wait_for_reply_s): socket = self.get_socket(address, "speaker") serialised_message = _serialise(message) socket.send(serialised_message) return _unserialise(self._receive_with_timeout(socket, wait_for_reply_s)) def send_reply_to(self, address, reply): socket = self.get_socket(address, "listener") reply = _serialise(reply) return socket.send(reply) def send_news_to(self, address, topic, data): socket = self.get_socket(address, "publisher") return socket.send_multipart(_serialise_for_pubsub(topic, data)) def wait_for_news_from(self, address, topic, wait_for_s, is_raw=False): if isinstance(address, list): addresses = address else: addresses = [address] socket = self.get_socket(addresses, "subscriber") if isinstance(topic, str): topics = [topic] else: topics = topic for t in topics: socket.set(zmq.SUBSCRIBE, t.encode(config.ENCODING)) try: result = self._receive_with_timeout(socket, wait_for_s, use_multipart=True) unserialised_result = _unserialise_for_pubsub(result, is_raw) return unserialised_result except (core.SocketTimedOutError, core.SocketInterruptedError): return None, None
tjguk/networkzero
networkzero/sockets.py
Sockets._receive_with_timeout
python
def _receive_with_timeout(self, socket, timeout_s, use_multipart=False): if timeout_s is config.FOREVER: timeout_ms = config.FOREVER else: timeout_ms = int(1000 * timeout_s) poller = zmq.Poller() poller.register(socket, zmq.POLLIN) ms_so_far = 0 try: for interval_ms in self.intervals_ms(timeout_ms): sockets = dict(poller.poll(interval_ms)) ms_so_far += interval_ms if socket in sockets: if use_multipart: return socket.recv_multipart() else: return socket.recv() else: raise core.SocketTimedOutError(timeout_s) except KeyboardInterrupt: raise core.SocketInterruptedError(ms_so_far / 1000.0)
Check for socket activity and either return what's received on the socket or time out if timeout_s expires without anything on the socket. This is implemented in loops of self.try_length_ms milliseconds to allow Ctrl-C handling to take place.
train
https://github.com/tjguk/networkzero/blob/0e3e81d2e9200b25a83ac07741612283599486d7/networkzero/sockets.py#L204-L232
[ "def intervals_ms(self, timeout_ms):\n \"\"\"Generate a series of interval lengths, in ms, which\n will add up to the number of ms in timeout_ms. If timeout_ms\n is None, keep returning intervals forever.\n \"\"\"\n if timeout_ms is config.FOREVER:\n while True:\n yield self.try_len...
class Sockets: try_length_ms = 500 # wait for .5 second at a time roles = { "listener" : zmq.REP, "speaker" : zmq.REQ, "publisher" : zmq.PUB, "subscriber" : zmq.SUB } def __init__(self): self._tls = threading.local() self._lock = threading.Lock() with self._lock: self._sockets = set() def get_socket(self, address, role): """Create or retrieve a socket of the right type, already connected to the address. Address (ip:port) must be fully specified at this point. core.address can be used to generate an address. """ # # If this thread doesn't yet have a sockets cache # in its local storage, create one here. # try: self._tls.sockets except AttributeError: self._tls.sockets = {} # Convert the address to a single canonical string. # # If a list of addresses is passed, turn it into a tuple # of canonical addresses for use as a dictionary key. # Otherwise convert it to a single canonical string. # if isinstance(address, list): caddress = tuple(core.address(a) for a in address) else: caddress = core.address(address) # # Each socket is identified for this thread by its address(es) # and the role the socket is playing (listener, publisher, etc.) # That is, within one thread, we are cacheing a read or a write # socket to the same address(es). # # The slight corner case from this is that if you attempt to # send to [addressA, addressB] and then to addressA and then # to [addressB, addressA], three separate sockets will be # created and used. # identifier = caddress if identifier not in self._tls.sockets: _logger.debug("%s does not exist in local sockets", identifier) # # If this is a listening / subscribing socket, it can only # be bound once, regardless of thread. Therefore keep a # threads-global list of addresses used and make sure this # one hasn't been used elsewhere. # if role in Socket.binding_roles: with self._lock: if identifier in self._sockets: raise core.SocketAlreadyExistsError("You cannot create a listening socket in more than one thread") else: self._sockets.add(identifier) type = self.roles[role] socket = context.socket(type) socket.role = role socket.address = caddress # # Do this last so that an exception earlier will result # in the socket not being cached # self._tls.sockets[identifier] = socket else: _logger.debug("%s already not exist in local sockets", identifier) # # Only return sockets created in this thread # socket = self._tls.sockets[identifier] return socket def intervals_ms(self, timeout_ms): """Generate a series of interval lengths, in ms, which will add up to the number of ms in timeout_ms. If timeout_ms is None, keep returning intervals forever. """ if timeout_ms is config.FOREVER: while True: yield self.try_length_ms else: whole_intervals, part_interval = divmod(timeout_ms, self.try_length_ms) for _ in range(whole_intervals): yield self.try_length_ms yield part_interval def _receive_with_timeout(self, socket, timeout_s, use_multipart=False): """Check for socket activity and either return what's received on the socket or time out if timeout_s expires without anything on the socket. This is implemented in loops of self.try_length_ms milliseconds to allow Ctrl-C handling to take place. """ if timeout_s is config.FOREVER: timeout_ms = config.FOREVER else: timeout_ms = int(1000 * timeout_s) poller = zmq.Poller() poller.register(socket, zmq.POLLIN) ms_so_far = 0 try: for interval_ms in self.intervals_ms(timeout_ms): sockets = dict(poller.poll(interval_ms)) ms_so_far += interval_ms if socket in sockets: if use_multipart: return socket.recv_multipart() else: return socket.recv() else: raise core.SocketTimedOutError(timeout_s) except KeyboardInterrupt: raise core.SocketInterruptedError(ms_so_far / 1000.0) def wait_for_message_from(self, address, wait_for_s): socket = self.get_socket(address, "listener") try: message = self._receive_with_timeout(socket, wait_for_s) except (core.SocketTimedOutError): return None else: return _unserialise(message) def send_message_to(self, address, message, wait_for_reply_s): socket = self.get_socket(address, "speaker") serialised_message = _serialise(message) socket.send(serialised_message) return _unserialise(self._receive_with_timeout(socket, wait_for_reply_s)) def send_reply_to(self, address, reply): socket = self.get_socket(address, "listener") reply = _serialise(reply) return socket.send(reply) def send_news_to(self, address, topic, data): socket = self.get_socket(address, "publisher") return socket.send_multipart(_serialise_for_pubsub(topic, data)) def wait_for_news_from(self, address, topic, wait_for_s, is_raw=False): if isinstance(address, list): addresses = address else: addresses = [address] socket = self.get_socket(addresses, "subscriber") if isinstance(topic, str): topics = [topic] else: topics = topic for t in topics: socket.set(zmq.SUBSCRIBE, t.encode(config.ENCODING)) try: result = self._receive_with_timeout(socket, wait_for_s, use_multipart=True) unserialised_result = _unserialise_for_pubsub(result, is_raw) return unserialised_result except (core.SocketTimedOutError, core.SocketInterruptedError): return None, None
tjguk/networkzero
networkzero/core.py
is_valid_ip_pattern
python
def is_valid_ip_pattern(ip): ip = ip.replace('*', '1') try: socket.inet_aton(ip) return True except socket.error: # Not a valid IPv4 address pattern return False
Check whether a string matches the outline of an IPv4 address, allowing "*" as a wildcard
train
https://github.com/tjguk/networkzero/blob/0e3e81d2e9200b25a83ac07741612283599486d7/networkzero/core.py#L101-L110
null
# -*- coding: utf-8 -*- import base64 import fnmatch import logging import random import shlex import socket try: import netifaces except ImportError: warnings.warn("Unable to import netifaces; using local fallback") from . import _netifaces as netifaces from . import config def get_logger(name): # # For now, this is just a hand-off to logging.getLogger # Later, though, we might want to add a null handler etc. # logger = logging.getLogger(name) logger.setLevel(logging.DEBUG) return logger _debug_logging_enabled = False def _enable_debug_logging(): global _debug_logging_enabled if not _debug_logging_enabled: logger = logging.getLogger("networkzero") handler = logging.FileHandler("network.log", "w", encoding="utf-8") handler.setFormatter(logging.Formatter("%(asctime)s %(threadName)s %(name)s %(levelname)s %(message)s")) handler.setLevel(logging.DEBUG) logger.addHandler(handler) _debug_logging_enabled = True _logger = get_logger(__name__) # # Common exceptions # class NetworkZeroError(Exception): pass class SocketAlreadyExistsError(NetworkZeroError): pass class SocketTimedOutError(NetworkZeroError): def __init__(self, n_seconds): self.n_seconds = n_seconds def __str__(self): return "Gave up waiting after %s seconds; this connection is now unusable" % self.n_seconds class SocketInterruptedError(NetworkZeroError): def __init__(self, after_n_seconds): self.after_n_seconds = after_n_seconds def __str__(self): return "Interrupted after %s seconds; this connection is now unusable" % self.after_n_seconds class AddressError(NetworkZeroError): pass class NoAddressFoundError(AddressError): pass class InvalidAddressError(NetworkZeroError): def __init__(self, address, errno=None): self.address = address self.errno = errno def __str__(self): message = "%s is not a valid address" % self.address if self.errno: message += "; the system returned an error of %d" % self.errno return message class DifferentThreadError(NetworkZeroError): pass # # Ports in the range 0xc000..0xffff are reserved # for dynamic allocation # PORT_POOL = list(config.DYNAMIC_PORTS) def split_address(address): if ":" in address: ip, _, port = address.partition(":") else: if address.isdigit(): ip, port = "", address else: ip, port = address, "" return ip, port def is_valid_port(port, port_range=range(65536)): try: return int(port) in port_range except ValueError: return False def is_valid_address(address, port_range=range(65536)): ip, port = split_address(address) return is_valid_ip_pattern(ip) and is_valid_port(port, port_range) def _find_ip4_broadcast_addresses(): """Yield each IP4 broadcast address, and the all-broadcast """ yield "255.255.255.255" for interface in netifaces.interfaces(): ifaddresses = netifaces.ifaddresses(interface) for family in ifaddresses: if family == netifaces.AF_INET: address_info = ifaddresses[family] for info in address_info: if "broadcast" in info: yield info['broadcast'] _ip4_addresses = None def _find_ip4_addresses(): """Find all the IP4 addresses currently bound to interfaces """ global _ip4_addresses proto = socket.AF_INET if _ip4_addresses is None: _ip4_addresses = [] # # Determine the interface for the default gateway # (if any) and, later, prioritise the INET address on # that interface. # default_gateway = netifaces.gateways()['default'] if proto in default_gateway: _, default_gateway_interface = default_gateway[proto] else: default_gateway_interface = None for interface in netifaces.interfaces(): for info in netifaces.ifaddresses(interface).get(netifaces.AF_INET, []): if info['addr']: if interface == default_gateway_interface: _ip4_addresses.insert(0, info['addr']) else: _ip4_addresses.append(info['addr']) return _ip4_addresses _ip4 = None _prefer = None def _find_ip4(prefer=None): global _ip4, _prefer # # Order the list of possible addresses on the machine: if any # address pattern is given as a preference (most -> least) # give it that weighting, otherwise treat all addresses # numerically. If no preference is given, prefer the most # likely useful local address range. # if prefer: _prefer = prefer else: _prefer = ["192.168.*"] def sorter(ip4): octets = [int(i) for i in ip4.split(".")] for n, pattern in enumerate(_prefer): if fnmatch.fnmatch(ip4, pattern): return n, octets else: # # Return the address itself if it doesn't match # a preference # return n + 1, octets ip4_addresses = _find_ip4_addresses() # # Pick an address allowing for user preference if stated # if not ip4_addresses: raise NoAddressFoundError else: # # Find the best match. If the user actually supplied a preference # list, assume an exact match is required to at least one of the # patterns. # ip4 = min(ip4_addresses, key=sorter) if prefer and not any(fnmatch.fnmatch(ip4, pattern) for pattern in prefer): raise NoAddressFoundError("No address matches any of: %s" % ", ".join(prefer)) else: _ip4 = ip4 return _ip4 def address(address=None): """Convert one of a number of inputs into a valid ip:port string. Elements which are not provided are filled in as follows: * IP Address: the system is asked for the set of IP addresses associated with the machine and the first one is used, preferring those matching `address` if it is a wildcard. * Port number: a random port is selected from the pool of dynamically-available port numbers. This means you can pass any of: nothing; a hostname; an IP address; an IP address with wildcards; a port number If an IP address is supplied but is invalid, an InvalidAddressError exception is raised. :param address: (optional) Any of: an IP address, a port number, or both :returns: a valid ip:port string for this machine """ address = str(address or "").strip() # # If the address is an ip:port pair, split into its component parts. # Otherwise, try to determine whether we're looking at an IP # or at a port and leave the other one blank # host_or_ip, port = split_address(address) # # If the port has been supplied, make sure it's numeric and that it's a valid # port number. If it hasn't been supplied, remove a random one from the pool # of possible dynamically-allocated ports and use that. # if port: try: port = int(port) except ValueError: raise AddressError("Port %s must be a number" % port) if port not in config.VALID_PORTS: raise AddressError("Port %d must be in range %d - %d" % ( port, min(config.VALID_PORTS), max(config.VALID_PORTS)) ) else: random.shuffle(PORT_POOL) port = PORT_POOL.pop() # # The address part could be an IP address (optionally including # wildcards to indicate a preference) or a hostname or nothing. # If it's a hostname we attempt to resolve it to an IP address. # It it's nothing or a wildcard we query the system for a matching IP address. # if (not host_or_ip) or is_valid_ip_pattern(host_or_ip): # # If a specific IP address is given, use that. # If an IP pattern is given (ie something with a wildcard in it) treat # that as no address with a preference for that wildcard. # prefer = None if "*" in host_or_ip: host_or_ip, prefer = None, [host_or_ip] # # If no IP (or only a wildcard) is specified, query the system for valid # addresses, preferring those which match the wildcard. NB if the preference # matches one we've previously used, we can return a cached address. But # different requests can specify different wildcard preferences. # if not host_or_ip: if _ip4 and _prefer == prefer: ip = _ip4 else: ip = _find_ip4(prefer) else: ip = host_or_ip else: # # Treat the string as a hostname and resolve to an IP4 address # try: ip = socket.gethostbyname(host_or_ip) except socket.gaierror as exc: _logger.error("gaierror %d for %s", exc.errno, host_or_ip) raise InvalidAddressError(host_or_ip, exc.errno) else: # # Bizarrely specific check because BT Internet "helpfully" # redirects DNS fails to this address which hosts a sponsored # landing page! # if ip == "92.242.132.15": raise InvalidAddressError(host_or_ip, 0) return "%s:%s" % (ip, port) def action_and_params(commandline): """Treat a command line as an action followed by parameter :param commandline: a string containing at least an action :returns: action, [param1, param2, ...] """ components = shlex.split(commandline) return components[0], components[1:] def bytes_to_string(data): """Take bytes and return a base64-encoded unicode string equivalent :param data: a bytes object :returns: base64-encoded unicode object """ return base64.b64encode(data).decode("ascii") def string_to_bytes(data): """Take a base64-encoded unicode string and return the equivalent bytes :param data: a base64-encoded unicode object :returns: the equivalent bytes """ return base64.b64decode(data.encode("ascii"))
tjguk/networkzero
networkzero/core.py
_find_ip4_broadcast_addresses
python
def _find_ip4_broadcast_addresses(): yield "255.255.255.255" for interface in netifaces.interfaces(): ifaddresses = netifaces.ifaddresses(interface) for family in ifaddresses: if family == netifaces.AF_INET: address_info = ifaddresses[family] for info in address_info: if "broadcast" in info: yield info['broadcast']
Yield each IP4 broadcast address, and the all-broadcast
train
https://github.com/tjguk/networkzero/blob/0e3e81d2e9200b25a83ac07741612283599486d7/networkzero/core.py#L122-L133
null
# -*- coding: utf-8 -*- import base64 import fnmatch import logging import random import shlex import socket try: import netifaces except ImportError: warnings.warn("Unable to import netifaces; using local fallback") from . import _netifaces as netifaces from . import config def get_logger(name): # # For now, this is just a hand-off to logging.getLogger # Later, though, we might want to add a null handler etc. # logger = logging.getLogger(name) logger.setLevel(logging.DEBUG) return logger _debug_logging_enabled = False def _enable_debug_logging(): global _debug_logging_enabled if not _debug_logging_enabled: logger = logging.getLogger("networkzero") handler = logging.FileHandler("network.log", "w", encoding="utf-8") handler.setFormatter(logging.Formatter("%(asctime)s %(threadName)s %(name)s %(levelname)s %(message)s")) handler.setLevel(logging.DEBUG) logger.addHandler(handler) _debug_logging_enabled = True _logger = get_logger(__name__) # # Common exceptions # class NetworkZeroError(Exception): pass class SocketAlreadyExistsError(NetworkZeroError): pass class SocketTimedOutError(NetworkZeroError): def __init__(self, n_seconds): self.n_seconds = n_seconds def __str__(self): return "Gave up waiting after %s seconds; this connection is now unusable" % self.n_seconds class SocketInterruptedError(NetworkZeroError): def __init__(self, after_n_seconds): self.after_n_seconds = after_n_seconds def __str__(self): return "Interrupted after %s seconds; this connection is now unusable" % self.after_n_seconds class AddressError(NetworkZeroError): pass class NoAddressFoundError(AddressError): pass class InvalidAddressError(NetworkZeroError): def __init__(self, address, errno=None): self.address = address self.errno = errno def __str__(self): message = "%s is not a valid address" % self.address if self.errno: message += "; the system returned an error of %d" % self.errno return message class DifferentThreadError(NetworkZeroError): pass # # Ports in the range 0xc000..0xffff are reserved # for dynamic allocation # PORT_POOL = list(config.DYNAMIC_PORTS) def split_address(address): if ":" in address: ip, _, port = address.partition(":") else: if address.isdigit(): ip, port = "", address else: ip, port = address, "" return ip, port def is_valid_ip_pattern(ip): """Check whether a string matches the outline of an IPv4 address, allowing "*" as a wildcard""" ip = ip.replace('*', '1') try: socket.inet_aton(ip) return True except socket.error: # Not a valid IPv4 address pattern return False def is_valid_port(port, port_range=range(65536)): try: return int(port) in port_range except ValueError: return False def is_valid_address(address, port_range=range(65536)): ip, port = split_address(address) return is_valid_ip_pattern(ip) and is_valid_port(port, port_range) _ip4_addresses = None def _find_ip4_addresses(): """Find all the IP4 addresses currently bound to interfaces """ global _ip4_addresses proto = socket.AF_INET if _ip4_addresses is None: _ip4_addresses = [] # # Determine the interface for the default gateway # (if any) and, later, prioritise the INET address on # that interface. # default_gateway = netifaces.gateways()['default'] if proto in default_gateway: _, default_gateway_interface = default_gateway[proto] else: default_gateway_interface = None for interface in netifaces.interfaces(): for info in netifaces.ifaddresses(interface).get(netifaces.AF_INET, []): if info['addr']: if interface == default_gateway_interface: _ip4_addresses.insert(0, info['addr']) else: _ip4_addresses.append(info['addr']) return _ip4_addresses _ip4 = None _prefer = None def _find_ip4(prefer=None): global _ip4, _prefer # # Order the list of possible addresses on the machine: if any # address pattern is given as a preference (most -> least) # give it that weighting, otherwise treat all addresses # numerically. If no preference is given, prefer the most # likely useful local address range. # if prefer: _prefer = prefer else: _prefer = ["192.168.*"] def sorter(ip4): octets = [int(i) for i in ip4.split(".")] for n, pattern in enumerate(_prefer): if fnmatch.fnmatch(ip4, pattern): return n, octets else: # # Return the address itself if it doesn't match # a preference # return n + 1, octets ip4_addresses = _find_ip4_addresses() # # Pick an address allowing for user preference if stated # if not ip4_addresses: raise NoAddressFoundError else: # # Find the best match. If the user actually supplied a preference # list, assume an exact match is required to at least one of the # patterns. # ip4 = min(ip4_addresses, key=sorter) if prefer and not any(fnmatch.fnmatch(ip4, pattern) for pattern in prefer): raise NoAddressFoundError("No address matches any of: %s" % ", ".join(prefer)) else: _ip4 = ip4 return _ip4 def address(address=None): """Convert one of a number of inputs into a valid ip:port string. Elements which are not provided are filled in as follows: * IP Address: the system is asked for the set of IP addresses associated with the machine and the first one is used, preferring those matching `address` if it is a wildcard. * Port number: a random port is selected from the pool of dynamically-available port numbers. This means you can pass any of: nothing; a hostname; an IP address; an IP address with wildcards; a port number If an IP address is supplied but is invalid, an InvalidAddressError exception is raised. :param address: (optional) Any of: an IP address, a port number, or both :returns: a valid ip:port string for this machine """ address = str(address or "").strip() # # If the address is an ip:port pair, split into its component parts. # Otherwise, try to determine whether we're looking at an IP # or at a port and leave the other one blank # host_or_ip, port = split_address(address) # # If the port has been supplied, make sure it's numeric and that it's a valid # port number. If it hasn't been supplied, remove a random one from the pool # of possible dynamically-allocated ports and use that. # if port: try: port = int(port) except ValueError: raise AddressError("Port %s must be a number" % port) if port not in config.VALID_PORTS: raise AddressError("Port %d must be in range %d - %d" % ( port, min(config.VALID_PORTS), max(config.VALID_PORTS)) ) else: random.shuffle(PORT_POOL) port = PORT_POOL.pop() # # The address part could be an IP address (optionally including # wildcards to indicate a preference) or a hostname or nothing. # If it's a hostname we attempt to resolve it to an IP address. # It it's nothing or a wildcard we query the system for a matching IP address. # if (not host_or_ip) or is_valid_ip_pattern(host_or_ip): # # If a specific IP address is given, use that. # If an IP pattern is given (ie something with a wildcard in it) treat # that as no address with a preference for that wildcard. # prefer = None if "*" in host_or_ip: host_or_ip, prefer = None, [host_or_ip] # # If no IP (or only a wildcard) is specified, query the system for valid # addresses, preferring those which match the wildcard. NB if the preference # matches one we've previously used, we can return a cached address. But # different requests can specify different wildcard preferences. # if not host_or_ip: if _ip4 and _prefer == prefer: ip = _ip4 else: ip = _find_ip4(prefer) else: ip = host_or_ip else: # # Treat the string as a hostname and resolve to an IP4 address # try: ip = socket.gethostbyname(host_or_ip) except socket.gaierror as exc: _logger.error("gaierror %d for %s", exc.errno, host_or_ip) raise InvalidAddressError(host_or_ip, exc.errno) else: # # Bizarrely specific check because BT Internet "helpfully" # redirects DNS fails to this address which hosts a sponsored # landing page! # if ip == "92.242.132.15": raise InvalidAddressError(host_or_ip, 0) return "%s:%s" % (ip, port) def action_and_params(commandline): """Treat a command line as an action followed by parameter :param commandline: a string containing at least an action :returns: action, [param1, param2, ...] """ components = shlex.split(commandline) return components[0], components[1:] def bytes_to_string(data): """Take bytes and return a base64-encoded unicode string equivalent :param data: a bytes object :returns: base64-encoded unicode object """ return base64.b64encode(data).decode("ascii") def string_to_bytes(data): """Take a base64-encoded unicode string and return the equivalent bytes :param data: a base64-encoded unicode object :returns: the equivalent bytes """ return base64.b64decode(data.encode("ascii"))
tjguk/networkzero
networkzero/core.py
_find_ip4_addresses
python
def _find_ip4_addresses(): global _ip4_addresses proto = socket.AF_INET if _ip4_addresses is None: _ip4_addresses = [] # # Determine the interface for the default gateway # (if any) and, later, prioritise the INET address on # that interface. # default_gateway = netifaces.gateways()['default'] if proto in default_gateway: _, default_gateway_interface = default_gateway[proto] else: default_gateway_interface = None for interface in netifaces.interfaces(): for info in netifaces.ifaddresses(interface).get(netifaces.AF_INET, []): if info['addr']: if interface == default_gateway_interface: _ip4_addresses.insert(0, info['addr']) else: _ip4_addresses.append(info['addr']) return _ip4_addresses
Find all the IP4 addresses currently bound to interfaces
train
https://github.com/tjguk/networkzero/blob/0e3e81d2e9200b25a83ac07741612283599486d7/networkzero/core.py#L136-L163
null
# -*- coding: utf-8 -*- import base64 import fnmatch import logging import random import shlex import socket try: import netifaces except ImportError: warnings.warn("Unable to import netifaces; using local fallback") from . import _netifaces as netifaces from . import config def get_logger(name): # # For now, this is just a hand-off to logging.getLogger # Later, though, we might want to add a null handler etc. # logger = logging.getLogger(name) logger.setLevel(logging.DEBUG) return logger _debug_logging_enabled = False def _enable_debug_logging(): global _debug_logging_enabled if not _debug_logging_enabled: logger = logging.getLogger("networkzero") handler = logging.FileHandler("network.log", "w", encoding="utf-8") handler.setFormatter(logging.Formatter("%(asctime)s %(threadName)s %(name)s %(levelname)s %(message)s")) handler.setLevel(logging.DEBUG) logger.addHandler(handler) _debug_logging_enabled = True _logger = get_logger(__name__) # # Common exceptions # class NetworkZeroError(Exception): pass class SocketAlreadyExistsError(NetworkZeroError): pass class SocketTimedOutError(NetworkZeroError): def __init__(self, n_seconds): self.n_seconds = n_seconds def __str__(self): return "Gave up waiting after %s seconds; this connection is now unusable" % self.n_seconds class SocketInterruptedError(NetworkZeroError): def __init__(self, after_n_seconds): self.after_n_seconds = after_n_seconds def __str__(self): return "Interrupted after %s seconds; this connection is now unusable" % self.after_n_seconds class AddressError(NetworkZeroError): pass class NoAddressFoundError(AddressError): pass class InvalidAddressError(NetworkZeroError): def __init__(self, address, errno=None): self.address = address self.errno = errno def __str__(self): message = "%s is not a valid address" % self.address if self.errno: message += "; the system returned an error of %d" % self.errno return message class DifferentThreadError(NetworkZeroError): pass # # Ports in the range 0xc000..0xffff are reserved # for dynamic allocation # PORT_POOL = list(config.DYNAMIC_PORTS) def split_address(address): if ":" in address: ip, _, port = address.partition(":") else: if address.isdigit(): ip, port = "", address else: ip, port = address, "" return ip, port def is_valid_ip_pattern(ip): """Check whether a string matches the outline of an IPv4 address, allowing "*" as a wildcard""" ip = ip.replace('*', '1') try: socket.inet_aton(ip) return True except socket.error: # Not a valid IPv4 address pattern return False def is_valid_port(port, port_range=range(65536)): try: return int(port) in port_range except ValueError: return False def is_valid_address(address, port_range=range(65536)): ip, port = split_address(address) return is_valid_ip_pattern(ip) and is_valid_port(port, port_range) def _find_ip4_broadcast_addresses(): """Yield each IP4 broadcast address, and the all-broadcast """ yield "255.255.255.255" for interface in netifaces.interfaces(): ifaddresses = netifaces.ifaddresses(interface) for family in ifaddresses: if family == netifaces.AF_INET: address_info = ifaddresses[family] for info in address_info: if "broadcast" in info: yield info['broadcast'] _ip4_addresses = None _ip4 = None _prefer = None def _find_ip4(prefer=None): global _ip4, _prefer # # Order the list of possible addresses on the machine: if any # address pattern is given as a preference (most -> least) # give it that weighting, otherwise treat all addresses # numerically. If no preference is given, prefer the most # likely useful local address range. # if prefer: _prefer = prefer else: _prefer = ["192.168.*"] def sorter(ip4): octets = [int(i) for i in ip4.split(".")] for n, pattern in enumerate(_prefer): if fnmatch.fnmatch(ip4, pattern): return n, octets else: # # Return the address itself if it doesn't match # a preference # return n + 1, octets ip4_addresses = _find_ip4_addresses() # # Pick an address allowing for user preference if stated # if not ip4_addresses: raise NoAddressFoundError else: # # Find the best match. If the user actually supplied a preference # list, assume an exact match is required to at least one of the # patterns. # ip4 = min(ip4_addresses, key=sorter) if prefer and not any(fnmatch.fnmatch(ip4, pattern) for pattern in prefer): raise NoAddressFoundError("No address matches any of: %s" % ", ".join(prefer)) else: _ip4 = ip4 return _ip4 def address(address=None): """Convert one of a number of inputs into a valid ip:port string. Elements which are not provided are filled in as follows: * IP Address: the system is asked for the set of IP addresses associated with the machine and the first one is used, preferring those matching `address` if it is a wildcard. * Port number: a random port is selected from the pool of dynamically-available port numbers. This means you can pass any of: nothing; a hostname; an IP address; an IP address with wildcards; a port number If an IP address is supplied but is invalid, an InvalidAddressError exception is raised. :param address: (optional) Any of: an IP address, a port number, or both :returns: a valid ip:port string for this machine """ address = str(address or "").strip() # # If the address is an ip:port pair, split into its component parts. # Otherwise, try to determine whether we're looking at an IP # or at a port and leave the other one blank # host_or_ip, port = split_address(address) # # If the port has been supplied, make sure it's numeric and that it's a valid # port number. If it hasn't been supplied, remove a random one from the pool # of possible dynamically-allocated ports and use that. # if port: try: port = int(port) except ValueError: raise AddressError("Port %s must be a number" % port) if port not in config.VALID_PORTS: raise AddressError("Port %d must be in range %d - %d" % ( port, min(config.VALID_PORTS), max(config.VALID_PORTS)) ) else: random.shuffle(PORT_POOL) port = PORT_POOL.pop() # # The address part could be an IP address (optionally including # wildcards to indicate a preference) or a hostname or nothing. # If it's a hostname we attempt to resolve it to an IP address. # It it's nothing or a wildcard we query the system for a matching IP address. # if (not host_or_ip) or is_valid_ip_pattern(host_or_ip): # # If a specific IP address is given, use that. # If an IP pattern is given (ie something with a wildcard in it) treat # that as no address with a preference for that wildcard. # prefer = None if "*" in host_or_ip: host_or_ip, prefer = None, [host_or_ip] # # If no IP (or only a wildcard) is specified, query the system for valid # addresses, preferring those which match the wildcard. NB if the preference # matches one we've previously used, we can return a cached address. But # different requests can specify different wildcard preferences. # if not host_or_ip: if _ip4 and _prefer == prefer: ip = _ip4 else: ip = _find_ip4(prefer) else: ip = host_or_ip else: # # Treat the string as a hostname and resolve to an IP4 address # try: ip = socket.gethostbyname(host_or_ip) except socket.gaierror as exc: _logger.error("gaierror %d for %s", exc.errno, host_or_ip) raise InvalidAddressError(host_or_ip, exc.errno) else: # # Bizarrely specific check because BT Internet "helpfully" # redirects DNS fails to this address which hosts a sponsored # landing page! # if ip == "92.242.132.15": raise InvalidAddressError(host_or_ip, 0) return "%s:%s" % (ip, port) def action_and_params(commandline): """Treat a command line as an action followed by parameter :param commandline: a string containing at least an action :returns: action, [param1, param2, ...] """ components = shlex.split(commandline) return components[0], components[1:] def bytes_to_string(data): """Take bytes and return a base64-encoded unicode string equivalent :param data: a bytes object :returns: base64-encoded unicode object """ return base64.b64encode(data).decode("ascii") def string_to_bytes(data): """Take a base64-encoded unicode string and return the equivalent bytes :param data: a base64-encoded unicode object :returns: the equivalent bytes """ return base64.b64decode(data.encode("ascii"))
tjguk/networkzero
networkzero/core.py
address
python
def address(address=None): address = str(address or "").strip() # # If the address is an ip:port pair, split into its component parts. # Otherwise, try to determine whether we're looking at an IP # or at a port and leave the other one blank # host_or_ip, port = split_address(address) # # If the port has been supplied, make sure it's numeric and that it's a valid # port number. If it hasn't been supplied, remove a random one from the pool # of possible dynamically-allocated ports and use that. # if port: try: port = int(port) except ValueError: raise AddressError("Port %s must be a number" % port) if port not in config.VALID_PORTS: raise AddressError("Port %d must be in range %d - %d" % ( port, min(config.VALID_PORTS), max(config.VALID_PORTS)) ) else: random.shuffle(PORT_POOL) port = PORT_POOL.pop() # # The address part could be an IP address (optionally including # wildcards to indicate a preference) or a hostname or nothing. # If it's a hostname we attempt to resolve it to an IP address. # It it's nothing or a wildcard we query the system for a matching IP address. # if (not host_or_ip) or is_valid_ip_pattern(host_or_ip): # # If a specific IP address is given, use that. # If an IP pattern is given (ie something with a wildcard in it) treat # that as no address with a preference for that wildcard. # prefer = None if "*" in host_or_ip: host_or_ip, prefer = None, [host_or_ip] # # If no IP (or only a wildcard) is specified, query the system for valid # addresses, preferring those which match the wildcard. NB if the preference # matches one we've previously used, we can return a cached address. But # different requests can specify different wildcard preferences. # if not host_or_ip: if _ip4 and _prefer == prefer: ip = _ip4 else: ip = _find_ip4(prefer) else: ip = host_or_ip else: # # Treat the string as a hostname and resolve to an IP4 address # try: ip = socket.gethostbyname(host_or_ip) except socket.gaierror as exc: _logger.error("gaierror %d for %s", exc.errno, host_or_ip) raise InvalidAddressError(host_or_ip, exc.errno) else: # # Bizarrely specific check because BT Internet "helpfully" # redirects DNS fails to this address which hosts a sponsored # landing page! # if ip == "92.242.132.15": raise InvalidAddressError(host_or_ip, 0) return "%s:%s" % (ip, port)
Convert one of a number of inputs into a valid ip:port string. Elements which are not provided are filled in as follows: * IP Address: the system is asked for the set of IP addresses associated with the machine and the first one is used, preferring those matching `address` if it is a wildcard. * Port number: a random port is selected from the pool of dynamically-available port numbers. This means you can pass any of: nothing; a hostname; an IP address; an IP address with wildcards; a port number If an IP address is supplied but is invalid, an InvalidAddressError exception is raised. :param address: (optional) Any of: an IP address, a port number, or both :returns: a valid ip:port string for this machine
train
https://github.com/tjguk/networkzero/blob/0e3e81d2e9200b25a83ac07741612283599486d7/networkzero/core.py#L212-L306
[ "def split_address(address):\n if \":\" in address:\n ip, _, port = address.partition(\":\")\n else:\n if address.isdigit():\n ip, port = \"\", address\n else:\n ip, port = address, \"\"\n return ip, port\n", "def is_valid_ip_pattern(ip):\n \"\"\"Check whethe...
# -*- coding: utf-8 -*- import base64 import fnmatch import logging import random import shlex import socket try: import netifaces except ImportError: warnings.warn("Unable to import netifaces; using local fallback") from . import _netifaces as netifaces from . import config def get_logger(name): # # For now, this is just a hand-off to logging.getLogger # Later, though, we might want to add a null handler etc. # logger = logging.getLogger(name) logger.setLevel(logging.DEBUG) return logger _debug_logging_enabled = False def _enable_debug_logging(): global _debug_logging_enabled if not _debug_logging_enabled: logger = logging.getLogger("networkzero") handler = logging.FileHandler("network.log", "w", encoding="utf-8") handler.setFormatter(logging.Formatter("%(asctime)s %(threadName)s %(name)s %(levelname)s %(message)s")) handler.setLevel(logging.DEBUG) logger.addHandler(handler) _debug_logging_enabled = True _logger = get_logger(__name__) # # Common exceptions # class NetworkZeroError(Exception): pass class SocketAlreadyExistsError(NetworkZeroError): pass class SocketTimedOutError(NetworkZeroError): def __init__(self, n_seconds): self.n_seconds = n_seconds def __str__(self): return "Gave up waiting after %s seconds; this connection is now unusable" % self.n_seconds class SocketInterruptedError(NetworkZeroError): def __init__(self, after_n_seconds): self.after_n_seconds = after_n_seconds def __str__(self): return "Interrupted after %s seconds; this connection is now unusable" % self.after_n_seconds class AddressError(NetworkZeroError): pass class NoAddressFoundError(AddressError): pass class InvalidAddressError(NetworkZeroError): def __init__(self, address, errno=None): self.address = address self.errno = errno def __str__(self): message = "%s is not a valid address" % self.address if self.errno: message += "; the system returned an error of %d" % self.errno return message class DifferentThreadError(NetworkZeroError): pass # # Ports in the range 0xc000..0xffff are reserved # for dynamic allocation # PORT_POOL = list(config.DYNAMIC_PORTS) def split_address(address): if ":" in address: ip, _, port = address.partition(":") else: if address.isdigit(): ip, port = "", address else: ip, port = address, "" return ip, port def is_valid_ip_pattern(ip): """Check whether a string matches the outline of an IPv4 address, allowing "*" as a wildcard""" ip = ip.replace('*', '1') try: socket.inet_aton(ip) return True except socket.error: # Not a valid IPv4 address pattern return False def is_valid_port(port, port_range=range(65536)): try: return int(port) in port_range except ValueError: return False def is_valid_address(address, port_range=range(65536)): ip, port = split_address(address) return is_valid_ip_pattern(ip) and is_valid_port(port, port_range) def _find_ip4_broadcast_addresses(): """Yield each IP4 broadcast address, and the all-broadcast """ yield "255.255.255.255" for interface in netifaces.interfaces(): ifaddresses = netifaces.ifaddresses(interface) for family in ifaddresses: if family == netifaces.AF_INET: address_info = ifaddresses[family] for info in address_info: if "broadcast" in info: yield info['broadcast'] _ip4_addresses = None def _find_ip4_addresses(): """Find all the IP4 addresses currently bound to interfaces """ global _ip4_addresses proto = socket.AF_INET if _ip4_addresses is None: _ip4_addresses = [] # # Determine the interface for the default gateway # (if any) and, later, prioritise the INET address on # that interface. # default_gateway = netifaces.gateways()['default'] if proto in default_gateway: _, default_gateway_interface = default_gateway[proto] else: default_gateway_interface = None for interface in netifaces.interfaces(): for info in netifaces.ifaddresses(interface).get(netifaces.AF_INET, []): if info['addr']: if interface == default_gateway_interface: _ip4_addresses.insert(0, info['addr']) else: _ip4_addresses.append(info['addr']) return _ip4_addresses _ip4 = None _prefer = None def _find_ip4(prefer=None): global _ip4, _prefer # # Order the list of possible addresses on the machine: if any # address pattern is given as a preference (most -> least) # give it that weighting, otherwise treat all addresses # numerically. If no preference is given, prefer the most # likely useful local address range. # if prefer: _prefer = prefer else: _prefer = ["192.168.*"] def sorter(ip4): octets = [int(i) for i in ip4.split(".")] for n, pattern in enumerate(_prefer): if fnmatch.fnmatch(ip4, pattern): return n, octets else: # # Return the address itself if it doesn't match # a preference # return n + 1, octets ip4_addresses = _find_ip4_addresses() # # Pick an address allowing for user preference if stated # if not ip4_addresses: raise NoAddressFoundError else: # # Find the best match. If the user actually supplied a preference # list, assume an exact match is required to at least one of the # patterns. # ip4 = min(ip4_addresses, key=sorter) if prefer and not any(fnmatch.fnmatch(ip4, pattern) for pattern in prefer): raise NoAddressFoundError("No address matches any of: %s" % ", ".join(prefer)) else: _ip4 = ip4 return _ip4 def action_and_params(commandline): """Treat a command line as an action followed by parameter :param commandline: a string containing at least an action :returns: action, [param1, param2, ...] """ components = shlex.split(commandline) return components[0], components[1:] def bytes_to_string(data): """Take bytes and return a base64-encoded unicode string equivalent :param data: a bytes object :returns: base64-encoded unicode object """ return base64.b64encode(data).decode("ascii") def string_to_bytes(data): """Take a base64-encoded unicode string and return the equivalent bytes :param data: a base64-encoded unicode object :returns: the equivalent bytes """ return base64.b64decode(data.encode("ascii"))
tjguk/networkzero
networkzero/discovery.py
_bind_with_timeout
python
def _bind_with_timeout(bind_function, args, n_tries=3, retry_interval_s=0.5): n_tries_left = n_tries while n_tries_left > 0: try: return bind_function(*args) except zmq.error.ZMQError as exc: _logger.warn("%s; %d tries remaining", exc, n_tries_left) n_tries_left -= 1 except OSError as exc: if exc.errno == errno.EADDRINUSE: _logger.warn("%s; %d tries remaining", exc, n_tries_left) n_tries_left -= 1 else: raise else: raise core.SocketAlreadyExistsError("Failed to bind after %s tries" % n_tries)
Attempt to bind a socket a number of times with a short interval in between Especially on Linux, crashing out of a networkzero process can leave the sockets lingering and unable to re-bind on startup. We give it a few goes here to see if we can bind within a couple of seconds.
train
https://github.com/tjguk/networkzero/blob/0e3e81d2e9200b25a83ac07741612283599486d7/networkzero/discovery.py#L80-L101
null
# -*- coding: utf-8 -*- """Advertise and collect advertisements of network services The discovery module offers: * A UDP broadcast socket which: - Listens for and keeps track of service adverts from this and other machines & processes - Broadcasts services advertised by this process * A ZeroMQ socket which allow any process on this machine to communicate with its broadcast socket In other words, we have a beacon which listens to instructions from processes on this machine while sending out and listening to adverts broadcast to/from all machines on the network. The beacon is started automatically in a daemon thread when the first attempt is made to advertise or discover. If another process already has a beacon running (ie if this beacon can't bind to its port) this beacon thread will shut down with no further action. The module-level functions to advertise and discover will open a connection to a ZeroMQ socket on this machine (which might be hosted by this or by another process) and will use this socket to send commands to the beacon thread which will update or return its internal list of advertised services. As an additional convenience, the :func:`advertise` function will, if given no specific address, generate a suitable ip:port pair by interrogating the system. This functionality is actually in :func:`networkzero.address` (qv). """ from __future__ import print_function import os, sys import atexit import collections import errno import json import logging import socket import threading import time import zmq from . import config from . import core from . import sockets _logger = core.get_logger(__name__) # # Continue is a sentinel value to indicate that a command # has completed its scheduled slice without producing a result # and without exceeding its overall timeout. # Continue = object() # # Empty is a sentinel to distinguish between no result and a result of None # Empty = object() def _unpack(message): return json.loads(message.decode(config.ENCODING)) def _pack(message): return json.dumps(message).encode(config.ENCODING) def timed_out(started_at, wait_for_s): # # If the wait time is the sentinel value FOREVER, never time out # Otherwise time out if the current time is more than wait_for_s seconds after the start time # if wait_for_s is config.FOREVER: return False else: return time.time() > started_at + wait_for_s def _bind_with_timeout(bind_function, args, n_tries=3, retry_interval_s=0.5): """Attempt to bind a socket a number of times with a short interval in between Especially on Linux, crashing out of a networkzero process can leave the sockets lingering and unable to re-bind on startup. We give it a few goes here to see if we can bind within a couple of seconds. """ n_tries_left = n_tries while n_tries_left > 0: try: return bind_function(*args) except zmq.error.ZMQError as exc: _logger.warn("%s; %d tries remaining", exc, n_tries_left) n_tries_left -= 1 except OSError as exc: if exc.errno == errno.EADDRINUSE: _logger.warn("%s; %d tries remaining", exc, n_tries_left) n_tries_left -= 1 else: raise else: raise core.SocketAlreadyExistsError("Failed to bind after %s tries" % n_tries) class _Service(object): """Convenience container with details of a service to be advertised Includes the name, address and when it is next due to be advertised and when it is due to expire if it was discovered. """ def __init__(self, name, address, ttl_s=None): self.name = name self.address = address self.ttl_s = ttl_s self.expires_at = None if ttl_s is None else (time.time() + ttl_s) self.advertise_at = 0 def __str__(self): return "_Service %s at %s due to advertise at %s and expire at %s" % ( self.name, self.address, time.ctime(self.advertise_at), time.ctime(self.expires_at) ) def __repr__(self): return "<%s: %s>" % (self.__class__.__name__, str(self)) def __eq__(self, other): return self.name == other.name class _Command(object): """Convenience container with details of a running command Includes the action ("discover", "advertise" etc.), its parameters, when it was started -- for timeout purposes -- and any response. This is used by the process_command functionality """ def __init__(self, action, params): self.action = action self.params = params self.started_at = time.time() self.response = Empty def __str__(self): return "_Command: %s (%s) started at %s -> %s" % (self.action, self.params, time.ctime(self.started_at), self.response) class _Beacon(threading.Thread): """Threaded beacon to: listen for adverts & broadcast adverts """ rpc_port = 9998 beacon_port = 9999 finder_timeout_s = 0.05 beacon_message_size = 256 time_between_broadcasts_s = config.BEACON_ADVERT_FREQUENCY_S def __init__(self, beacon_port=None): threading.Thread.__init__(self) self.setDaemon(True) self._stop_event = threading.Event() self._is_paused = False self.beacon_port = beacon_port or self.__class__.beacon_port _logger.debug("Using beacon port %s", self.beacon_port) # # Services we're advertising # self._services_to_advertise = collections.deque() # # Broadcast adverts which we've received (some of which will be our own) # self._services_found = {} # # _Command requests are collected on one queue # _Command responses are added to another # self._command = None # # Set the socket up to broadcast datagrams over UDP # self.broadcast_addresses = set(core._find_ip4_broadcast_addresses()) self.socket = socket.socket(socket.AF_INET, socket.SOCK_DGRAM, socket.IPPROTO_UDP) self.socket.setsockopt(socket.SOL_SOCKET, socket.SO_BROADCAST, 1) self.socket.bind(("", self.beacon_port)) # # Add the raw UDP socket to a ZeroMQ socket poller so we can check whether # it's received anything as part of the beacon's main event loop. # self.socket_fd = self.socket.fileno() self.poller = zmq.Poller() self.poller.register(self.socket, zmq.POLLIN) self.rpc = sockets.context.socket(zmq.REP) # # To avoid problems when restarting a beacon not long after it's been # closed, force the socket to shut down regardless about 1 second after # it's been closed. # self.rpc.linger = 1000 _bind_with_timeout(self.rpc.bind, ("tcp://127.0.0.1:%s" % self.rpc_port,)) def stop(self): _logger.debug("About to stop") self._stop_event.set() # # Commands available via RPC are methods whose name starts with "do_" # def do_advertise(self, started_at, name, address, fail_if_exists, ttl_s): _logger.debug("Advertise %s on %s %s TTL=%s", name, address, fail_if_exists, ttl_s) canonical_address = core.address(address) for service in self._services_to_advertise: if service.name == name: if fail_if_exists: _logger.error("_Service %s already exists on %s", name, service.address) return None else: _logger.warn("Superseding service %s which already exists on %s", name, service.address) service = _Service(name, canonical_address, ttl_s) self._services_to_advertise.append(service) # # As a shortcut, automatically "discover" any services we ourselves are advertising # self._services_found[name] = service return canonical_address def do_unadvertise(self, started_at, name): _logger.debug("Unadvertise %s", name) for service in self._services_to_advertise: if service.name == name: self._services_to_advertise.remove(service) break else: _logger.warn("No advert found for %s", name) _logger.debug("Services now: %s", self._services_to_advertise) def do_pause(self, started_at): _logger.debug("Pause") self._is_paused = True def do_resume(self, started_at): _logger.debug("Resume") self._is_paused = False def do_discover(self, started_at, name, wait_for_s): _logger.debug("Discover %s waiting for %s seconds", name, wait_for_s) discovered = self._services_found.get(name) # # If we've got a match, return it. Otherwise: # * If we're due to wait for ever, continue # * If we're out of time return None # * Otherwise we've still got time left: continue # if discovered: return discovered.address if timed_out(started_at, wait_for_s): return None else: return Continue def do_discover_all(self, started_at): _logger.debug("Discover all") return [(service.name, service.address) for service in self._services_found.values()] def do_reset(self, started_at): _logger.debug("Reset") self.do_pause(started_at) self._services_found.clear() self._services_to_advertise.clear() self.do_resume(started_at) def do_stop(self, started_at): _logger.debug("Stop") self.stop() def listen_for_one_advert(self): events = dict(self.poller.poll(1000 * self.finder_timeout_s)) if self.socket_fd not in events: return message, source = self.socket.recvfrom(self.beacon_message_size) _logger.debug("Broadcast message received: %r", message) service_name, service_address, ttl_s = _unpack(message) service = _Service(service_name, service_address, ttl_s) self._services_found[service_name] = service def broadcast_one_advert(self): if self._services_to_advertise: next_service = self._services_to_advertise[0] if next_service.advertise_at < time.time(): _logger.debug("%s due to advertise at %s", next_service.name, time.ctime(next_service.advertise_at)) message = _pack([next_service.name, next_service.address, next_service.ttl_s]) for broadcast_address in self.broadcast_addresses: _logger.debug("Advertising on %s", broadcast_address) self.socket.sendto(message, 0, (broadcast_address, self.beacon_port)) next_service.advertise_at = time.time() + self.time_between_broadcasts_s self._services_to_advertise.rotate(-1) def remove_expired_adverts(self): for name, service in list(self._services_found.items()): # # A service with an empty expiry time never expired # if service.expires_at is None: continue if service.expires_at <= time.time(): _logger.warn("Removing advert for %s which expired at %s", name, time.ctime(service.expires_at)) del self._services_found[name] def poll_command_request(self): """If the command RPC socket has an incoming request, separate it into its action and its params and put it on the command request queue. """ try: message = self.rpc.recv(zmq.NOBLOCK) except zmq.ZMQError as exc: if exc.errno == zmq.EAGAIN: return else: raise _logger.debug("Received command %s", message) segments = _unpack(message) action, params = segments[0], segments[1:] _logger.debug("Adding %s, %s to the request queue", action, params) self._command = _Command(action, params) def process_command(self): if not self._command: return else: _logger.debug("process_command: %s", self._command.action) command = self._command _logger.debug("Picked %s, %s, %s", self._command.action, self._command.params, self._command.started_at) function = getattr(self, "do_" + command.action.lower(), None) if not function: raise NotImplementedError("%s is not a valid action") else: try: result = function(command.started_at, *command.params) except: _logger.exception("Problem calling %s with %s", command.action, command.params) result = None _logger.debug("result = %s", result) # # result will be Continue if the action cannot be completed # (eg a discovery) but its time is not yet expired. Leave # the command on the stack for now. # if result is Continue: return # # If we get a result, add the result to the response # queue and pop the request off the stack. # self._command.response = result def poll_command_reponse(self): """If the latest request has a response, issue it as a reply to the RPC socket. """ if self._command.response is not Empty: _logger.debug("Sending response %s", self._command.response) self.rpc.send(_pack(self._command.response)) self._command = None def run(self): _logger.info("Starting discovery") while not self._stop_event.wait(0): try: # # If we're not already processing one, check for an command # to advertise/discover from a local process. # if not self._command: self.poll_command_request() # # If we're paused no adverts will be broadcast. Adverts # will be received and stale ones expired # if not self._is_paused: # # Broadcast the first advert whose advertising schedule # has arrived # self.broadcast_one_advert() # # See if an advert broadcast has arrived # self.listen_for_one_advert() # # See if any adverts have expired # self.remove_expired_adverts() # # If we're processing a command, see if it's complete # if self._command: self.process_command() self.poll_command_reponse() except: _logger.exception("Problem in beacon thread") break _logger.info("Ending discovery") self.rpc.close() self.socket.close() _beacon = None _remote_beacon = object() def _start_beacon(port=None): """Start a beacon thread within this process if no beacon is currently running on this machine. In general this is called automatically when an attempt is made to advertise or discover. It might be convenient, though, to call this function directly if you want to have a process whose only job is to host this beacon so that it doesn't shut down when other processes shut down. """ global _beacon if _beacon is None: _logger.debug("About to start beacon with port %s", port) try: _beacon = _Beacon(port) except (OSError, socket.error) as exc: if exc.errno == errno.EADDRINUSE: _logger.warn("Beacon already active on this machine") # # _remote_beacon is simply a not-None sentinel value # to distinguish between the case where we have not # yet started a beacon and where we have found one # in another process. # _beacon = _remote_beacon else: raise else: _beacon.start() def _stop_beacon(): # # Mostly for testing: shutdown the beacon if it's running # locally and clear it globally so the next attempt will # start fresh. # global _beacon if _beacon and _beacon is not _remote_beacon: _beacon.stop() _beacon.join() _beacon = None def _rpc(action, *args, **kwargs): _logger.debug("About to send rpc request %s with args %s, kwargs %s", action, args, kwargs) wait_for_s = kwargs.pop("wait_for_s", 5) with sockets.context.socket(zmq.REQ) as socket: # # To avoid problems when restarting a beacon not long after it's been # closed, force the socket to shut down regardless about 1 second after # it's been closed. # socket.connect("tcp://localhost:%s" % _Beacon.rpc_port) socket.send(_pack([action] + list(args))) reply = sockets._sockets._receive_with_timeout(socket, wait_for_s) return _unpack(reply) def _pause(): return _rpc("pause") def _resume(): return _rpc("resume") _services_advertised = {} def advertise(name, address=None, fail_if_exists=False, ttl_s=config.ADVERT_TTL_S): """Advertise a name at an address Start to advertise service `name` at address `address`. If the address is not supplied, one is constructed and this is returned by the function. ie this is a typical use:: address = nw0.advertise("myservice") :param name: any text :param address: either "ip:port" or None :param fail_if_exists: fail if this name is already registered? :param ttl_s: the advert will persist for this many seconds other beacons :returns: the address given or constructed """ _start_beacon() address = _rpc("advertise", name, address, fail_if_exists, ttl_s) _services_advertised[name] = address return address def _unadvertise_all(): """Remove all adverts """ for name in _services_advertised: try: _unadvertise(name) except core.SocketTimedOutError: _logger.warn("Timed out trying to unadvertise") break atexit.register(_unadvertise_all) def _unadvertise(name): """Remove the advert for a name This is intended for internal use only at the moment. When a process exits it can remove adverts for its services from the beacon running on that machine. (Of course, if the beacon thread is part of of the same service, all its adverts will cease). """ _start_beacon() return _rpc("unadvertise", name) def discover(name, wait_for_s=60): """Discover a service by name Look for an advert to a named service:: address = nw0.discover("myservice") :param name: any text :param wait_for_s: how many seconds to wait before giving up :returns: the address found or None """ _start_beacon() # # It's possible to enter a deadlock situation where the first # process fires off a discovery request and waits for the # second process to advertise. But the second process has to # connect to the rpc port of the first process' beacon and # its advertisement is queued behind the pending discovery. # # To give both a chance of succeeding we operate in bursts, # allowing them to interleave. # t0 = time.time() while True: discovery = _rpc("discover", name, 0.5) if discovery: return discovery if timed_out(t0, wait_for_s): return None def discover_all(): """Produce a list of all known services and their addresses Ask for all known services as a list of 2-tuples: (name, address) This could, eg, be used to form a dictionary of services:: services = dict(nw0.discover_all()) :returns: a list of 2-tuples [(name, address), ...] """ _start_beacon() return _rpc("discover_all") def discover_group(group, separator="/", exclude=None): """Produce a list of all services and their addresses in a group A group is an optional form of namespace within the discovery mechanism. If an advertised name has the form <group><sep><name> it is deemed to belong to <group>. Note that the service's name is still the full string <group><sep><name>. The group concept is simply for discovery and to assist differentiation, eg, in a classroom group. :param group: the name of a group prefix :param separator: the separator character [/] :param exclude: an iterable of names to exclude (or None) :returns: a list of 2-tuples [(name, address), ...] """ _start_beacon() if exclude is None: names_to_exclude = set() else: names_to_exclude = set(exclude) all_discovered = _rpc("discover_all") return [(name, address) for (name, address) in all_discovered if name.startswith("%s%s" % (group, separator)) and name not in names_to_exclude ] def reset_beacon(): """Clear the adverts which the beacon is carrying (This is mostly useful when testing, to get a fresh start) """ _start_beacon() return _rpc("reset") if __name__ == '__main__': params = [arg.lower() for arg in sys.argv] if "--debug" in params: logging_level = logging.DEBUG else: logging_level = logging.INFO handler = logging.StreamHandler() handler.setLevel(logging_level) handler.setFormatter(logging.Formatter("%(levelname)s - %(message)s")) _logger.addHandler(handler) _start_beacon() _logger.info("Beacon started at %s", time.asctime()) while True: time.sleep(1)
tjguk/networkzero
networkzero/discovery.py
_start_beacon
python
def _start_beacon(port=None): global _beacon if _beacon is None: _logger.debug("About to start beacon with port %s", port) try: _beacon = _Beacon(port) except (OSError, socket.error) as exc: if exc.errno == errno.EADDRINUSE: _logger.warn("Beacon already active on this machine") # # _remote_beacon is simply a not-None sentinel value # to distinguish between the case where we have not # yet started a beacon and where we have found one # in another process. # _beacon = _remote_beacon else: raise else: _beacon.start()
Start a beacon thread within this process if no beacon is currently running on this machine. In general this is called automatically when an attempt is made to advertise or discover. It might be convenient, though, to call this function directly if you want to have a process whose only job is to host this beacon so that it doesn't shut down when other processes shut down.
train
https://github.com/tjguk/networkzero/blob/0e3e81d2e9200b25a83ac07741612283599486d7/networkzero/discovery.py#L433-L461
null
# -*- coding: utf-8 -*- """Advertise and collect advertisements of network services The discovery module offers: * A UDP broadcast socket which: - Listens for and keeps track of service adverts from this and other machines & processes - Broadcasts services advertised by this process * A ZeroMQ socket which allow any process on this machine to communicate with its broadcast socket In other words, we have a beacon which listens to instructions from processes on this machine while sending out and listening to adverts broadcast to/from all machines on the network. The beacon is started automatically in a daemon thread when the first attempt is made to advertise or discover. If another process already has a beacon running (ie if this beacon can't bind to its port) this beacon thread will shut down with no further action. The module-level functions to advertise and discover will open a connection to a ZeroMQ socket on this machine (which might be hosted by this or by another process) and will use this socket to send commands to the beacon thread which will update or return its internal list of advertised services. As an additional convenience, the :func:`advertise` function will, if given no specific address, generate a suitable ip:port pair by interrogating the system. This functionality is actually in :func:`networkzero.address` (qv). """ from __future__ import print_function import os, sys import atexit import collections import errno import json import logging import socket import threading import time import zmq from . import config from . import core from . import sockets _logger = core.get_logger(__name__) # # Continue is a sentinel value to indicate that a command # has completed its scheduled slice without producing a result # and without exceeding its overall timeout. # Continue = object() # # Empty is a sentinel to distinguish between no result and a result of None # Empty = object() def _unpack(message): return json.loads(message.decode(config.ENCODING)) def _pack(message): return json.dumps(message).encode(config.ENCODING) def timed_out(started_at, wait_for_s): # # If the wait time is the sentinel value FOREVER, never time out # Otherwise time out if the current time is more than wait_for_s seconds after the start time # if wait_for_s is config.FOREVER: return False else: return time.time() > started_at + wait_for_s def _bind_with_timeout(bind_function, args, n_tries=3, retry_interval_s=0.5): """Attempt to bind a socket a number of times with a short interval in between Especially on Linux, crashing out of a networkzero process can leave the sockets lingering and unable to re-bind on startup. We give it a few goes here to see if we can bind within a couple of seconds. """ n_tries_left = n_tries while n_tries_left > 0: try: return bind_function(*args) except zmq.error.ZMQError as exc: _logger.warn("%s; %d tries remaining", exc, n_tries_left) n_tries_left -= 1 except OSError as exc: if exc.errno == errno.EADDRINUSE: _logger.warn("%s; %d tries remaining", exc, n_tries_left) n_tries_left -= 1 else: raise else: raise core.SocketAlreadyExistsError("Failed to bind after %s tries" % n_tries) class _Service(object): """Convenience container with details of a service to be advertised Includes the name, address and when it is next due to be advertised and when it is due to expire if it was discovered. """ def __init__(self, name, address, ttl_s=None): self.name = name self.address = address self.ttl_s = ttl_s self.expires_at = None if ttl_s is None else (time.time() + ttl_s) self.advertise_at = 0 def __str__(self): return "_Service %s at %s due to advertise at %s and expire at %s" % ( self.name, self.address, time.ctime(self.advertise_at), time.ctime(self.expires_at) ) def __repr__(self): return "<%s: %s>" % (self.__class__.__name__, str(self)) def __eq__(self, other): return self.name == other.name class _Command(object): """Convenience container with details of a running command Includes the action ("discover", "advertise" etc.), its parameters, when it was started -- for timeout purposes -- and any response. This is used by the process_command functionality """ def __init__(self, action, params): self.action = action self.params = params self.started_at = time.time() self.response = Empty def __str__(self): return "_Command: %s (%s) started at %s -> %s" % (self.action, self.params, time.ctime(self.started_at), self.response) class _Beacon(threading.Thread): """Threaded beacon to: listen for adverts & broadcast adverts """ rpc_port = 9998 beacon_port = 9999 finder_timeout_s = 0.05 beacon_message_size = 256 time_between_broadcasts_s = config.BEACON_ADVERT_FREQUENCY_S def __init__(self, beacon_port=None): threading.Thread.__init__(self) self.setDaemon(True) self._stop_event = threading.Event() self._is_paused = False self.beacon_port = beacon_port or self.__class__.beacon_port _logger.debug("Using beacon port %s", self.beacon_port) # # Services we're advertising # self._services_to_advertise = collections.deque() # # Broadcast adverts which we've received (some of which will be our own) # self._services_found = {} # # _Command requests are collected on one queue # _Command responses are added to another # self._command = None # # Set the socket up to broadcast datagrams over UDP # self.broadcast_addresses = set(core._find_ip4_broadcast_addresses()) self.socket = socket.socket(socket.AF_INET, socket.SOCK_DGRAM, socket.IPPROTO_UDP) self.socket.setsockopt(socket.SOL_SOCKET, socket.SO_BROADCAST, 1) self.socket.bind(("", self.beacon_port)) # # Add the raw UDP socket to a ZeroMQ socket poller so we can check whether # it's received anything as part of the beacon's main event loop. # self.socket_fd = self.socket.fileno() self.poller = zmq.Poller() self.poller.register(self.socket, zmq.POLLIN) self.rpc = sockets.context.socket(zmq.REP) # # To avoid problems when restarting a beacon not long after it's been # closed, force the socket to shut down regardless about 1 second after # it's been closed. # self.rpc.linger = 1000 _bind_with_timeout(self.rpc.bind, ("tcp://127.0.0.1:%s" % self.rpc_port,)) def stop(self): _logger.debug("About to stop") self._stop_event.set() # # Commands available via RPC are methods whose name starts with "do_" # def do_advertise(self, started_at, name, address, fail_if_exists, ttl_s): _logger.debug("Advertise %s on %s %s TTL=%s", name, address, fail_if_exists, ttl_s) canonical_address = core.address(address) for service in self._services_to_advertise: if service.name == name: if fail_if_exists: _logger.error("_Service %s already exists on %s", name, service.address) return None else: _logger.warn("Superseding service %s which already exists on %s", name, service.address) service = _Service(name, canonical_address, ttl_s) self._services_to_advertise.append(service) # # As a shortcut, automatically "discover" any services we ourselves are advertising # self._services_found[name] = service return canonical_address def do_unadvertise(self, started_at, name): _logger.debug("Unadvertise %s", name) for service in self._services_to_advertise: if service.name == name: self._services_to_advertise.remove(service) break else: _logger.warn("No advert found for %s", name) _logger.debug("Services now: %s", self._services_to_advertise) def do_pause(self, started_at): _logger.debug("Pause") self._is_paused = True def do_resume(self, started_at): _logger.debug("Resume") self._is_paused = False def do_discover(self, started_at, name, wait_for_s): _logger.debug("Discover %s waiting for %s seconds", name, wait_for_s) discovered = self._services_found.get(name) # # If we've got a match, return it. Otherwise: # * If we're due to wait for ever, continue # * If we're out of time return None # * Otherwise we've still got time left: continue # if discovered: return discovered.address if timed_out(started_at, wait_for_s): return None else: return Continue def do_discover_all(self, started_at): _logger.debug("Discover all") return [(service.name, service.address) for service in self._services_found.values()] def do_reset(self, started_at): _logger.debug("Reset") self.do_pause(started_at) self._services_found.clear() self._services_to_advertise.clear() self.do_resume(started_at) def do_stop(self, started_at): _logger.debug("Stop") self.stop() def listen_for_one_advert(self): events = dict(self.poller.poll(1000 * self.finder_timeout_s)) if self.socket_fd not in events: return message, source = self.socket.recvfrom(self.beacon_message_size) _logger.debug("Broadcast message received: %r", message) service_name, service_address, ttl_s = _unpack(message) service = _Service(service_name, service_address, ttl_s) self._services_found[service_name] = service def broadcast_one_advert(self): if self._services_to_advertise: next_service = self._services_to_advertise[0] if next_service.advertise_at < time.time(): _logger.debug("%s due to advertise at %s", next_service.name, time.ctime(next_service.advertise_at)) message = _pack([next_service.name, next_service.address, next_service.ttl_s]) for broadcast_address in self.broadcast_addresses: _logger.debug("Advertising on %s", broadcast_address) self.socket.sendto(message, 0, (broadcast_address, self.beacon_port)) next_service.advertise_at = time.time() + self.time_between_broadcasts_s self._services_to_advertise.rotate(-1) def remove_expired_adverts(self): for name, service in list(self._services_found.items()): # # A service with an empty expiry time never expired # if service.expires_at is None: continue if service.expires_at <= time.time(): _logger.warn("Removing advert for %s which expired at %s", name, time.ctime(service.expires_at)) del self._services_found[name] def poll_command_request(self): """If the command RPC socket has an incoming request, separate it into its action and its params and put it on the command request queue. """ try: message = self.rpc.recv(zmq.NOBLOCK) except zmq.ZMQError as exc: if exc.errno == zmq.EAGAIN: return else: raise _logger.debug("Received command %s", message) segments = _unpack(message) action, params = segments[0], segments[1:] _logger.debug("Adding %s, %s to the request queue", action, params) self._command = _Command(action, params) def process_command(self): if not self._command: return else: _logger.debug("process_command: %s", self._command.action) command = self._command _logger.debug("Picked %s, %s, %s", self._command.action, self._command.params, self._command.started_at) function = getattr(self, "do_" + command.action.lower(), None) if not function: raise NotImplementedError("%s is not a valid action") else: try: result = function(command.started_at, *command.params) except: _logger.exception("Problem calling %s with %s", command.action, command.params) result = None _logger.debug("result = %s", result) # # result will be Continue if the action cannot be completed # (eg a discovery) but its time is not yet expired. Leave # the command on the stack for now. # if result is Continue: return # # If we get a result, add the result to the response # queue and pop the request off the stack. # self._command.response = result def poll_command_reponse(self): """If the latest request has a response, issue it as a reply to the RPC socket. """ if self._command.response is not Empty: _logger.debug("Sending response %s", self._command.response) self.rpc.send(_pack(self._command.response)) self._command = None def run(self): _logger.info("Starting discovery") while not self._stop_event.wait(0): try: # # If we're not already processing one, check for an command # to advertise/discover from a local process. # if not self._command: self.poll_command_request() # # If we're paused no adverts will be broadcast. Adverts # will be received and stale ones expired # if not self._is_paused: # # Broadcast the first advert whose advertising schedule # has arrived # self.broadcast_one_advert() # # See if an advert broadcast has arrived # self.listen_for_one_advert() # # See if any adverts have expired # self.remove_expired_adverts() # # If we're processing a command, see if it's complete # if self._command: self.process_command() self.poll_command_reponse() except: _logger.exception("Problem in beacon thread") break _logger.info("Ending discovery") self.rpc.close() self.socket.close() _beacon = None _remote_beacon = object() def _start_beacon(port=None): """Start a beacon thread within this process if no beacon is currently running on this machine. In general this is called automatically when an attempt is made to advertise or discover. It might be convenient, though, to call this function directly if you want to have a process whose only job is to host this beacon so that it doesn't shut down when other processes shut down. """ global _beacon if _beacon is None: _logger.debug("About to start beacon with port %s", port) try: _beacon = _Beacon(port) except (OSError, socket.error) as exc: if exc.errno == errno.EADDRINUSE: _logger.warn("Beacon already active on this machine") # # _remote_beacon is simply a not-None sentinel value # to distinguish between the case where we have not # yet started a beacon and where we have found one # in another process. # _beacon = _remote_beacon else: raise else: _beacon.start() def _stop_beacon(): # # Mostly for testing: shutdown the beacon if it's running # locally and clear it globally so the next attempt will # start fresh. # global _beacon if _beacon and _beacon is not _remote_beacon: _beacon.stop() _beacon.join() _beacon = None def _rpc(action, *args, **kwargs): _logger.debug("About to send rpc request %s with args %s, kwargs %s", action, args, kwargs) wait_for_s = kwargs.pop("wait_for_s", 5) with sockets.context.socket(zmq.REQ) as socket: # # To avoid problems when restarting a beacon not long after it's been # closed, force the socket to shut down regardless about 1 second after # it's been closed. # socket.connect("tcp://localhost:%s" % _Beacon.rpc_port) socket.send(_pack([action] + list(args))) reply = sockets._sockets._receive_with_timeout(socket, wait_for_s) return _unpack(reply) def _pause(): return _rpc("pause") def _resume(): return _rpc("resume") _services_advertised = {} def advertise(name, address=None, fail_if_exists=False, ttl_s=config.ADVERT_TTL_S): """Advertise a name at an address Start to advertise service `name` at address `address`. If the address is not supplied, one is constructed and this is returned by the function. ie this is a typical use:: address = nw0.advertise("myservice") :param name: any text :param address: either "ip:port" or None :param fail_if_exists: fail if this name is already registered? :param ttl_s: the advert will persist for this many seconds other beacons :returns: the address given or constructed """ _start_beacon() address = _rpc("advertise", name, address, fail_if_exists, ttl_s) _services_advertised[name] = address return address def _unadvertise_all(): """Remove all adverts """ for name in _services_advertised: try: _unadvertise(name) except core.SocketTimedOutError: _logger.warn("Timed out trying to unadvertise") break atexit.register(_unadvertise_all) def _unadvertise(name): """Remove the advert for a name This is intended for internal use only at the moment. When a process exits it can remove adverts for its services from the beacon running on that machine. (Of course, if the beacon thread is part of of the same service, all its adverts will cease). """ _start_beacon() return _rpc("unadvertise", name) def discover(name, wait_for_s=60): """Discover a service by name Look for an advert to a named service:: address = nw0.discover("myservice") :param name: any text :param wait_for_s: how many seconds to wait before giving up :returns: the address found or None """ _start_beacon() # # It's possible to enter a deadlock situation where the first # process fires off a discovery request and waits for the # second process to advertise. But the second process has to # connect to the rpc port of the first process' beacon and # its advertisement is queued behind the pending discovery. # # To give both a chance of succeeding we operate in bursts, # allowing them to interleave. # t0 = time.time() while True: discovery = _rpc("discover", name, 0.5) if discovery: return discovery if timed_out(t0, wait_for_s): return None def discover_all(): """Produce a list of all known services and their addresses Ask for all known services as a list of 2-tuples: (name, address) This could, eg, be used to form a dictionary of services:: services = dict(nw0.discover_all()) :returns: a list of 2-tuples [(name, address), ...] """ _start_beacon() return _rpc("discover_all") def discover_group(group, separator="/", exclude=None): """Produce a list of all services and their addresses in a group A group is an optional form of namespace within the discovery mechanism. If an advertised name has the form <group><sep><name> it is deemed to belong to <group>. Note that the service's name is still the full string <group><sep><name>. The group concept is simply for discovery and to assist differentiation, eg, in a classroom group. :param group: the name of a group prefix :param separator: the separator character [/] :param exclude: an iterable of names to exclude (or None) :returns: a list of 2-tuples [(name, address), ...] """ _start_beacon() if exclude is None: names_to_exclude = set() else: names_to_exclude = set(exclude) all_discovered = _rpc("discover_all") return [(name, address) for (name, address) in all_discovered if name.startswith("%s%s" % (group, separator)) and name not in names_to_exclude ] def reset_beacon(): """Clear the adverts which the beacon is carrying (This is mostly useful when testing, to get a fresh start) """ _start_beacon() return _rpc("reset") if __name__ == '__main__': params = [arg.lower() for arg in sys.argv] if "--debug" in params: logging_level = logging.DEBUG else: logging_level = logging.INFO handler = logging.StreamHandler() handler.setLevel(logging_level) handler.setFormatter(logging.Formatter("%(levelname)s - %(message)s")) _logger.addHandler(handler) _start_beacon() _logger.info("Beacon started at %s", time.asctime()) while True: time.sleep(1)
tjguk/networkzero
networkzero/discovery.py
advertise
python
def advertise(name, address=None, fail_if_exists=False, ttl_s=config.ADVERT_TTL_S): _start_beacon() address = _rpc("advertise", name, address, fail_if_exists, ttl_s) _services_advertised[name] = address return address
Advertise a name at an address Start to advertise service `name` at address `address`. If the address is not supplied, one is constructed and this is returned by the function. ie this is a typical use:: address = nw0.advertise("myservice") :param name: any text :param address: either "ip:port" or None :param fail_if_exists: fail if this name is already registered? :param ttl_s: the advert will persist for this many seconds other beacons :returns: the address given or constructed
train
https://github.com/tjguk/networkzero/blob/0e3e81d2e9200b25a83ac07741612283599486d7/networkzero/discovery.py#L497-L515
[ "def _start_beacon(port=None):\n \"\"\"Start a beacon thread within this process if no beacon is currently\n running on this machine.\n\n In general this is called automatically when an attempt is made to\n advertise or discover. It might be convenient, though, to call this\n function directly if you...
# -*- coding: utf-8 -*- """Advertise and collect advertisements of network services The discovery module offers: * A UDP broadcast socket which: - Listens for and keeps track of service adverts from this and other machines & processes - Broadcasts services advertised by this process * A ZeroMQ socket which allow any process on this machine to communicate with its broadcast socket In other words, we have a beacon which listens to instructions from processes on this machine while sending out and listening to adverts broadcast to/from all machines on the network. The beacon is started automatically in a daemon thread when the first attempt is made to advertise or discover. If another process already has a beacon running (ie if this beacon can't bind to its port) this beacon thread will shut down with no further action. The module-level functions to advertise and discover will open a connection to a ZeroMQ socket on this machine (which might be hosted by this or by another process) and will use this socket to send commands to the beacon thread which will update or return its internal list of advertised services. As an additional convenience, the :func:`advertise` function will, if given no specific address, generate a suitable ip:port pair by interrogating the system. This functionality is actually in :func:`networkzero.address` (qv). """ from __future__ import print_function import os, sys import atexit import collections import errno import json import logging import socket import threading import time import zmq from . import config from . import core from . import sockets _logger = core.get_logger(__name__) # # Continue is a sentinel value to indicate that a command # has completed its scheduled slice without producing a result # and without exceeding its overall timeout. # Continue = object() # # Empty is a sentinel to distinguish between no result and a result of None # Empty = object() def _unpack(message): return json.loads(message.decode(config.ENCODING)) def _pack(message): return json.dumps(message).encode(config.ENCODING) def timed_out(started_at, wait_for_s): # # If the wait time is the sentinel value FOREVER, never time out # Otherwise time out if the current time is more than wait_for_s seconds after the start time # if wait_for_s is config.FOREVER: return False else: return time.time() > started_at + wait_for_s def _bind_with_timeout(bind_function, args, n_tries=3, retry_interval_s=0.5): """Attempt to bind a socket a number of times with a short interval in between Especially on Linux, crashing out of a networkzero process can leave the sockets lingering and unable to re-bind on startup. We give it a few goes here to see if we can bind within a couple of seconds. """ n_tries_left = n_tries while n_tries_left > 0: try: return bind_function(*args) except zmq.error.ZMQError as exc: _logger.warn("%s; %d tries remaining", exc, n_tries_left) n_tries_left -= 1 except OSError as exc: if exc.errno == errno.EADDRINUSE: _logger.warn("%s; %d tries remaining", exc, n_tries_left) n_tries_left -= 1 else: raise else: raise core.SocketAlreadyExistsError("Failed to bind after %s tries" % n_tries) class _Service(object): """Convenience container with details of a service to be advertised Includes the name, address and when it is next due to be advertised and when it is due to expire if it was discovered. """ def __init__(self, name, address, ttl_s=None): self.name = name self.address = address self.ttl_s = ttl_s self.expires_at = None if ttl_s is None else (time.time() + ttl_s) self.advertise_at = 0 def __str__(self): return "_Service %s at %s due to advertise at %s and expire at %s" % ( self.name, self.address, time.ctime(self.advertise_at), time.ctime(self.expires_at) ) def __repr__(self): return "<%s: %s>" % (self.__class__.__name__, str(self)) def __eq__(self, other): return self.name == other.name class _Command(object): """Convenience container with details of a running command Includes the action ("discover", "advertise" etc.), its parameters, when it was started -- for timeout purposes -- and any response. This is used by the process_command functionality """ def __init__(self, action, params): self.action = action self.params = params self.started_at = time.time() self.response = Empty def __str__(self): return "_Command: %s (%s) started at %s -> %s" % (self.action, self.params, time.ctime(self.started_at), self.response) class _Beacon(threading.Thread): """Threaded beacon to: listen for adverts & broadcast adverts """ rpc_port = 9998 beacon_port = 9999 finder_timeout_s = 0.05 beacon_message_size = 256 time_between_broadcasts_s = config.BEACON_ADVERT_FREQUENCY_S def __init__(self, beacon_port=None): threading.Thread.__init__(self) self.setDaemon(True) self._stop_event = threading.Event() self._is_paused = False self.beacon_port = beacon_port or self.__class__.beacon_port _logger.debug("Using beacon port %s", self.beacon_port) # # Services we're advertising # self._services_to_advertise = collections.deque() # # Broadcast adverts which we've received (some of which will be our own) # self._services_found = {} # # _Command requests are collected on one queue # _Command responses are added to another # self._command = None # # Set the socket up to broadcast datagrams over UDP # self.broadcast_addresses = set(core._find_ip4_broadcast_addresses()) self.socket = socket.socket(socket.AF_INET, socket.SOCK_DGRAM, socket.IPPROTO_UDP) self.socket.setsockopt(socket.SOL_SOCKET, socket.SO_BROADCAST, 1) self.socket.bind(("", self.beacon_port)) # # Add the raw UDP socket to a ZeroMQ socket poller so we can check whether # it's received anything as part of the beacon's main event loop. # self.socket_fd = self.socket.fileno() self.poller = zmq.Poller() self.poller.register(self.socket, zmq.POLLIN) self.rpc = sockets.context.socket(zmq.REP) # # To avoid problems when restarting a beacon not long after it's been # closed, force the socket to shut down regardless about 1 second after # it's been closed. # self.rpc.linger = 1000 _bind_with_timeout(self.rpc.bind, ("tcp://127.0.0.1:%s" % self.rpc_port,)) def stop(self): _logger.debug("About to stop") self._stop_event.set() # # Commands available via RPC are methods whose name starts with "do_" # def do_advertise(self, started_at, name, address, fail_if_exists, ttl_s): _logger.debug("Advertise %s on %s %s TTL=%s", name, address, fail_if_exists, ttl_s) canonical_address = core.address(address) for service in self._services_to_advertise: if service.name == name: if fail_if_exists: _logger.error("_Service %s already exists on %s", name, service.address) return None else: _logger.warn("Superseding service %s which already exists on %s", name, service.address) service = _Service(name, canonical_address, ttl_s) self._services_to_advertise.append(service) # # As a shortcut, automatically "discover" any services we ourselves are advertising # self._services_found[name] = service return canonical_address def do_unadvertise(self, started_at, name): _logger.debug("Unadvertise %s", name) for service in self._services_to_advertise: if service.name == name: self._services_to_advertise.remove(service) break else: _logger.warn("No advert found for %s", name) _logger.debug("Services now: %s", self._services_to_advertise) def do_pause(self, started_at): _logger.debug("Pause") self._is_paused = True def do_resume(self, started_at): _logger.debug("Resume") self._is_paused = False def do_discover(self, started_at, name, wait_for_s): _logger.debug("Discover %s waiting for %s seconds", name, wait_for_s) discovered = self._services_found.get(name) # # If we've got a match, return it. Otherwise: # * If we're due to wait for ever, continue # * If we're out of time return None # * Otherwise we've still got time left: continue # if discovered: return discovered.address if timed_out(started_at, wait_for_s): return None else: return Continue def do_discover_all(self, started_at): _logger.debug("Discover all") return [(service.name, service.address) for service in self._services_found.values()] def do_reset(self, started_at): _logger.debug("Reset") self.do_pause(started_at) self._services_found.clear() self._services_to_advertise.clear() self.do_resume(started_at) def do_stop(self, started_at): _logger.debug("Stop") self.stop() def listen_for_one_advert(self): events = dict(self.poller.poll(1000 * self.finder_timeout_s)) if self.socket_fd not in events: return message, source = self.socket.recvfrom(self.beacon_message_size) _logger.debug("Broadcast message received: %r", message) service_name, service_address, ttl_s = _unpack(message) service = _Service(service_name, service_address, ttl_s) self._services_found[service_name] = service def broadcast_one_advert(self): if self._services_to_advertise: next_service = self._services_to_advertise[0] if next_service.advertise_at < time.time(): _logger.debug("%s due to advertise at %s", next_service.name, time.ctime(next_service.advertise_at)) message = _pack([next_service.name, next_service.address, next_service.ttl_s]) for broadcast_address in self.broadcast_addresses: _logger.debug("Advertising on %s", broadcast_address) self.socket.sendto(message, 0, (broadcast_address, self.beacon_port)) next_service.advertise_at = time.time() + self.time_between_broadcasts_s self._services_to_advertise.rotate(-1) def remove_expired_adverts(self): for name, service in list(self._services_found.items()): # # A service with an empty expiry time never expired # if service.expires_at is None: continue if service.expires_at <= time.time(): _logger.warn("Removing advert for %s which expired at %s", name, time.ctime(service.expires_at)) del self._services_found[name] def poll_command_request(self): """If the command RPC socket has an incoming request, separate it into its action and its params and put it on the command request queue. """ try: message = self.rpc.recv(zmq.NOBLOCK) except zmq.ZMQError as exc: if exc.errno == zmq.EAGAIN: return else: raise _logger.debug("Received command %s", message) segments = _unpack(message) action, params = segments[0], segments[1:] _logger.debug("Adding %s, %s to the request queue", action, params) self._command = _Command(action, params) def process_command(self): if not self._command: return else: _logger.debug("process_command: %s", self._command.action) command = self._command _logger.debug("Picked %s, %s, %s", self._command.action, self._command.params, self._command.started_at) function = getattr(self, "do_" + command.action.lower(), None) if not function: raise NotImplementedError("%s is not a valid action") else: try: result = function(command.started_at, *command.params) except: _logger.exception("Problem calling %s with %s", command.action, command.params) result = None _logger.debug("result = %s", result) # # result will be Continue if the action cannot be completed # (eg a discovery) but its time is not yet expired. Leave # the command on the stack for now. # if result is Continue: return # # If we get a result, add the result to the response # queue and pop the request off the stack. # self._command.response = result def poll_command_reponse(self): """If the latest request has a response, issue it as a reply to the RPC socket. """ if self._command.response is not Empty: _logger.debug("Sending response %s", self._command.response) self.rpc.send(_pack(self._command.response)) self._command = None def run(self): _logger.info("Starting discovery") while not self._stop_event.wait(0): try: # # If we're not already processing one, check for an command # to advertise/discover from a local process. # if not self._command: self.poll_command_request() # # If we're paused no adverts will be broadcast. Adverts # will be received and stale ones expired # if not self._is_paused: # # Broadcast the first advert whose advertising schedule # has arrived # self.broadcast_one_advert() # # See if an advert broadcast has arrived # self.listen_for_one_advert() # # See if any adverts have expired # self.remove_expired_adverts() # # If we're processing a command, see if it's complete # if self._command: self.process_command() self.poll_command_reponse() except: _logger.exception("Problem in beacon thread") break _logger.info("Ending discovery") self.rpc.close() self.socket.close() _beacon = None _remote_beacon = object() def _start_beacon(port=None): """Start a beacon thread within this process if no beacon is currently running on this machine. In general this is called automatically when an attempt is made to advertise or discover. It might be convenient, though, to call this function directly if you want to have a process whose only job is to host this beacon so that it doesn't shut down when other processes shut down. """ global _beacon if _beacon is None: _logger.debug("About to start beacon with port %s", port) try: _beacon = _Beacon(port) except (OSError, socket.error) as exc: if exc.errno == errno.EADDRINUSE: _logger.warn("Beacon already active on this machine") # # _remote_beacon is simply a not-None sentinel value # to distinguish between the case where we have not # yet started a beacon and where we have found one # in another process. # _beacon = _remote_beacon else: raise else: _beacon.start() def _stop_beacon(): # # Mostly for testing: shutdown the beacon if it's running # locally and clear it globally so the next attempt will # start fresh. # global _beacon if _beacon and _beacon is not _remote_beacon: _beacon.stop() _beacon.join() _beacon = None def _rpc(action, *args, **kwargs): _logger.debug("About to send rpc request %s with args %s, kwargs %s", action, args, kwargs) wait_for_s = kwargs.pop("wait_for_s", 5) with sockets.context.socket(zmq.REQ) as socket: # # To avoid problems when restarting a beacon not long after it's been # closed, force the socket to shut down regardless about 1 second after # it's been closed. # socket.connect("tcp://localhost:%s" % _Beacon.rpc_port) socket.send(_pack([action] + list(args))) reply = sockets._sockets._receive_with_timeout(socket, wait_for_s) return _unpack(reply) def _pause(): return _rpc("pause") def _resume(): return _rpc("resume") _services_advertised = {} def advertise(name, address=None, fail_if_exists=False, ttl_s=config.ADVERT_TTL_S): """Advertise a name at an address Start to advertise service `name` at address `address`. If the address is not supplied, one is constructed and this is returned by the function. ie this is a typical use:: address = nw0.advertise("myservice") :param name: any text :param address: either "ip:port" or None :param fail_if_exists: fail if this name is already registered? :param ttl_s: the advert will persist for this many seconds other beacons :returns: the address given or constructed """ _start_beacon() address = _rpc("advertise", name, address, fail_if_exists, ttl_s) _services_advertised[name] = address return address def _unadvertise_all(): """Remove all adverts """ for name in _services_advertised: try: _unadvertise(name) except core.SocketTimedOutError: _logger.warn("Timed out trying to unadvertise") break atexit.register(_unadvertise_all) def _unadvertise(name): """Remove the advert for a name This is intended for internal use only at the moment. When a process exits it can remove adverts for its services from the beacon running on that machine. (Of course, if the beacon thread is part of of the same service, all its adverts will cease). """ _start_beacon() return _rpc("unadvertise", name) def discover(name, wait_for_s=60): """Discover a service by name Look for an advert to a named service:: address = nw0.discover("myservice") :param name: any text :param wait_for_s: how many seconds to wait before giving up :returns: the address found or None """ _start_beacon() # # It's possible to enter a deadlock situation where the first # process fires off a discovery request and waits for the # second process to advertise. But the second process has to # connect to the rpc port of the first process' beacon and # its advertisement is queued behind the pending discovery. # # To give both a chance of succeeding we operate in bursts, # allowing them to interleave. # t0 = time.time() while True: discovery = _rpc("discover", name, 0.5) if discovery: return discovery if timed_out(t0, wait_for_s): return None def discover_all(): """Produce a list of all known services and their addresses Ask for all known services as a list of 2-tuples: (name, address) This could, eg, be used to form a dictionary of services:: services = dict(nw0.discover_all()) :returns: a list of 2-tuples [(name, address), ...] """ _start_beacon() return _rpc("discover_all") def discover_group(group, separator="/", exclude=None): """Produce a list of all services and their addresses in a group A group is an optional form of namespace within the discovery mechanism. If an advertised name has the form <group><sep><name> it is deemed to belong to <group>. Note that the service's name is still the full string <group><sep><name>. The group concept is simply for discovery and to assist differentiation, eg, in a classroom group. :param group: the name of a group prefix :param separator: the separator character [/] :param exclude: an iterable of names to exclude (or None) :returns: a list of 2-tuples [(name, address), ...] """ _start_beacon() if exclude is None: names_to_exclude = set() else: names_to_exclude = set(exclude) all_discovered = _rpc("discover_all") return [(name, address) for (name, address) in all_discovered if name.startswith("%s%s" % (group, separator)) and name not in names_to_exclude ] def reset_beacon(): """Clear the adverts which the beacon is carrying (This is mostly useful when testing, to get a fresh start) """ _start_beacon() return _rpc("reset") if __name__ == '__main__': params = [arg.lower() for arg in sys.argv] if "--debug" in params: logging_level = logging.DEBUG else: logging_level = logging.INFO handler = logging.StreamHandler() handler.setLevel(logging_level) handler.setFormatter(logging.Formatter("%(levelname)s - %(message)s")) _logger.addHandler(handler) _start_beacon() _logger.info("Beacon started at %s", time.asctime()) while True: time.sleep(1)
tjguk/networkzero
networkzero/discovery.py
discover
python
def discover(name, wait_for_s=60): _start_beacon() # # It's possible to enter a deadlock situation where the first # process fires off a discovery request and waits for the # second process to advertise. But the second process has to # connect to the rpc port of the first process' beacon and # its advertisement is queued behind the pending discovery. # # To give both a chance of succeeding we operate in bursts, # allowing them to interleave. # t0 = time.time() while True: discovery = _rpc("discover", name, 0.5) if discovery: return discovery if timed_out(t0, wait_for_s): return None
Discover a service by name Look for an advert to a named service:: address = nw0.discover("myservice") :param name: any text :param wait_for_s: how many seconds to wait before giving up :returns: the address found or None
train
https://github.com/tjguk/networkzero/blob/0e3e81d2e9200b25a83ac07741612283599486d7/networkzero/discovery.py#L539-L567
[ "def timed_out(started_at, wait_for_s):\n #\n # If the wait time is the sentinel value FOREVER, never time out\n # Otherwise time out if the current time is more than wait_for_s seconds after the start time\n #\n if wait_for_s is config.FOREVER:\n return False\n else:\n return time.t...
# -*- coding: utf-8 -*- """Advertise and collect advertisements of network services The discovery module offers: * A UDP broadcast socket which: - Listens for and keeps track of service adverts from this and other machines & processes - Broadcasts services advertised by this process * A ZeroMQ socket which allow any process on this machine to communicate with its broadcast socket In other words, we have a beacon which listens to instructions from processes on this machine while sending out and listening to adverts broadcast to/from all machines on the network. The beacon is started automatically in a daemon thread when the first attempt is made to advertise or discover. If another process already has a beacon running (ie if this beacon can't bind to its port) this beacon thread will shut down with no further action. The module-level functions to advertise and discover will open a connection to a ZeroMQ socket on this machine (which might be hosted by this or by another process) and will use this socket to send commands to the beacon thread which will update or return its internal list of advertised services. As an additional convenience, the :func:`advertise` function will, if given no specific address, generate a suitable ip:port pair by interrogating the system. This functionality is actually in :func:`networkzero.address` (qv). """ from __future__ import print_function import os, sys import atexit import collections import errno import json import logging import socket import threading import time import zmq from . import config from . import core from . import sockets _logger = core.get_logger(__name__) # # Continue is a sentinel value to indicate that a command # has completed its scheduled slice without producing a result # and without exceeding its overall timeout. # Continue = object() # # Empty is a sentinel to distinguish between no result and a result of None # Empty = object() def _unpack(message): return json.loads(message.decode(config.ENCODING)) def _pack(message): return json.dumps(message).encode(config.ENCODING) def timed_out(started_at, wait_for_s): # # If the wait time is the sentinel value FOREVER, never time out # Otherwise time out if the current time is more than wait_for_s seconds after the start time # if wait_for_s is config.FOREVER: return False else: return time.time() > started_at + wait_for_s def _bind_with_timeout(bind_function, args, n_tries=3, retry_interval_s=0.5): """Attempt to bind a socket a number of times with a short interval in between Especially on Linux, crashing out of a networkzero process can leave the sockets lingering and unable to re-bind on startup. We give it a few goes here to see if we can bind within a couple of seconds. """ n_tries_left = n_tries while n_tries_left > 0: try: return bind_function(*args) except zmq.error.ZMQError as exc: _logger.warn("%s; %d tries remaining", exc, n_tries_left) n_tries_left -= 1 except OSError as exc: if exc.errno == errno.EADDRINUSE: _logger.warn("%s; %d tries remaining", exc, n_tries_left) n_tries_left -= 1 else: raise else: raise core.SocketAlreadyExistsError("Failed to bind after %s tries" % n_tries) class _Service(object): """Convenience container with details of a service to be advertised Includes the name, address and when it is next due to be advertised and when it is due to expire if it was discovered. """ def __init__(self, name, address, ttl_s=None): self.name = name self.address = address self.ttl_s = ttl_s self.expires_at = None if ttl_s is None else (time.time() + ttl_s) self.advertise_at = 0 def __str__(self): return "_Service %s at %s due to advertise at %s and expire at %s" % ( self.name, self.address, time.ctime(self.advertise_at), time.ctime(self.expires_at) ) def __repr__(self): return "<%s: %s>" % (self.__class__.__name__, str(self)) def __eq__(self, other): return self.name == other.name class _Command(object): """Convenience container with details of a running command Includes the action ("discover", "advertise" etc.), its parameters, when it was started -- for timeout purposes -- and any response. This is used by the process_command functionality """ def __init__(self, action, params): self.action = action self.params = params self.started_at = time.time() self.response = Empty def __str__(self): return "_Command: %s (%s) started at %s -> %s" % (self.action, self.params, time.ctime(self.started_at), self.response) class _Beacon(threading.Thread): """Threaded beacon to: listen for adverts & broadcast adverts """ rpc_port = 9998 beacon_port = 9999 finder_timeout_s = 0.05 beacon_message_size = 256 time_between_broadcasts_s = config.BEACON_ADVERT_FREQUENCY_S def __init__(self, beacon_port=None): threading.Thread.__init__(self) self.setDaemon(True) self._stop_event = threading.Event() self._is_paused = False self.beacon_port = beacon_port or self.__class__.beacon_port _logger.debug("Using beacon port %s", self.beacon_port) # # Services we're advertising # self._services_to_advertise = collections.deque() # # Broadcast adverts which we've received (some of which will be our own) # self._services_found = {} # # _Command requests are collected on one queue # _Command responses are added to another # self._command = None # # Set the socket up to broadcast datagrams over UDP # self.broadcast_addresses = set(core._find_ip4_broadcast_addresses()) self.socket = socket.socket(socket.AF_INET, socket.SOCK_DGRAM, socket.IPPROTO_UDP) self.socket.setsockopt(socket.SOL_SOCKET, socket.SO_BROADCAST, 1) self.socket.bind(("", self.beacon_port)) # # Add the raw UDP socket to a ZeroMQ socket poller so we can check whether # it's received anything as part of the beacon's main event loop. # self.socket_fd = self.socket.fileno() self.poller = zmq.Poller() self.poller.register(self.socket, zmq.POLLIN) self.rpc = sockets.context.socket(zmq.REP) # # To avoid problems when restarting a beacon not long after it's been # closed, force the socket to shut down regardless about 1 second after # it's been closed. # self.rpc.linger = 1000 _bind_with_timeout(self.rpc.bind, ("tcp://127.0.0.1:%s" % self.rpc_port,)) def stop(self): _logger.debug("About to stop") self._stop_event.set() # # Commands available via RPC are methods whose name starts with "do_" # def do_advertise(self, started_at, name, address, fail_if_exists, ttl_s): _logger.debug("Advertise %s on %s %s TTL=%s", name, address, fail_if_exists, ttl_s) canonical_address = core.address(address) for service in self._services_to_advertise: if service.name == name: if fail_if_exists: _logger.error("_Service %s already exists on %s", name, service.address) return None else: _logger.warn("Superseding service %s which already exists on %s", name, service.address) service = _Service(name, canonical_address, ttl_s) self._services_to_advertise.append(service) # # As a shortcut, automatically "discover" any services we ourselves are advertising # self._services_found[name] = service return canonical_address def do_unadvertise(self, started_at, name): _logger.debug("Unadvertise %s", name) for service in self._services_to_advertise: if service.name == name: self._services_to_advertise.remove(service) break else: _logger.warn("No advert found for %s", name) _logger.debug("Services now: %s", self._services_to_advertise) def do_pause(self, started_at): _logger.debug("Pause") self._is_paused = True def do_resume(self, started_at): _logger.debug("Resume") self._is_paused = False def do_discover(self, started_at, name, wait_for_s): _logger.debug("Discover %s waiting for %s seconds", name, wait_for_s) discovered = self._services_found.get(name) # # If we've got a match, return it. Otherwise: # * If we're due to wait for ever, continue # * If we're out of time return None # * Otherwise we've still got time left: continue # if discovered: return discovered.address if timed_out(started_at, wait_for_s): return None else: return Continue def do_discover_all(self, started_at): _logger.debug("Discover all") return [(service.name, service.address) for service in self._services_found.values()] def do_reset(self, started_at): _logger.debug("Reset") self.do_pause(started_at) self._services_found.clear() self._services_to_advertise.clear() self.do_resume(started_at) def do_stop(self, started_at): _logger.debug("Stop") self.stop() def listen_for_one_advert(self): events = dict(self.poller.poll(1000 * self.finder_timeout_s)) if self.socket_fd not in events: return message, source = self.socket.recvfrom(self.beacon_message_size) _logger.debug("Broadcast message received: %r", message) service_name, service_address, ttl_s = _unpack(message) service = _Service(service_name, service_address, ttl_s) self._services_found[service_name] = service def broadcast_one_advert(self): if self._services_to_advertise: next_service = self._services_to_advertise[0] if next_service.advertise_at < time.time(): _logger.debug("%s due to advertise at %s", next_service.name, time.ctime(next_service.advertise_at)) message = _pack([next_service.name, next_service.address, next_service.ttl_s]) for broadcast_address in self.broadcast_addresses: _logger.debug("Advertising on %s", broadcast_address) self.socket.sendto(message, 0, (broadcast_address, self.beacon_port)) next_service.advertise_at = time.time() + self.time_between_broadcasts_s self._services_to_advertise.rotate(-1) def remove_expired_adverts(self): for name, service in list(self._services_found.items()): # # A service with an empty expiry time never expired # if service.expires_at is None: continue if service.expires_at <= time.time(): _logger.warn("Removing advert for %s which expired at %s", name, time.ctime(service.expires_at)) del self._services_found[name] def poll_command_request(self): """If the command RPC socket has an incoming request, separate it into its action and its params and put it on the command request queue. """ try: message = self.rpc.recv(zmq.NOBLOCK) except zmq.ZMQError as exc: if exc.errno == zmq.EAGAIN: return else: raise _logger.debug("Received command %s", message) segments = _unpack(message) action, params = segments[0], segments[1:] _logger.debug("Adding %s, %s to the request queue", action, params) self._command = _Command(action, params) def process_command(self): if not self._command: return else: _logger.debug("process_command: %s", self._command.action) command = self._command _logger.debug("Picked %s, %s, %s", self._command.action, self._command.params, self._command.started_at) function = getattr(self, "do_" + command.action.lower(), None) if not function: raise NotImplementedError("%s is not a valid action") else: try: result = function(command.started_at, *command.params) except: _logger.exception("Problem calling %s with %s", command.action, command.params) result = None _logger.debug("result = %s", result) # # result will be Continue if the action cannot be completed # (eg a discovery) but its time is not yet expired. Leave # the command on the stack for now. # if result is Continue: return # # If we get a result, add the result to the response # queue and pop the request off the stack. # self._command.response = result def poll_command_reponse(self): """If the latest request has a response, issue it as a reply to the RPC socket. """ if self._command.response is not Empty: _logger.debug("Sending response %s", self._command.response) self.rpc.send(_pack(self._command.response)) self._command = None def run(self): _logger.info("Starting discovery") while not self._stop_event.wait(0): try: # # If we're not already processing one, check for an command # to advertise/discover from a local process. # if not self._command: self.poll_command_request() # # If we're paused no adverts will be broadcast. Adverts # will be received and stale ones expired # if not self._is_paused: # # Broadcast the first advert whose advertising schedule # has arrived # self.broadcast_one_advert() # # See if an advert broadcast has arrived # self.listen_for_one_advert() # # See if any adverts have expired # self.remove_expired_adverts() # # If we're processing a command, see if it's complete # if self._command: self.process_command() self.poll_command_reponse() except: _logger.exception("Problem in beacon thread") break _logger.info("Ending discovery") self.rpc.close() self.socket.close() _beacon = None _remote_beacon = object() def _start_beacon(port=None): """Start a beacon thread within this process if no beacon is currently running on this machine. In general this is called automatically when an attempt is made to advertise or discover. It might be convenient, though, to call this function directly if you want to have a process whose only job is to host this beacon so that it doesn't shut down when other processes shut down. """ global _beacon if _beacon is None: _logger.debug("About to start beacon with port %s", port) try: _beacon = _Beacon(port) except (OSError, socket.error) as exc: if exc.errno == errno.EADDRINUSE: _logger.warn("Beacon already active on this machine") # # _remote_beacon is simply a not-None sentinel value # to distinguish between the case where we have not # yet started a beacon and where we have found one # in another process. # _beacon = _remote_beacon else: raise else: _beacon.start() def _stop_beacon(): # # Mostly for testing: shutdown the beacon if it's running # locally and clear it globally so the next attempt will # start fresh. # global _beacon if _beacon and _beacon is not _remote_beacon: _beacon.stop() _beacon.join() _beacon = None def _rpc(action, *args, **kwargs): _logger.debug("About to send rpc request %s with args %s, kwargs %s", action, args, kwargs) wait_for_s = kwargs.pop("wait_for_s", 5) with sockets.context.socket(zmq.REQ) as socket: # # To avoid problems when restarting a beacon not long after it's been # closed, force the socket to shut down regardless about 1 second after # it's been closed. # socket.connect("tcp://localhost:%s" % _Beacon.rpc_port) socket.send(_pack([action] + list(args))) reply = sockets._sockets._receive_with_timeout(socket, wait_for_s) return _unpack(reply) def _pause(): return _rpc("pause") def _resume(): return _rpc("resume") _services_advertised = {} def advertise(name, address=None, fail_if_exists=False, ttl_s=config.ADVERT_TTL_S): """Advertise a name at an address Start to advertise service `name` at address `address`. If the address is not supplied, one is constructed and this is returned by the function. ie this is a typical use:: address = nw0.advertise("myservice") :param name: any text :param address: either "ip:port" or None :param fail_if_exists: fail if this name is already registered? :param ttl_s: the advert will persist for this many seconds other beacons :returns: the address given or constructed """ _start_beacon() address = _rpc("advertise", name, address, fail_if_exists, ttl_s) _services_advertised[name] = address return address def _unadvertise_all(): """Remove all adverts """ for name in _services_advertised: try: _unadvertise(name) except core.SocketTimedOutError: _logger.warn("Timed out trying to unadvertise") break atexit.register(_unadvertise_all) def _unadvertise(name): """Remove the advert for a name This is intended for internal use only at the moment. When a process exits it can remove adverts for its services from the beacon running on that machine. (Of course, if the beacon thread is part of of the same service, all its adverts will cease). """ _start_beacon() return _rpc("unadvertise", name) def discover(name, wait_for_s=60): """Discover a service by name Look for an advert to a named service:: address = nw0.discover("myservice") :param name: any text :param wait_for_s: how many seconds to wait before giving up :returns: the address found or None """ _start_beacon() # # It's possible to enter a deadlock situation where the first # process fires off a discovery request and waits for the # second process to advertise. But the second process has to # connect to the rpc port of the first process' beacon and # its advertisement is queued behind the pending discovery. # # To give both a chance of succeeding we operate in bursts, # allowing them to interleave. # t0 = time.time() while True: discovery = _rpc("discover", name, 0.5) if discovery: return discovery if timed_out(t0, wait_for_s): return None def discover_all(): """Produce a list of all known services and their addresses Ask for all known services as a list of 2-tuples: (name, address) This could, eg, be used to form a dictionary of services:: services = dict(nw0.discover_all()) :returns: a list of 2-tuples [(name, address), ...] """ _start_beacon() return _rpc("discover_all") def discover_group(group, separator="/", exclude=None): """Produce a list of all services and their addresses in a group A group is an optional form of namespace within the discovery mechanism. If an advertised name has the form <group><sep><name> it is deemed to belong to <group>. Note that the service's name is still the full string <group><sep><name>. The group concept is simply for discovery and to assist differentiation, eg, in a classroom group. :param group: the name of a group prefix :param separator: the separator character [/] :param exclude: an iterable of names to exclude (or None) :returns: a list of 2-tuples [(name, address), ...] """ _start_beacon() if exclude is None: names_to_exclude = set() else: names_to_exclude = set(exclude) all_discovered = _rpc("discover_all") return [(name, address) for (name, address) in all_discovered if name.startswith("%s%s" % (group, separator)) and name not in names_to_exclude ] def reset_beacon(): """Clear the adverts which the beacon is carrying (This is mostly useful when testing, to get a fresh start) """ _start_beacon() return _rpc("reset") if __name__ == '__main__': params = [arg.lower() for arg in sys.argv] if "--debug" in params: logging_level = logging.DEBUG else: logging_level = logging.INFO handler = logging.StreamHandler() handler.setLevel(logging_level) handler.setFormatter(logging.Formatter("%(levelname)s - %(message)s")) _logger.addHandler(handler) _start_beacon() _logger.info("Beacon started at %s", time.asctime()) while True: time.sleep(1)
tjguk/networkzero
networkzero/discovery.py
discover_group
python
def discover_group(group, separator="/", exclude=None): _start_beacon() if exclude is None: names_to_exclude = set() else: names_to_exclude = set(exclude) all_discovered = _rpc("discover_all") return [(name, address) for (name, address) in all_discovered if name.startswith("%s%s" % (group, separator)) and name not in names_to_exclude ]
Produce a list of all services and their addresses in a group A group is an optional form of namespace within the discovery mechanism. If an advertised name has the form <group><sep><name> it is deemed to belong to <group>. Note that the service's name is still the full string <group><sep><name>. The group concept is simply for discovery and to assist differentiation, eg, in a classroom group. :param group: the name of a group prefix :param separator: the separator character [/] :param exclude: an iterable of names to exclude (or None) :returns: a list of 2-tuples [(name, address), ...]
train
https://github.com/tjguk/networkzero/blob/0e3e81d2e9200b25a83ac07741612283599486d7/networkzero/discovery.py#L582-L607
[ "def _start_beacon(port=None):\n \"\"\"Start a beacon thread within this process if no beacon is currently\n running on this machine.\n\n In general this is called automatically when an attempt is made to\n advertise or discover. It might be convenient, though, to call this\n function directly if you...
# -*- coding: utf-8 -*- """Advertise and collect advertisements of network services The discovery module offers: * A UDP broadcast socket which: - Listens for and keeps track of service adverts from this and other machines & processes - Broadcasts services advertised by this process * A ZeroMQ socket which allow any process on this machine to communicate with its broadcast socket In other words, we have a beacon which listens to instructions from processes on this machine while sending out and listening to adverts broadcast to/from all machines on the network. The beacon is started automatically in a daemon thread when the first attempt is made to advertise or discover. If another process already has a beacon running (ie if this beacon can't bind to its port) this beacon thread will shut down with no further action. The module-level functions to advertise and discover will open a connection to a ZeroMQ socket on this machine (which might be hosted by this or by another process) and will use this socket to send commands to the beacon thread which will update or return its internal list of advertised services. As an additional convenience, the :func:`advertise` function will, if given no specific address, generate a suitable ip:port pair by interrogating the system. This functionality is actually in :func:`networkzero.address` (qv). """ from __future__ import print_function import os, sys import atexit import collections import errno import json import logging import socket import threading import time import zmq from . import config from . import core from . import sockets _logger = core.get_logger(__name__) # # Continue is a sentinel value to indicate that a command # has completed its scheduled slice without producing a result # and without exceeding its overall timeout. # Continue = object() # # Empty is a sentinel to distinguish between no result and a result of None # Empty = object() def _unpack(message): return json.loads(message.decode(config.ENCODING)) def _pack(message): return json.dumps(message).encode(config.ENCODING) def timed_out(started_at, wait_for_s): # # If the wait time is the sentinel value FOREVER, never time out # Otherwise time out if the current time is more than wait_for_s seconds after the start time # if wait_for_s is config.FOREVER: return False else: return time.time() > started_at + wait_for_s def _bind_with_timeout(bind_function, args, n_tries=3, retry_interval_s=0.5): """Attempt to bind a socket a number of times with a short interval in between Especially on Linux, crashing out of a networkzero process can leave the sockets lingering and unable to re-bind on startup. We give it a few goes here to see if we can bind within a couple of seconds. """ n_tries_left = n_tries while n_tries_left > 0: try: return bind_function(*args) except zmq.error.ZMQError as exc: _logger.warn("%s; %d tries remaining", exc, n_tries_left) n_tries_left -= 1 except OSError as exc: if exc.errno == errno.EADDRINUSE: _logger.warn("%s; %d tries remaining", exc, n_tries_left) n_tries_left -= 1 else: raise else: raise core.SocketAlreadyExistsError("Failed to bind after %s tries" % n_tries) class _Service(object): """Convenience container with details of a service to be advertised Includes the name, address and when it is next due to be advertised and when it is due to expire if it was discovered. """ def __init__(self, name, address, ttl_s=None): self.name = name self.address = address self.ttl_s = ttl_s self.expires_at = None if ttl_s is None else (time.time() + ttl_s) self.advertise_at = 0 def __str__(self): return "_Service %s at %s due to advertise at %s and expire at %s" % ( self.name, self.address, time.ctime(self.advertise_at), time.ctime(self.expires_at) ) def __repr__(self): return "<%s: %s>" % (self.__class__.__name__, str(self)) def __eq__(self, other): return self.name == other.name class _Command(object): """Convenience container with details of a running command Includes the action ("discover", "advertise" etc.), its parameters, when it was started -- for timeout purposes -- and any response. This is used by the process_command functionality """ def __init__(self, action, params): self.action = action self.params = params self.started_at = time.time() self.response = Empty def __str__(self): return "_Command: %s (%s) started at %s -> %s" % (self.action, self.params, time.ctime(self.started_at), self.response) class _Beacon(threading.Thread): """Threaded beacon to: listen for adverts & broadcast adverts """ rpc_port = 9998 beacon_port = 9999 finder_timeout_s = 0.05 beacon_message_size = 256 time_between_broadcasts_s = config.BEACON_ADVERT_FREQUENCY_S def __init__(self, beacon_port=None): threading.Thread.__init__(self) self.setDaemon(True) self._stop_event = threading.Event() self._is_paused = False self.beacon_port = beacon_port or self.__class__.beacon_port _logger.debug("Using beacon port %s", self.beacon_port) # # Services we're advertising # self._services_to_advertise = collections.deque() # # Broadcast adverts which we've received (some of which will be our own) # self._services_found = {} # # _Command requests are collected on one queue # _Command responses are added to another # self._command = None # # Set the socket up to broadcast datagrams over UDP # self.broadcast_addresses = set(core._find_ip4_broadcast_addresses()) self.socket = socket.socket(socket.AF_INET, socket.SOCK_DGRAM, socket.IPPROTO_UDP) self.socket.setsockopt(socket.SOL_SOCKET, socket.SO_BROADCAST, 1) self.socket.bind(("", self.beacon_port)) # # Add the raw UDP socket to a ZeroMQ socket poller so we can check whether # it's received anything as part of the beacon's main event loop. # self.socket_fd = self.socket.fileno() self.poller = zmq.Poller() self.poller.register(self.socket, zmq.POLLIN) self.rpc = sockets.context.socket(zmq.REP) # # To avoid problems when restarting a beacon not long after it's been # closed, force the socket to shut down regardless about 1 second after # it's been closed. # self.rpc.linger = 1000 _bind_with_timeout(self.rpc.bind, ("tcp://127.0.0.1:%s" % self.rpc_port,)) def stop(self): _logger.debug("About to stop") self._stop_event.set() # # Commands available via RPC are methods whose name starts with "do_" # def do_advertise(self, started_at, name, address, fail_if_exists, ttl_s): _logger.debug("Advertise %s on %s %s TTL=%s", name, address, fail_if_exists, ttl_s) canonical_address = core.address(address) for service in self._services_to_advertise: if service.name == name: if fail_if_exists: _logger.error("_Service %s already exists on %s", name, service.address) return None else: _logger.warn("Superseding service %s which already exists on %s", name, service.address) service = _Service(name, canonical_address, ttl_s) self._services_to_advertise.append(service) # # As a shortcut, automatically "discover" any services we ourselves are advertising # self._services_found[name] = service return canonical_address def do_unadvertise(self, started_at, name): _logger.debug("Unadvertise %s", name) for service in self._services_to_advertise: if service.name == name: self._services_to_advertise.remove(service) break else: _logger.warn("No advert found for %s", name) _logger.debug("Services now: %s", self._services_to_advertise) def do_pause(self, started_at): _logger.debug("Pause") self._is_paused = True def do_resume(self, started_at): _logger.debug("Resume") self._is_paused = False def do_discover(self, started_at, name, wait_for_s): _logger.debug("Discover %s waiting for %s seconds", name, wait_for_s) discovered = self._services_found.get(name) # # If we've got a match, return it. Otherwise: # * If we're due to wait for ever, continue # * If we're out of time return None # * Otherwise we've still got time left: continue # if discovered: return discovered.address if timed_out(started_at, wait_for_s): return None else: return Continue def do_discover_all(self, started_at): _logger.debug("Discover all") return [(service.name, service.address) for service in self._services_found.values()] def do_reset(self, started_at): _logger.debug("Reset") self.do_pause(started_at) self._services_found.clear() self._services_to_advertise.clear() self.do_resume(started_at) def do_stop(self, started_at): _logger.debug("Stop") self.stop() def listen_for_one_advert(self): events = dict(self.poller.poll(1000 * self.finder_timeout_s)) if self.socket_fd not in events: return message, source = self.socket.recvfrom(self.beacon_message_size) _logger.debug("Broadcast message received: %r", message) service_name, service_address, ttl_s = _unpack(message) service = _Service(service_name, service_address, ttl_s) self._services_found[service_name] = service def broadcast_one_advert(self): if self._services_to_advertise: next_service = self._services_to_advertise[0] if next_service.advertise_at < time.time(): _logger.debug("%s due to advertise at %s", next_service.name, time.ctime(next_service.advertise_at)) message = _pack([next_service.name, next_service.address, next_service.ttl_s]) for broadcast_address in self.broadcast_addresses: _logger.debug("Advertising on %s", broadcast_address) self.socket.sendto(message, 0, (broadcast_address, self.beacon_port)) next_service.advertise_at = time.time() + self.time_between_broadcasts_s self._services_to_advertise.rotate(-1) def remove_expired_adverts(self): for name, service in list(self._services_found.items()): # # A service with an empty expiry time never expired # if service.expires_at is None: continue if service.expires_at <= time.time(): _logger.warn("Removing advert for %s which expired at %s", name, time.ctime(service.expires_at)) del self._services_found[name] def poll_command_request(self): """If the command RPC socket has an incoming request, separate it into its action and its params and put it on the command request queue. """ try: message = self.rpc.recv(zmq.NOBLOCK) except zmq.ZMQError as exc: if exc.errno == zmq.EAGAIN: return else: raise _logger.debug("Received command %s", message) segments = _unpack(message) action, params = segments[0], segments[1:] _logger.debug("Adding %s, %s to the request queue", action, params) self._command = _Command(action, params) def process_command(self): if not self._command: return else: _logger.debug("process_command: %s", self._command.action) command = self._command _logger.debug("Picked %s, %s, %s", self._command.action, self._command.params, self._command.started_at) function = getattr(self, "do_" + command.action.lower(), None) if not function: raise NotImplementedError("%s is not a valid action") else: try: result = function(command.started_at, *command.params) except: _logger.exception("Problem calling %s with %s", command.action, command.params) result = None _logger.debug("result = %s", result) # # result will be Continue if the action cannot be completed # (eg a discovery) but its time is not yet expired. Leave # the command on the stack for now. # if result is Continue: return # # If we get a result, add the result to the response # queue and pop the request off the stack. # self._command.response = result def poll_command_reponse(self): """If the latest request has a response, issue it as a reply to the RPC socket. """ if self._command.response is not Empty: _logger.debug("Sending response %s", self._command.response) self.rpc.send(_pack(self._command.response)) self._command = None def run(self): _logger.info("Starting discovery") while not self._stop_event.wait(0): try: # # If we're not already processing one, check for an command # to advertise/discover from a local process. # if not self._command: self.poll_command_request() # # If we're paused no adverts will be broadcast. Adverts # will be received and stale ones expired # if not self._is_paused: # # Broadcast the first advert whose advertising schedule # has arrived # self.broadcast_one_advert() # # See if an advert broadcast has arrived # self.listen_for_one_advert() # # See if any adverts have expired # self.remove_expired_adverts() # # If we're processing a command, see if it's complete # if self._command: self.process_command() self.poll_command_reponse() except: _logger.exception("Problem in beacon thread") break _logger.info("Ending discovery") self.rpc.close() self.socket.close() _beacon = None _remote_beacon = object() def _start_beacon(port=None): """Start a beacon thread within this process if no beacon is currently running on this machine. In general this is called automatically when an attempt is made to advertise or discover. It might be convenient, though, to call this function directly if you want to have a process whose only job is to host this beacon so that it doesn't shut down when other processes shut down. """ global _beacon if _beacon is None: _logger.debug("About to start beacon with port %s", port) try: _beacon = _Beacon(port) except (OSError, socket.error) as exc: if exc.errno == errno.EADDRINUSE: _logger.warn("Beacon already active on this machine") # # _remote_beacon is simply a not-None sentinel value # to distinguish between the case where we have not # yet started a beacon and where we have found one # in another process. # _beacon = _remote_beacon else: raise else: _beacon.start() def _stop_beacon(): # # Mostly for testing: shutdown the beacon if it's running # locally and clear it globally so the next attempt will # start fresh. # global _beacon if _beacon and _beacon is not _remote_beacon: _beacon.stop() _beacon.join() _beacon = None def _rpc(action, *args, **kwargs): _logger.debug("About to send rpc request %s with args %s, kwargs %s", action, args, kwargs) wait_for_s = kwargs.pop("wait_for_s", 5) with sockets.context.socket(zmq.REQ) as socket: # # To avoid problems when restarting a beacon not long after it's been # closed, force the socket to shut down regardless about 1 second after # it's been closed. # socket.connect("tcp://localhost:%s" % _Beacon.rpc_port) socket.send(_pack([action] + list(args))) reply = sockets._sockets._receive_with_timeout(socket, wait_for_s) return _unpack(reply) def _pause(): return _rpc("pause") def _resume(): return _rpc("resume") _services_advertised = {} def advertise(name, address=None, fail_if_exists=False, ttl_s=config.ADVERT_TTL_S): """Advertise a name at an address Start to advertise service `name` at address `address`. If the address is not supplied, one is constructed and this is returned by the function. ie this is a typical use:: address = nw0.advertise("myservice") :param name: any text :param address: either "ip:port" or None :param fail_if_exists: fail if this name is already registered? :param ttl_s: the advert will persist for this many seconds other beacons :returns: the address given or constructed """ _start_beacon() address = _rpc("advertise", name, address, fail_if_exists, ttl_s) _services_advertised[name] = address return address def _unadvertise_all(): """Remove all adverts """ for name in _services_advertised: try: _unadvertise(name) except core.SocketTimedOutError: _logger.warn("Timed out trying to unadvertise") break atexit.register(_unadvertise_all) def _unadvertise(name): """Remove the advert for a name This is intended for internal use only at the moment. When a process exits it can remove adverts for its services from the beacon running on that machine. (Of course, if the beacon thread is part of of the same service, all its adverts will cease). """ _start_beacon() return _rpc("unadvertise", name) def discover(name, wait_for_s=60): """Discover a service by name Look for an advert to a named service:: address = nw0.discover("myservice") :param name: any text :param wait_for_s: how many seconds to wait before giving up :returns: the address found or None """ _start_beacon() # # It's possible to enter a deadlock situation where the first # process fires off a discovery request and waits for the # second process to advertise. But the second process has to # connect to the rpc port of the first process' beacon and # its advertisement is queued behind the pending discovery. # # To give both a chance of succeeding we operate in bursts, # allowing them to interleave. # t0 = time.time() while True: discovery = _rpc("discover", name, 0.5) if discovery: return discovery if timed_out(t0, wait_for_s): return None def discover_all(): """Produce a list of all known services and their addresses Ask for all known services as a list of 2-tuples: (name, address) This could, eg, be used to form a dictionary of services:: services = dict(nw0.discover_all()) :returns: a list of 2-tuples [(name, address), ...] """ _start_beacon() return _rpc("discover_all") def discover_group(group, separator="/", exclude=None): """Produce a list of all services and their addresses in a group A group is an optional form of namespace within the discovery mechanism. If an advertised name has the form <group><sep><name> it is deemed to belong to <group>. Note that the service's name is still the full string <group><sep><name>. The group concept is simply for discovery and to assist differentiation, eg, in a classroom group. :param group: the name of a group prefix :param separator: the separator character [/] :param exclude: an iterable of names to exclude (or None) :returns: a list of 2-tuples [(name, address), ...] """ _start_beacon() if exclude is None: names_to_exclude = set() else: names_to_exclude = set(exclude) all_discovered = _rpc("discover_all") return [(name, address) for (name, address) in all_discovered if name.startswith("%s%s" % (group, separator)) and name not in names_to_exclude ] def reset_beacon(): """Clear the adverts which the beacon is carrying (This is mostly useful when testing, to get a fresh start) """ _start_beacon() return _rpc("reset") if __name__ == '__main__': params = [arg.lower() for arg in sys.argv] if "--debug" in params: logging_level = logging.DEBUG else: logging_level = logging.INFO handler = logging.StreamHandler() handler.setLevel(logging_level) handler.setFormatter(logging.Formatter("%(levelname)s - %(message)s")) _logger.addHandler(handler) _start_beacon() _logger.info("Beacon started at %s", time.asctime()) while True: time.sleep(1)
tjguk/networkzero
networkzero/discovery.py
_Beacon.poll_command_request
python
def poll_command_request(self): try: message = self.rpc.recv(zmq.NOBLOCK) except zmq.ZMQError as exc: if exc.errno == zmq.EAGAIN: return else: raise _logger.debug("Received command %s", message) segments = _unpack(message) action, params = segments[0], segments[1:] _logger.debug("Adding %s, %s to the request queue", action, params) self._command = _Command(action, params)
If the command RPC socket has an incoming request, separate it into its action and its params and put it on the command request queue.
train
https://github.com/tjguk/networkzero/blob/0e3e81d2e9200b25a83ac07741612283599486d7/networkzero/discovery.py#L320-L337
[ "def _unpack(message):\n return json.loads(message.decode(config.ENCODING))\n" ]
class _Beacon(threading.Thread): """Threaded beacon to: listen for adverts & broadcast adverts """ rpc_port = 9998 beacon_port = 9999 finder_timeout_s = 0.05 beacon_message_size = 256 time_between_broadcasts_s = config.BEACON_ADVERT_FREQUENCY_S def __init__(self, beacon_port=None): threading.Thread.__init__(self) self.setDaemon(True) self._stop_event = threading.Event() self._is_paused = False self.beacon_port = beacon_port or self.__class__.beacon_port _logger.debug("Using beacon port %s", self.beacon_port) # # Services we're advertising # self._services_to_advertise = collections.deque() # # Broadcast adverts which we've received (some of which will be our own) # self._services_found = {} # # _Command requests are collected on one queue # _Command responses are added to another # self._command = None # # Set the socket up to broadcast datagrams over UDP # self.broadcast_addresses = set(core._find_ip4_broadcast_addresses()) self.socket = socket.socket(socket.AF_INET, socket.SOCK_DGRAM, socket.IPPROTO_UDP) self.socket.setsockopt(socket.SOL_SOCKET, socket.SO_BROADCAST, 1) self.socket.bind(("", self.beacon_port)) # # Add the raw UDP socket to a ZeroMQ socket poller so we can check whether # it's received anything as part of the beacon's main event loop. # self.socket_fd = self.socket.fileno() self.poller = zmq.Poller() self.poller.register(self.socket, zmq.POLLIN) self.rpc = sockets.context.socket(zmq.REP) # # To avoid problems when restarting a beacon not long after it's been # closed, force the socket to shut down regardless about 1 second after # it's been closed. # self.rpc.linger = 1000 _bind_with_timeout(self.rpc.bind, ("tcp://127.0.0.1:%s" % self.rpc_port,)) def stop(self): _logger.debug("About to stop") self._stop_event.set() # # Commands available via RPC are methods whose name starts with "do_" # def do_advertise(self, started_at, name, address, fail_if_exists, ttl_s): _logger.debug("Advertise %s on %s %s TTL=%s", name, address, fail_if_exists, ttl_s) canonical_address = core.address(address) for service in self._services_to_advertise: if service.name == name: if fail_if_exists: _logger.error("_Service %s already exists on %s", name, service.address) return None else: _logger.warn("Superseding service %s which already exists on %s", name, service.address) service = _Service(name, canonical_address, ttl_s) self._services_to_advertise.append(service) # # As a shortcut, automatically "discover" any services we ourselves are advertising # self._services_found[name] = service return canonical_address def do_unadvertise(self, started_at, name): _logger.debug("Unadvertise %s", name) for service in self._services_to_advertise: if service.name == name: self._services_to_advertise.remove(service) break else: _logger.warn("No advert found for %s", name) _logger.debug("Services now: %s", self._services_to_advertise) def do_pause(self, started_at): _logger.debug("Pause") self._is_paused = True def do_resume(self, started_at): _logger.debug("Resume") self._is_paused = False def do_discover(self, started_at, name, wait_for_s): _logger.debug("Discover %s waiting for %s seconds", name, wait_for_s) discovered = self._services_found.get(name) # # If we've got a match, return it. Otherwise: # * If we're due to wait for ever, continue # * If we're out of time return None # * Otherwise we've still got time left: continue # if discovered: return discovered.address if timed_out(started_at, wait_for_s): return None else: return Continue def do_discover_all(self, started_at): _logger.debug("Discover all") return [(service.name, service.address) for service in self._services_found.values()] def do_reset(self, started_at): _logger.debug("Reset") self.do_pause(started_at) self._services_found.clear() self._services_to_advertise.clear() self.do_resume(started_at) def do_stop(self, started_at): _logger.debug("Stop") self.stop() def listen_for_one_advert(self): events = dict(self.poller.poll(1000 * self.finder_timeout_s)) if self.socket_fd not in events: return message, source = self.socket.recvfrom(self.beacon_message_size) _logger.debug("Broadcast message received: %r", message) service_name, service_address, ttl_s = _unpack(message) service = _Service(service_name, service_address, ttl_s) self._services_found[service_name] = service def broadcast_one_advert(self): if self._services_to_advertise: next_service = self._services_to_advertise[0] if next_service.advertise_at < time.time(): _logger.debug("%s due to advertise at %s", next_service.name, time.ctime(next_service.advertise_at)) message = _pack([next_service.name, next_service.address, next_service.ttl_s]) for broadcast_address in self.broadcast_addresses: _logger.debug("Advertising on %s", broadcast_address) self.socket.sendto(message, 0, (broadcast_address, self.beacon_port)) next_service.advertise_at = time.time() + self.time_between_broadcasts_s self._services_to_advertise.rotate(-1) def remove_expired_adverts(self): for name, service in list(self._services_found.items()): # # A service with an empty expiry time never expired # if service.expires_at is None: continue if service.expires_at <= time.time(): _logger.warn("Removing advert for %s which expired at %s", name, time.ctime(service.expires_at)) del self._services_found[name] def process_command(self): if not self._command: return else: _logger.debug("process_command: %s", self._command.action) command = self._command _logger.debug("Picked %s, %s, %s", self._command.action, self._command.params, self._command.started_at) function = getattr(self, "do_" + command.action.lower(), None) if not function: raise NotImplementedError("%s is not a valid action") else: try: result = function(command.started_at, *command.params) except: _logger.exception("Problem calling %s with %s", command.action, command.params) result = None _logger.debug("result = %s", result) # # result will be Continue if the action cannot be completed # (eg a discovery) but its time is not yet expired. Leave # the command on the stack for now. # if result is Continue: return # # If we get a result, add the result to the response # queue and pop the request off the stack. # self._command.response = result def poll_command_reponse(self): """If the latest request has a response, issue it as a reply to the RPC socket. """ if self._command.response is not Empty: _logger.debug("Sending response %s", self._command.response) self.rpc.send(_pack(self._command.response)) self._command = None def run(self): _logger.info("Starting discovery") while not self._stop_event.wait(0): try: # # If we're not already processing one, check for an command # to advertise/discover from a local process. # if not self._command: self.poll_command_request() # # If we're paused no adverts will be broadcast. Adverts # will be received and stale ones expired # if not self._is_paused: # # Broadcast the first advert whose advertising schedule # has arrived # self.broadcast_one_advert() # # See if an advert broadcast has arrived # self.listen_for_one_advert() # # See if any adverts have expired # self.remove_expired_adverts() # # If we're processing a command, see if it's complete # if self._command: self.process_command() self.poll_command_reponse() except: _logger.exception("Problem in beacon thread") break _logger.info("Ending discovery") self.rpc.close() self.socket.close()
tjguk/networkzero
networkzero/discovery.py
_Beacon.poll_command_reponse
python
def poll_command_reponse(self): if self._command.response is not Empty: _logger.debug("Sending response %s", self._command.response) self.rpc.send(_pack(self._command.response)) self._command = None
If the latest request has a response, issue it as a reply to the RPC socket.
train
https://github.com/tjguk/networkzero/blob/0e3e81d2e9200b25a83ac07741612283599486d7/networkzero/discovery.py#L373-L380
[ "def _pack(message):\n return json.dumps(message).encode(config.ENCODING)\n" ]
class _Beacon(threading.Thread): """Threaded beacon to: listen for adverts & broadcast adverts """ rpc_port = 9998 beacon_port = 9999 finder_timeout_s = 0.05 beacon_message_size = 256 time_between_broadcasts_s = config.BEACON_ADVERT_FREQUENCY_S def __init__(self, beacon_port=None): threading.Thread.__init__(self) self.setDaemon(True) self._stop_event = threading.Event() self._is_paused = False self.beacon_port = beacon_port or self.__class__.beacon_port _logger.debug("Using beacon port %s", self.beacon_port) # # Services we're advertising # self._services_to_advertise = collections.deque() # # Broadcast adverts which we've received (some of which will be our own) # self._services_found = {} # # _Command requests are collected on one queue # _Command responses are added to another # self._command = None # # Set the socket up to broadcast datagrams over UDP # self.broadcast_addresses = set(core._find_ip4_broadcast_addresses()) self.socket = socket.socket(socket.AF_INET, socket.SOCK_DGRAM, socket.IPPROTO_UDP) self.socket.setsockopt(socket.SOL_SOCKET, socket.SO_BROADCAST, 1) self.socket.bind(("", self.beacon_port)) # # Add the raw UDP socket to a ZeroMQ socket poller so we can check whether # it's received anything as part of the beacon's main event loop. # self.socket_fd = self.socket.fileno() self.poller = zmq.Poller() self.poller.register(self.socket, zmq.POLLIN) self.rpc = sockets.context.socket(zmq.REP) # # To avoid problems when restarting a beacon not long after it's been # closed, force the socket to shut down regardless about 1 second after # it's been closed. # self.rpc.linger = 1000 _bind_with_timeout(self.rpc.bind, ("tcp://127.0.0.1:%s" % self.rpc_port,)) def stop(self): _logger.debug("About to stop") self._stop_event.set() # # Commands available via RPC are methods whose name starts with "do_" # def do_advertise(self, started_at, name, address, fail_if_exists, ttl_s): _logger.debug("Advertise %s on %s %s TTL=%s", name, address, fail_if_exists, ttl_s) canonical_address = core.address(address) for service in self._services_to_advertise: if service.name == name: if fail_if_exists: _logger.error("_Service %s already exists on %s", name, service.address) return None else: _logger.warn("Superseding service %s which already exists on %s", name, service.address) service = _Service(name, canonical_address, ttl_s) self._services_to_advertise.append(service) # # As a shortcut, automatically "discover" any services we ourselves are advertising # self._services_found[name] = service return canonical_address def do_unadvertise(self, started_at, name): _logger.debug("Unadvertise %s", name) for service in self._services_to_advertise: if service.name == name: self._services_to_advertise.remove(service) break else: _logger.warn("No advert found for %s", name) _logger.debug("Services now: %s", self._services_to_advertise) def do_pause(self, started_at): _logger.debug("Pause") self._is_paused = True def do_resume(self, started_at): _logger.debug("Resume") self._is_paused = False def do_discover(self, started_at, name, wait_for_s): _logger.debug("Discover %s waiting for %s seconds", name, wait_for_s) discovered = self._services_found.get(name) # # If we've got a match, return it. Otherwise: # * If we're due to wait for ever, continue # * If we're out of time return None # * Otherwise we've still got time left: continue # if discovered: return discovered.address if timed_out(started_at, wait_for_s): return None else: return Continue def do_discover_all(self, started_at): _logger.debug("Discover all") return [(service.name, service.address) for service in self._services_found.values()] def do_reset(self, started_at): _logger.debug("Reset") self.do_pause(started_at) self._services_found.clear() self._services_to_advertise.clear() self.do_resume(started_at) def do_stop(self, started_at): _logger.debug("Stop") self.stop() def listen_for_one_advert(self): events = dict(self.poller.poll(1000 * self.finder_timeout_s)) if self.socket_fd not in events: return message, source = self.socket.recvfrom(self.beacon_message_size) _logger.debug("Broadcast message received: %r", message) service_name, service_address, ttl_s = _unpack(message) service = _Service(service_name, service_address, ttl_s) self._services_found[service_name] = service def broadcast_one_advert(self): if self._services_to_advertise: next_service = self._services_to_advertise[0] if next_service.advertise_at < time.time(): _logger.debug("%s due to advertise at %s", next_service.name, time.ctime(next_service.advertise_at)) message = _pack([next_service.name, next_service.address, next_service.ttl_s]) for broadcast_address in self.broadcast_addresses: _logger.debug("Advertising on %s", broadcast_address) self.socket.sendto(message, 0, (broadcast_address, self.beacon_port)) next_service.advertise_at = time.time() + self.time_between_broadcasts_s self._services_to_advertise.rotate(-1) def remove_expired_adverts(self): for name, service in list(self._services_found.items()): # # A service with an empty expiry time never expired # if service.expires_at is None: continue if service.expires_at <= time.time(): _logger.warn("Removing advert for %s which expired at %s", name, time.ctime(service.expires_at)) del self._services_found[name] def poll_command_request(self): """If the command RPC socket has an incoming request, separate it into its action and its params and put it on the command request queue. """ try: message = self.rpc.recv(zmq.NOBLOCK) except zmq.ZMQError as exc: if exc.errno == zmq.EAGAIN: return else: raise _logger.debug("Received command %s", message) segments = _unpack(message) action, params = segments[0], segments[1:] _logger.debug("Adding %s, %s to the request queue", action, params) self._command = _Command(action, params) def process_command(self): if not self._command: return else: _logger.debug("process_command: %s", self._command.action) command = self._command _logger.debug("Picked %s, %s, %s", self._command.action, self._command.params, self._command.started_at) function = getattr(self, "do_" + command.action.lower(), None) if not function: raise NotImplementedError("%s is not a valid action") else: try: result = function(command.started_at, *command.params) except: _logger.exception("Problem calling %s with %s", command.action, command.params) result = None _logger.debug("result = %s", result) # # result will be Continue if the action cannot be completed # (eg a discovery) but its time is not yet expired. Leave # the command on the stack for now. # if result is Continue: return # # If we get a result, add the result to the response # queue and pop the request off the stack. # self._command.response = result def run(self): _logger.info("Starting discovery") while not self._stop_event.wait(0): try: # # If we're not already processing one, check for an command # to advertise/discover from a local process. # if not self._command: self.poll_command_request() # # If we're paused no adverts will be broadcast. Adverts # will be received and stale ones expired # if not self._is_paused: # # Broadcast the first advert whose advertising schedule # has arrived # self.broadcast_one_advert() # # See if an advert broadcast has arrived # self.listen_for_one_advert() # # See if any adverts have expired # self.remove_expired_adverts() # # If we're processing a command, see if it's complete # if self._command: self.process_command() self.poll_command_reponse() except: _logger.exception("Problem in beacon thread") break _logger.info("Ending discovery") self.rpc.close() self.socket.close()
tjguk/networkzero
networkzero/messenger.py
send_message_to
python
def send_message_to(address, message=EMPTY, wait_for_reply_s=config.FOREVER): _logger.info("Sending message %s to %s", message, address) if isinstance(address, list): raise core.InvalidAddressError("Multiple addresses are not allowed") return sockets._sockets.send_message_to(address, message, wait_for_reply_s)
Send a message and return the reply :param address: a nw0 address (eg from `nw0.discover`) :param message: any simple Python object, including text & tuples :param wait_for_reply_s: how many seconds to wait for a reply [default: forever] :returns: the reply returned from the address or None if out of time
train
https://github.com/tjguk/networkzero/blob/0e3e81d2e9200b25a83ac07741612283599486d7/networkzero/messenger.py#L11-L23
[ "def send_message_to(self, address, message, wait_for_reply_s):\n socket = self.get_socket(address, \"speaker\")\n serialised_message = _serialise(message)\n socket.send(serialised_message)\n return _unserialise(self._receive_with_timeout(socket, wait_for_reply_s))\n" ]
# -*- coding: utf-8 -*- import uuid from . import config from . import core from . import sockets _logger = core.get_logger(__name__) EMPTY = None def send_message_to(address, message=EMPTY, wait_for_reply_s=config.FOREVER): """Send a message and return the reply :param address: a nw0 address (eg from `nw0.discover`) :param message: any simple Python object, including text & tuples :param wait_for_reply_s: how many seconds to wait for a reply [default: forever] :returns: the reply returned from the address or None if out of time """ _logger.info("Sending message %s to %s", message, address) if isinstance(address, list): raise core.InvalidAddressError("Multiple addresses are not allowed") return sockets._sockets.send_message_to(address, message, wait_for_reply_s) def wait_for_message_from(address, wait_for_s=config.FOREVER, autoreply=False): """Wait for a message :param address: a nw0 address (eg from `nw0.advertise`) :param wait_for_s: how many seconds to wait for a message before giving up [default: forever] :param autoreply: whether to send an empty reply [default: No] :returns: the message received from another address or None if out of time """ _logger.info("Waiting for message on %s for %s secs", address, wait_for_s) message = sockets._sockets.wait_for_message_from(address, wait_for_s) if message is not None and autoreply: sockets._sockets.send_reply_to(address, EMPTY) return message def send_reply_to(address, reply=EMPTY): """Reply to a message previously received :param address: a nw0 address (eg from `nw0.advertise`) :param reply: any simple Python object, including text & tuples """ _logger.debug("Sending reply %s to %s", reply, address) return sockets._sockets.send_reply_to(address, reply) def send_news_to(address, topic, data=None): """Publish news to all subscribers :param address: a nw0 address, eg from `nw0.advertise` :param topic: any text object :param data: any simple Python object including test & tuples [default: empty] """ _logger.info("Publish topic %s with data %s to %s", topic, data, address) return sockets._sockets.send_news_to(address, topic, data) def wait_for_news_from(address, prefix=config.EVERYTHING, wait_for_s=config.FOREVER, is_raw=False): """Wait for news whose topic starts with `prefix`. :param address: a nw0 address, eg from `nw0.discover` :param prefix: any text object [default: all messages] :param wait_for_s: how many seconds to wait before giving up [default: forever] :returns: a 2-tuple of (topic, data) or (None, None) if out of time """ _logger.info("Listen on %s for news matching %s waiting for %s secs", address, prefix, wait_for_s) return sockets._sockets.wait_for_news_from(address, prefix, wait_for_s, is_raw)
tjguk/networkzero
networkzero/messenger.py
wait_for_message_from
python
def wait_for_message_from(address, wait_for_s=config.FOREVER, autoreply=False): _logger.info("Waiting for message on %s for %s secs", address, wait_for_s) message = sockets._sockets.wait_for_message_from(address, wait_for_s) if message is not None and autoreply: sockets._sockets.send_reply_to(address, EMPTY) return message
Wait for a message :param address: a nw0 address (eg from `nw0.advertise`) :param wait_for_s: how many seconds to wait for a message before giving up [default: forever] :param autoreply: whether to send an empty reply [default: No] :returns: the message received from another address or None if out of time
train
https://github.com/tjguk/networkzero/blob/0e3e81d2e9200b25a83ac07741612283599486d7/networkzero/messenger.py#L25-L38
[ "def wait_for_message_from(self, address, wait_for_s):\n socket = self.get_socket(address, \"listener\")\n try:\n message = self._receive_with_timeout(socket, wait_for_s)\n except (core.SocketTimedOutError):\n return None\n else:\n return _unserialise(message)\n", "def send_reply_...
# -*- coding: utf-8 -*- import uuid from . import config from . import core from . import sockets _logger = core.get_logger(__name__) EMPTY = None def send_message_to(address, message=EMPTY, wait_for_reply_s=config.FOREVER): """Send a message and return the reply :param address: a nw0 address (eg from `nw0.discover`) :param message: any simple Python object, including text & tuples :param wait_for_reply_s: how many seconds to wait for a reply [default: forever] :returns: the reply returned from the address or None if out of time """ _logger.info("Sending message %s to %s", message, address) if isinstance(address, list): raise core.InvalidAddressError("Multiple addresses are not allowed") return sockets._sockets.send_message_to(address, message, wait_for_reply_s) def wait_for_message_from(address, wait_for_s=config.FOREVER, autoreply=False): """Wait for a message :param address: a nw0 address (eg from `nw0.advertise`) :param wait_for_s: how many seconds to wait for a message before giving up [default: forever] :param autoreply: whether to send an empty reply [default: No] :returns: the message received from another address or None if out of time """ _logger.info("Waiting for message on %s for %s secs", address, wait_for_s) message = sockets._sockets.wait_for_message_from(address, wait_for_s) if message is not None and autoreply: sockets._sockets.send_reply_to(address, EMPTY) return message def send_reply_to(address, reply=EMPTY): """Reply to a message previously received :param address: a nw0 address (eg from `nw0.advertise`) :param reply: any simple Python object, including text & tuples """ _logger.debug("Sending reply %s to %s", reply, address) return sockets._sockets.send_reply_to(address, reply) def send_news_to(address, topic, data=None): """Publish news to all subscribers :param address: a nw0 address, eg from `nw0.advertise` :param topic: any text object :param data: any simple Python object including test & tuples [default: empty] """ _logger.info("Publish topic %s with data %s to %s", topic, data, address) return sockets._sockets.send_news_to(address, topic, data) def wait_for_news_from(address, prefix=config.EVERYTHING, wait_for_s=config.FOREVER, is_raw=False): """Wait for news whose topic starts with `prefix`. :param address: a nw0 address, eg from `nw0.discover` :param prefix: any text object [default: all messages] :param wait_for_s: how many seconds to wait before giving up [default: forever] :returns: a 2-tuple of (topic, data) or (None, None) if out of time """ _logger.info("Listen on %s for news matching %s waiting for %s secs", address, prefix, wait_for_s) return sockets._sockets.wait_for_news_from(address, prefix, wait_for_s, is_raw)
tjguk/networkzero
networkzero/messenger.py
send_reply_to
python
def send_reply_to(address, reply=EMPTY): _logger.debug("Sending reply %s to %s", reply, address) return sockets._sockets.send_reply_to(address, reply)
Reply to a message previously received :param address: a nw0 address (eg from `nw0.advertise`) :param reply: any simple Python object, including text & tuples
train
https://github.com/tjguk/networkzero/blob/0e3e81d2e9200b25a83ac07741612283599486d7/networkzero/messenger.py#L40-L47
[ "def send_reply_to(self, address, reply):\n socket = self.get_socket(address, \"listener\")\n reply = _serialise(reply)\n return socket.send(reply)\n" ]
# -*- coding: utf-8 -*- import uuid from . import config from . import core from . import sockets _logger = core.get_logger(__name__) EMPTY = None def send_message_to(address, message=EMPTY, wait_for_reply_s=config.FOREVER): """Send a message and return the reply :param address: a nw0 address (eg from `nw0.discover`) :param message: any simple Python object, including text & tuples :param wait_for_reply_s: how many seconds to wait for a reply [default: forever] :returns: the reply returned from the address or None if out of time """ _logger.info("Sending message %s to %s", message, address) if isinstance(address, list): raise core.InvalidAddressError("Multiple addresses are not allowed") return sockets._sockets.send_message_to(address, message, wait_for_reply_s) def wait_for_message_from(address, wait_for_s=config.FOREVER, autoreply=False): """Wait for a message :param address: a nw0 address (eg from `nw0.advertise`) :param wait_for_s: how many seconds to wait for a message before giving up [default: forever] :param autoreply: whether to send an empty reply [default: No] :returns: the message received from another address or None if out of time """ _logger.info("Waiting for message on %s for %s secs", address, wait_for_s) message = sockets._sockets.wait_for_message_from(address, wait_for_s) if message is not None and autoreply: sockets._sockets.send_reply_to(address, EMPTY) return message def send_reply_to(address, reply=EMPTY): """Reply to a message previously received :param address: a nw0 address (eg from `nw0.advertise`) :param reply: any simple Python object, including text & tuples """ _logger.debug("Sending reply %s to %s", reply, address) return sockets._sockets.send_reply_to(address, reply) def send_news_to(address, topic, data=None): """Publish news to all subscribers :param address: a nw0 address, eg from `nw0.advertise` :param topic: any text object :param data: any simple Python object including test & tuples [default: empty] """ _logger.info("Publish topic %s with data %s to %s", topic, data, address) return sockets._sockets.send_news_to(address, topic, data) def wait_for_news_from(address, prefix=config.EVERYTHING, wait_for_s=config.FOREVER, is_raw=False): """Wait for news whose topic starts with `prefix`. :param address: a nw0 address, eg from `nw0.discover` :param prefix: any text object [default: all messages] :param wait_for_s: how many seconds to wait before giving up [default: forever] :returns: a 2-tuple of (topic, data) or (None, None) if out of time """ _logger.info("Listen on %s for news matching %s waiting for %s secs", address, prefix, wait_for_s) return sockets._sockets.wait_for_news_from(address, prefix, wait_for_s, is_raw)
tjguk/networkzero
networkzero/messenger.py
send_news_to
python
def send_news_to(address, topic, data=None): _logger.info("Publish topic %s with data %s to %s", topic, data, address) return sockets._sockets.send_news_to(address, topic, data)
Publish news to all subscribers :param address: a nw0 address, eg from `nw0.advertise` :param topic: any text object :param data: any simple Python object including test & tuples [default: empty]
train
https://github.com/tjguk/networkzero/blob/0e3e81d2e9200b25a83ac07741612283599486d7/networkzero/messenger.py#L49-L57
[ "def send_news_to(self, address, topic, data):\n socket = self.get_socket(address, \"publisher\")\n return socket.send_multipart(_serialise_for_pubsub(topic, data))\n" ]
# -*- coding: utf-8 -*- import uuid from . import config from . import core from . import sockets _logger = core.get_logger(__name__) EMPTY = None def send_message_to(address, message=EMPTY, wait_for_reply_s=config.FOREVER): """Send a message and return the reply :param address: a nw0 address (eg from `nw0.discover`) :param message: any simple Python object, including text & tuples :param wait_for_reply_s: how many seconds to wait for a reply [default: forever] :returns: the reply returned from the address or None if out of time """ _logger.info("Sending message %s to %s", message, address) if isinstance(address, list): raise core.InvalidAddressError("Multiple addresses are not allowed") return sockets._sockets.send_message_to(address, message, wait_for_reply_s) def wait_for_message_from(address, wait_for_s=config.FOREVER, autoreply=False): """Wait for a message :param address: a nw0 address (eg from `nw0.advertise`) :param wait_for_s: how many seconds to wait for a message before giving up [default: forever] :param autoreply: whether to send an empty reply [default: No] :returns: the message received from another address or None if out of time """ _logger.info("Waiting for message on %s for %s secs", address, wait_for_s) message = sockets._sockets.wait_for_message_from(address, wait_for_s) if message is not None and autoreply: sockets._sockets.send_reply_to(address, EMPTY) return message def send_reply_to(address, reply=EMPTY): """Reply to a message previously received :param address: a nw0 address (eg from `nw0.advertise`) :param reply: any simple Python object, including text & tuples """ _logger.debug("Sending reply %s to %s", reply, address) return sockets._sockets.send_reply_to(address, reply) def send_news_to(address, topic, data=None): """Publish news to all subscribers :param address: a nw0 address, eg from `nw0.advertise` :param topic: any text object :param data: any simple Python object including test & tuples [default: empty] """ _logger.info("Publish topic %s with data %s to %s", topic, data, address) return sockets._sockets.send_news_to(address, topic, data) def wait_for_news_from(address, prefix=config.EVERYTHING, wait_for_s=config.FOREVER, is_raw=False): """Wait for news whose topic starts with `prefix`. :param address: a nw0 address, eg from `nw0.discover` :param prefix: any text object [default: all messages] :param wait_for_s: how many seconds to wait before giving up [default: forever] :returns: a 2-tuple of (topic, data) or (None, None) if out of time """ _logger.info("Listen on %s for news matching %s waiting for %s secs", address, prefix, wait_for_s) return sockets._sockets.wait_for_news_from(address, prefix, wait_for_s, is_raw)
tjguk/networkzero
networkzero/messenger.py
wait_for_news_from
python
def wait_for_news_from(address, prefix=config.EVERYTHING, wait_for_s=config.FOREVER, is_raw=False): _logger.info("Listen on %s for news matching %s waiting for %s secs", address, prefix, wait_for_s) return sockets._sockets.wait_for_news_from(address, prefix, wait_for_s, is_raw)
Wait for news whose topic starts with `prefix`. :param address: a nw0 address, eg from `nw0.discover` :param prefix: any text object [default: all messages] :param wait_for_s: how many seconds to wait before giving up [default: forever] :returns: a 2-tuple of (topic, data) or (None, None) if out of time
train
https://github.com/tjguk/networkzero/blob/0e3e81d2e9200b25a83ac07741612283599486d7/networkzero/messenger.py#L59-L69
[ "def wait_for_news_from(self, address, topic, wait_for_s, is_raw=False):\n if isinstance(address, list):\n addresses = address\n else:\n addresses = [address]\n socket = self.get_socket(addresses, \"subscriber\")\n if isinstance(topic, str):\n topics = [topic]\n else:\n to...
# -*- coding: utf-8 -*- import uuid from . import config from . import core from . import sockets _logger = core.get_logger(__name__) EMPTY = None def send_message_to(address, message=EMPTY, wait_for_reply_s=config.FOREVER): """Send a message and return the reply :param address: a nw0 address (eg from `nw0.discover`) :param message: any simple Python object, including text & tuples :param wait_for_reply_s: how many seconds to wait for a reply [default: forever] :returns: the reply returned from the address or None if out of time """ _logger.info("Sending message %s to %s", message, address) if isinstance(address, list): raise core.InvalidAddressError("Multiple addresses are not allowed") return sockets._sockets.send_message_to(address, message, wait_for_reply_s) def wait_for_message_from(address, wait_for_s=config.FOREVER, autoreply=False): """Wait for a message :param address: a nw0 address (eg from `nw0.advertise`) :param wait_for_s: how many seconds to wait for a message before giving up [default: forever] :param autoreply: whether to send an empty reply [default: No] :returns: the message received from another address or None if out of time """ _logger.info("Waiting for message on %s for %s secs", address, wait_for_s) message = sockets._sockets.wait_for_message_from(address, wait_for_s) if message is not None and autoreply: sockets._sockets.send_reply_to(address, EMPTY) return message def send_reply_to(address, reply=EMPTY): """Reply to a message previously received :param address: a nw0 address (eg from `nw0.advertise`) :param reply: any simple Python object, including text & tuples """ _logger.debug("Sending reply %s to %s", reply, address) return sockets._sockets.send_reply_to(address, reply) def send_news_to(address, topic, data=None): """Publish news to all subscribers :param address: a nw0 address, eg from `nw0.advertise` :param topic: any text object :param data: any simple Python object including test & tuples [default: empty] """ _logger.info("Publish topic %s with data %s to %s", topic, data, address) return sockets._sockets.send_news_to(address, topic, data) def wait_for_news_from(address, prefix=config.EVERYTHING, wait_for_s=config.FOREVER, is_raw=False): """Wait for news whose topic starts with `prefix`. :param address: a nw0 address, eg from `nw0.discover` :param prefix: any text object [default: all messages] :param wait_for_s: how many seconds to wait before giving up [default: forever] :returns: a 2-tuple of (topic, data) or (None, None) if out of time """ _logger.info("Listen on %s for news matching %s waiting for %s secs", address, prefix, wait_for_s) return sockets._sockets.wait_for_news_from(address, prefix, wait_for_s, is_raw)
agile4you/bottle-neck
bottle_neck/routing.py
Route.wrap_callable
python
def wrap_callable(cls, uri, methods, callable_obj): if isinstance(callable_obj, HandlerMeta): callable_obj.base_endpoint = uri callable_obj.is_valid = True return callable_obj if isinstance(callable_obj, types.FunctionType): return cls(uri=uri, methods=methods, callable_obj=callable_obj) raise RouteError("Invalid handler type.")
Wraps function-based callable_obj into a `Route` instance, else proxies a `bottle_neck.handlers.BaseHandler` subclass instance. Args: uri (str): The uri relative path. methods (tuple): A tuple of valid method strings. callable_obj (instance): The callable object. Returns: A route instance. Raises: RouteError for invalid callable object type.
train
https://github.com/agile4you/bottle-neck/blob/ebc670a4b178255473d68e9b4122ba04e38f4810/bottle_neck/routing.py#L38-L61
null
class Route(object): """Base Route interface. It wraps function-based web handlers in order to provide a same interface for `bottle_neck.routing.Router` class functionality. """ __slots__ = ('uri', 'methods', 'callable_obj') @classmethod def __init__(self, uri, methods, callable_obj): self.uri = uri, self.methods = methods, self.callable_obj = callable_obj @property def is_valid(self): args = [self.uri, self.methods, self.callable_obj] return all([arg for arg in args]) def register_app(self, app): """Register the route object to a `bottle.Bottle` app instance. Args: app (instance): Returns: Route instance (for chaining purposes) """ app.route(self.uri, methods=self.methods)(self.callable_obj) return self
agile4you/bottle-neck
bottle_neck/routing.py
Route.register_app
python
def register_app(self, app): app.route(self.uri, methods=self.methods)(self.callable_obj) return self
Register the route object to a `bottle.Bottle` app instance. Args: app (instance): Returns: Route instance (for chaining purposes)
train
https://github.com/agile4you/bottle-neck/blob/ebc670a4b178255473d68e9b4122ba04e38f4810/bottle_neck/routing.py#L73-L84
null
class Route(object): """Base Route interface. It wraps function-based web handlers in order to provide a same interface for `bottle_neck.routing.Router` class functionality. """ __slots__ = ('uri', 'methods', 'callable_obj') @classmethod def wrap_callable(cls, uri, methods, callable_obj): """Wraps function-based callable_obj into a `Route` instance, else proxies a `bottle_neck.handlers.BaseHandler` subclass instance. Args: uri (str): The uri relative path. methods (tuple): A tuple of valid method strings. callable_obj (instance): The callable object. Returns: A route instance. Raises: RouteError for invalid callable object type. """ if isinstance(callable_obj, HandlerMeta): callable_obj.base_endpoint = uri callable_obj.is_valid = True return callable_obj if isinstance(callable_obj, types.FunctionType): return cls(uri=uri, methods=methods, callable_obj=callable_obj) raise RouteError("Invalid handler type.") def __init__(self, uri, methods, callable_obj): self.uri = uri, self.methods = methods, self.callable_obj = callable_obj @property def is_valid(self): args = [self.uri, self.methods, self.callable_obj] return all([arg for arg in args])
agile4you/bottle-neck
bottle_neck/routing.py
Router.register_handler
python
def register_handler(self, callable_obj, entrypoint, methods=('GET',)): router_obj = Route.wrap_callable( uri=entrypoint, methods=methods, callable_obj=callable_obj ) if router_obj.is_valid: self._routes.add(router_obj) return self raise RouteError( # pragma: no cover "Missing params: methods: {} - entrypoint: {}".format( methods, entrypoint ) )
Register a handler callable to a specific route. Args: entrypoint (str): The uri relative path. methods (tuple): A tuple of valid method strings. callable_obj (callable): The callable object. Returns: The Router instance (for chaining purposes). Raises: RouteError, for missing routing params or invalid callable object type.
train
https://github.com/agile4you/bottle-neck/blob/ebc670a4b178255473d68e9b4122ba04e38f4810/bottle_neck/routing.py#L98-L128
[ "def wrap_callable(cls, uri, methods, callable_obj):\n \"\"\"Wraps function-based callable_obj into a `Route` instance, else\n proxies a `bottle_neck.handlers.BaseHandler` subclass instance.\n\n Args:\n uri (str): The uri relative path.\n methods (tuple): A tuple of valid method strings.\n ...
class Router(object): """Base Router class for bottle.py WSGI applications. """ def __init__(self): self._routes = set() @property def routes(self): # pragma: no cover return self._routes def register_handler(self, callable_obj, entrypoint, methods=('GET',)): """Register a handler callable to a specific route. Args: entrypoint (str): The uri relative path. methods (tuple): A tuple of valid method strings. callable_obj (callable): The callable object. Returns: The Router instance (for chaining purposes). Raises: RouteError, for missing routing params or invalid callable object type. """ router_obj = Route.wrap_callable( uri=entrypoint, methods=methods, callable_obj=callable_obj ) if router_obj.is_valid: self._routes.add(router_obj) return self raise RouteError( # pragma: no cover "Missing params: methods: {} - entrypoint: {}".format( methods, entrypoint ) ) def mount(self, app=None): """Mounts all registered routes to a bottle.py application instance. Args: app (instance): A `bottle.Bottle()` application instance. Returns: The Router instance (for chaining purposes). """ for endpoint in self._routes: endpoint.register_app(app) return self def __repr__(self): # pragma: no cover return 'Router object: total {} routes'.format(len(self)) def __len__(self): return len(self._routes) def __iter__(self): # pragma: no cover for route in self._routes: yield route
agile4you/bottle-neck
bottle_neck/routing.py
Router.mount
python
def mount(self, app=None): for endpoint in self._routes: endpoint.register_app(app) return self
Mounts all registered routes to a bottle.py application instance. Args: app (instance): A `bottle.Bottle()` application instance. Returns: The Router instance (for chaining purposes).
train
https://github.com/agile4you/bottle-neck/blob/ebc670a4b178255473d68e9b4122ba04e38f4810/bottle_neck/routing.py#L130-L142
null
class Router(object): """Base Router class for bottle.py WSGI applications. """ def __init__(self): self._routes = set() @property def routes(self): # pragma: no cover return self._routes def register_handler(self, callable_obj, entrypoint, methods=('GET',)): """Register a handler callable to a specific route. Args: entrypoint (str): The uri relative path. methods (tuple): A tuple of valid method strings. callable_obj (callable): The callable object. Returns: The Router instance (for chaining purposes). Raises: RouteError, for missing routing params or invalid callable object type. """ router_obj = Route.wrap_callable( uri=entrypoint, methods=methods, callable_obj=callable_obj ) if router_obj.is_valid: self._routes.add(router_obj) return self raise RouteError( # pragma: no cover "Missing params: methods: {} - entrypoint: {}".format( methods, entrypoint ) ) def __repr__(self): # pragma: no cover return 'Router object: total {} routes'.format(len(self)) def __len__(self): return len(self._routes) def __iter__(self): # pragma: no cover for route in self._routes: yield route
agile4you/bottle-neck
bottle_neck/response.py
WSResponse.from_status
python
def from_status(cls, status_line, msg=None): method = getattr(cls, status_line.lower()[4:].replace(' ', '_')) return method(msg)
Returns a class method from bottle.HTTPError.status_line attribute. Useful for patching `bottle.HTTPError` for web services. Args: status_line (str): bottle.HTTPError.status_line text. msg: The message data for response. Returns: Class method based on status_line arg. Examples: >>> status_line = '401 Unauthorized' >>> error_msg = 'Get out!' >>> resp = WSResponse.from_status(status_line, error_msg) >>> resp['errors'] ['Get out!'] >>> resp['status_text'] 'Unauthorized'
train
https://github.com/agile4you/bottle-neck/blob/ebc670a4b178255473d68e9b4122ba04e38f4810/bottle_neck/response.py#L111-L132
null
class WSResponse(object): """Base web service response class. WSResponse class provides a unified API for HTTP responses. The Response body skeleton looks like this:: { "status_code": 200, "status_txt": "OK", "response_data": { "id": 65234 "username": "pav" "email": "pav@geotagaeroview.com" }, "errors": [] } Attributes: status_code (int): Response HTTP status code. data (object): Response key/value data. errors (dict): Response key/value errors. Examples: >>> response = WSResponse( ... status_code=200, ... data='Hi' ... ) ... >>> response.to_json OrderedDict([('status_code', 200), ('status_text', 'OK'), ('data', 'Hi'), ('errors', [])]) """ __slots__ = ['status_code', 'data', 'errors'] expose_status = True response = bottle.response def __init__(self, status_code=200, data=None, errors=None): if status_code not in dict(HTTP_CODES) or\ not isinstance(errors, six.string_types + (list, tuple, type(None),)): raise WSResponseError('Invalid Response initialization.') self.status_code = status_code self.data = data or {} if isinstance(errors, (six.string_types, )): errors = [errors] self.errors = errors def __repr__(self): # pragma: no cover return "WebService Response: status={}, data={}".format( self.status_code, str(self.data) ) def __eq__(self, other): # pragma: no cover assert isinstance(other, WSResponse), 'Invalid Type for eq operator.' return self.status_code == other.status_code and \ self.data == self.data __str__ = __repr__ @classmethod @classmethod def ok(cls, data): """Shortcut API for HTTP 200 `OK` response. Args: data (object): Response key/value data. Returns WSResponse Instance. """ return cls( status_code=200, data=data ).to_json @classmethod def created(cls, data=None): """Shortcut API for HTTP 201 `Created` response. Args: data (object): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '201 Created' return cls(201, data=data).to_json @classmethod def not_modified(cls, errors=None): """Shortcut API for HTTP 304 `Not Modified` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '304 Not Modified' return cls(304, None, errors).to_json @classmethod def bad_request(cls, errors=None): """Shortcut API for HTTP 400 `Bad Request` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '400 Bad Request' return cls(400, errors=errors).to_json @classmethod def unauthorized(cls, errors=None): """Shortcut API for HTTP 401 `Unauthorized` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '401 Unauthorized' return cls(401, errors=errors).to_json @classmethod def forbidden(cls, errors=None): """Shortcut API for HTTP 403 `Forbidden` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '403 Forbidden' return cls(403, errors=errors).to_json @classmethod def not_found(cls, errors=None): """Shortcut API for HTTP 404 `Not found` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '404 Not Found' return cls(404, None, errors).to_json @classmethod def method_not_allowed(cls, errors=None): """Shortcut API for HTTP 405 `Method not allowed` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '405 Method Not Allowed' return cls(405, None, errors).to_json @classmethod def not_implemented(cls, errors=None): """Shortcut API for HTTP 501 `Not Implemented` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '501 Not Implemented' return cls(501, None, errors).to_json @classmethod def service_unavailable(cls, errors=None): """Shortcut API for HTTP 503 `Service Unavailable` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '503 Service Unavailable' return cls(503, None, errors).to_json @property def to_json(self): """Short cut for JSON response service data. Returns: Dict that implements JSON interface. """ web_resp = collections.OrderedDict() web_resp['status_code'] = self.status_code web_resp['status_text'] = dict(HTTP_CODES).get(self.status_code) web_resp['data'] = self.data if self.data is not None else {} web_resp['errors'] = self.errors or [] return web_resp
agile4you/bottle-neck
bottle_neck/response.py
WSResponse.created
python
def created(cls, data=None): if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '201 Created' return cls(201, data=data).to_json
Shortcut API for HTTP 201 `Created` response. Args: data (object): Response key/value data. Returns: WSResponse Instance.
train
https://github.com/agile4you/bottle-neck/blob/ebc670a4b178255473d68e9b4122ba04e38f4810/bottle_neck/response.py#L150-L163
null
class WSResponse(object): """Base web service response class. WSResponse class provides a unified API for HTTP responses. The Response body skeleton looks like this:: { "status_code": 200, "status_txt": "OK", "response_data": { "id": 65234 "username": "pav" "email": "pav@geotagaeroview.com" }, "errors": [] } Attributes: status_code (int): Response HTTP status code. data (object): Response key/value data. errors (dict): Response key/value errors. Examples: >>> response = WSResponse( ... status_code=200, ... data='Hi' ... ) ... >>> response.to_json OrderedDict([('status_code', 200), ('status_text', 'OK'), ('data', 'Hi'), ('errors', [])]) """ __slots__ = ['status_code', 'data', 'errors'] expose_status = True response = bottle.response def __init__(self, status_code=200, data=None, errors=None): if status_code not in dict(HTTP_CODES) or\ not isinstance(errors, six.string_types + (list, tuple, type(None),)): raise WSResponseError('Invalid Response initialization.') self.status_code = status_code self.data = data or {} if isinstance(errors, (six.string_types, )): errors = [errors] self.errors = errors def __repr__(self): # pragma: no cover return "WebService Response: status={}, data={}".format( self.status_code, str(self.data) ) def __eq__(self, other): # pragma: no cover assert isinstance(other, WSResponse), 'Invalid Type for eq operator.' return self.status_code == other.status_code and \ self.data == self.data __str__ = __repr__ @classmethod def from_status(cls, status_line, msg=None): """Returns a class method from bottle.HTTPError.status_line attribute. Useful for patching `bottle.HTTPError` for web services. Args: status_line (str): bottle.HTTPError.status_line text. msg: The message data for response. Returns: Class method based on status_line arg. Examples: >>> status_line = '401 Unauthorized' >>> error_msg = 'Get out!' >>> resp = WSResponse.from_status(status_line, error_msg) >>> resp['errors'] ['Get out!'] >>> resp['status_text'] 'Unauthorized' """ method = getattr(cls, status_line.lower()[4:].replace(' ', '_')) return method(msg) @classmethod def ok(cls, data): """Shortcut API for HTTP 200 `OK` response. Args: data (object): Response key/value data. Returns WSResponse Instance. """ return cls( status_code=200, data=data ).to_json @classmethod @classmethod def not_modified(cls, errors=None): """Shortcut API for HTTP 304 `Not Modified` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '304 Not Modified' return cls(304, None, errors).to_json @classmethod def bad_request(cls, errors=None): """Shortcut API for HTTP 400 `Bad Request` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '400 Bad Request' return cls(400, errors=errors).to_json @classmethod def unauthorized(cls, errors=None): """Shortcut API for HTTP 401 `Unauthorized` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '401 Unauthorized' return cls(401, errors=errors).to_json @classmethod def forbidden(cls, errors=None): """Shortcut API for HTTP 403 `Forbidden` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '403 Forbidden' return cls(403, errors=errors).to_json @classmethod def not_found(cls, errors=None): """Shortcut API for HTTP 404 `Not found` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '404 Not Found' return cls(404, None, errors).to_json @classmethod def method_not_allowed(cls, errors=None): """Shortcut API for HTTP 405 `Method not allowed` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '405 Method Not Allowed' return cls(405, None, errors).to_json @classmethod def not_implemented(cls, errors=None): """Shortcut API for HTTP 501 `Not Implemented` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '501 Not Implemented' return cls(501, None, errors).to_json @classmethod def service_unavailable(cls, errors=None): """Shortcut API for HTTP 503 `Service Unavailable` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '503 Service Unavailable' return cls(503, None, errors).to_json @property def to_json(self): """Short cut for JSON response service data. Returns: Dict that implements JSON interface. """ web_resp = collections.OrderedDict() web_resp['status_code'] = self.status_code web_resp['status_text'] = dict(HTTP_CODES).get(self.status_code) web_resp['data'] = self.data if self.data is not None else {} web_resp['errors'] = self.errors or [] return web_resp
agile4you/bottle-neck
bottle_neck/response.py
WSResponse.not_modified
python
def not_modified(cls, errors=None): if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '304 Not Modified' return cls(304, None, errors).to_json
Shortcut API for HTTP 304 `Not Modified` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance.
train
https://github.com/agile4you/bottle-neck/blob/ebc670a4b178255473d68e9b4122ba04e38f4810/bottle_neck/response.py#L166-L179
null
class WSResponse(object): """Base web service response class. WSResponse class provides a unified API for HTTP responses. The Response body skeleton looks like this:: { "status_code": 200, "status_txt": "OK", "response_data": { "id": 65234 "username": "pav" "email": "pav@geotagaeroview.com" }, "errors": [] } Attributes: status_code (int): Response HTTP status code. data (object): Response key/value data. errors (dict): Response key/value errors. Examples: >>> response = WSResponse( ... status_code=200, ... data='Hi' ... ) ... >>> response.to_json OrderedDict([('status_code', 200), ('status_text', 'OK'), ('data', 'Hi'), ('errors', [])]) """ __slots__ = ['status_code', 'data', 'errors'] expose_status = True response = bottle.response def __init__(self, status_code=200, data=None, errors=None): if status_code not in dict(HTTP_CODES) or\ not isinstance(errors, six.string_types + (list, tuple, type(None),)): raise WSResponseError('Invalid Response initialization.') self.status_code = status_code self.data = data or {} if isinstance(errors, (six.string_types, )): errors = [errors] self.errors = errors def __repr__(self): # pragma: no cover return "WebService Response: status={}, data={}".format( self.status_code, str(self.data) ) def __eq__(self, other): # pragma: no cover assert isinstance(other, WSResponse), 'Invalid Type for eq operator.' return self.status_code == other.status_code and \ self.data == self.data __str__ = __repr__ @classmethod def from_status(cls, status_line, msg=None): """Returns a class method from bottle.HTTPError.status_line attribute. Useful for patching `bottle.HTTPError` for web services. Args: status_line (str): bottle.HTTPError.status_line text. msg: The message data for response. Returns: Class method based on status_line arg. Examples: >>> status_line = '401 Unauthorized' >>> error_msg = 'Get out!' >>> resp = WSResponse.from_status(status_line, error_msg) >>> resp['errors'] ['Get out!'] >>> resp['status_text'] 'Unauthorized' """ method = getattr(cls, status_line.lower()[4:].replace(' ', '_')) return method(msg) @classmethod def ok(cls, data): """Shortcut API for HTTP 200 `OK` response. Args: data (object): Response key/value data. Returns WSResponse Instance. """ return cls( status_code=200, data=data ).to_json @classmethod def created(cls, data=None): """Shortcut API for HTTP 201 `Created` response. Args: data (object): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '201 Created' return cls(201, data=data).to_json @classmethod @classmethod def bad_request(cls, errors=None): """Shortcut API for HTTP 400 `Bad Request` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '400 Bad Request' return cls(400, errors=errors).to_json @classmethod def unauthorized(cls, errors=None): """Shortcut API for HTTP 401 `Unauthorized` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '401 Unauthorized' return cls(401, errors=errors).to_json @classmethod def forbidden(cls, errors=None): """Shortcut API for HTTP 403 `Forbidden` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '403 Forbidden' return cls(403, errors=errors).to_json @classmethod def not_found(cls, errors=None): """Shortcut API for HTTP 404 `Not found` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '404 Not Found' return cls(404, None, errors).to_json @classmethod def method_not_allowed(cls, errors=None): """Shortcut API for HTTP 405 `Method not allowed` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '405 Method Not Allowed' return cls(405, None, errors).to_json @classmethod def not_implemented(cls, errors=None): """Shortcut API for HTTP 501 `Not Implemented` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '501 Not Implemented' return cls(501, None, errors).to_json @classmethod def service_unavailable(cls, errors=None): """Shortcut API for HTTP 503 `Service Unavailable` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '503 Service Unavailable' return cls(503, None, errors).to_json @property def to_json(self): """Short cut for JSON response service data. Returns: Dict that implements JSON interface. """ web_resp = collections.OrderedDict() web_resp['status_code'] = self.status_code web_resp['status_text'] = dict(HTTP_CODES).get(self.status_code) web_resp['data'] = self.data if self.data is not None else {} web_resp['errors'] = self.errors or [] return web_resp
agile4you/bottle-neck
bottle_neck/response.py
WSResponse.bad_request
python
def bad_request(cls, errors=None): if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '400 Bad Request' return cls(400, errors=errors).to_json
Shortcut API for HTTP 400 `Bad Request` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance.
train
https://github.com/agile4you/bottle-neck/blob/ebc670a4b178255473d68e9b4122ba04e38f4810/bottle_neck/response.py#L182-L195
null
class WSResponse(object): """Base web service response class. WSResponse class provides a unified API for HTTP responses. The Response body skeleton looks like this:: { "status_code": 200, "status_txt": "OK", "response_data": { "id": 65234 "username": "pav" "email": "pav@geotagaeroview.com" }, "errors": [] } Attributes: status_code (int): Response HTTP status code. data (object): Response key/value data. errors (dict): Response key/value errors. Examples: >>> response = WSResponse( ... status_code=200, ... data='Hi' ... ) ... >>> response.to_json OrderedDict([('status_code', 200), ('status_text', 'OK'), ('data', 'Hi'), ('errors', [])]) """ __slots__ = ['status_code', 'data', 'errors'] expose_status = True response = bottle.response def __init__(self, status_code=200, data=None, errors=None): if status_code not in dict(HTTP_CODES) or\ not isinstance(errors, six.string_types + (list, tuple, type(None),)): raise WSResponseError('Invalid Response initialization.') self.status_code = status_code self.data = data or {} if isinstance(errors, (six.string_types, )): errors = [errors] self.errors = errors def __repr__(self): # pragma: no cover return "WebService Response: status={}, data={}".format( self.status_code, str(self.data) ) def __eq__(self, other): # pragma: no cover assert isinstance(other, WSResponse), 'Invalid Type for eq operator.' return self.status_code == other.status_code and \ self.data == self.data __str__ = __repr__ @classmethod def from_status(cls, status_line, msg=None): """Returns a class method from bottle.HTTPError.status_line attribute. Useful for patching `bottle.HTTPError` for web services. Args: status_line (str): bottle.HTTPError.status_line text. msg: The message data for response. Returns: Class method based on status_line arg. Examples: >>> status_line = '401 Unauthorized' >>> error_msg = 'Get out!' >>> resp = WSResponse.from_status(status_line, error_msg) >>> resp['errors'] ['Get out!'] >>> resp['status_text'] 'Unauthorized' """ method = getattr(cls, status_line.lower()[4:].replace(' ', '_')) return method(msg) @classmethod def ok(cls, data): """Shortcut API for HTTP 200 `OK` response. Args: data (object): Response key/value data. Returns WSResponse Instance. """ return cls( status_code=200, data=data ).to_json @classmethod def created(cls, data=None): """Shortcut API for HTTP 201 `Created` response. Args: data (object): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '201 Created' return cls(201, data=data).to_json @classmethod def not_modified(cls, errors=None): """Shortcut API for HTTP 304 `Not Modified` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '304 Not Modified' return cls(304, None, errors).to_json @classmethod @classmethod def unauthorized(cls, errors=None): """Shortcut API for HTTP 401 `Unauthorized` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '401 Unauthorized' return cls(401, errors=errors).to_json @classmethod def forbidden(cls, errors=None): """Shortcut API for HTTP 403 `Forbidden` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '403 Forbidden' return cls(403, errors=errors).to_json @classmethod def not_found(cls, errors=None): """Shortcut API for HTTP 404 `Not found` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '404 Not Found' return cls(404, None, errors).to_json @classmethod def method_not_allowed(cls, errors=None): """Shortcut API for HTTP 405 `Method not allowed` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '405 Method Not Allowed' return cls(405, None, errors).to_json @classmethod def not_implemented(cls, errors=None): """Shortcut API for HTTP 501 `Not Implemented` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '501 Not Implemented' return cls(501, None, errors).to_json @classmethod def service_unavailable(cls, errors=None): """Shortcut API for HTTP 503 `Service Unavailable` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '503 Service Unavailable' return cls(503, None, errors).to_json @property def to_json(self): """Short cut for JSON response service data. Returns: Dict that implements JSON interface. """ web_resp = collections.OrderedDict() web_resp['status_code'] = self.status_code web_resp['status_text'] = dict(HTTP_CODES).get(self.status_code) web_resp['data'] = self.data if self.data is not None else {} web_resp['errors'] = self.errors or [] return web_resp
agile4you/bottle-neck
bottle_neck/response.py
WSResponse.unauthorized
python
def unauthorized(cls, errors=None): if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '401 Unauthorized' return cls(401, errors=errors).to_json
Shortcut API for HTTP 401 `Unauthorized` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance.
train
https://github.com/agile4you/bottle-neck/blob/ebc670a4b178255473d68e9b4122ba04e38f4810/bottle_neck/response.py#L198-L211
null
class WSResponse(object): """Base web service response class. WSResponse class provides a unified API for HTTP responses. The Response body skeleton looks like this:: { "status_code": 200, "status_txt": "OK", "response_data": { "id": 65234 "username": "pav" "email": "pav@geotagaeroview.com" }, "errors": [] } Attributes: status_code (int): Response HTTP status code. data (object): Response key/value data. errors (dict): Response key/value errors. Examples: >>> response = WSResponse( ... status_code=200, ... data='Hi' ... ) ... >>> response.to_json OrderedDict([('status_code', 200), ('status_text', 'OK'), ('data', 'Hi'), ('errors', [])]) """ __slots__ = ['status_code', 'data', 'errors'] expose_status = True response = bottle.response def __init__(self, status_code=200, data=None, errors=None): if status_code not in dict(HTTP_CODES) or\ not isinstance(errors, six.string_types + (list, tuple, type(None),)): raise WSResponseError('Invalid Response initialization.') self.status_code = status_code self.data = data or {} if isinstance(errors, (six.string_types, )): errors = [errors] self.errors = errors def __repr__(self): # pragma: no cover return "WebService Response: status={}, data={}".format( self.status_code, str(self.data) ) def __eq__(self, other): # pragma: no cover assert isinstance(other, WSResponse), 'Invalid Type for eq operator.' return self.status_code == other.status_code and \ self.data == self.data __str__ = __repr__ @classmethod def from_status(cls, status_line, msg=None): """Returns a class method from bottle.HTTPError.status_line attribute. Useful for patching `bottle.HTTPError` for web services. Args: status_line (str): bottle.HTTPError.status_line text. msg: The message data for response. Returns: Class method based on status_line arg. Examples: >>> status_line = '401 Unauthorized' >>> error_msg = 'Get out!' >>> resp = WSResponse.from_status(status_line, error_msg) >>> resp['errors'] ['Get out!'] >>> resp['status_text'] 'Unauthorized' """ method = getattr(cls, status_line.lower()[4:].replace(' ', '_')) return method(msg) @classmethod def ok(cls, data): """Shortcut API for HTTP 200 `OK` response. Args: data (object): Response key/value data. Returns WSResponse Instance. """ return cls( status_code=200, data=data ).to_json @classmethod def created(cls, data=None): """Shortcut API for HTTP 201 `Created` response. Args: data (object): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '201 Created' return cls(201, data=data).to_json @classmethod def not_modified(cls, errors=None): """Shortcut API for HTTP 304 `Not Modified` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '304 Not Modified' return cls(304, None, errors).to_json @classmethod def bad_request(cls, errors=None): """Shortcut API for HTTP 400 `Bad Request` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '400 Bad Request' return cls(400, errors=errors).to_json @classmethod @classmethod def forbidden(cls, errors=None): """Shortcut API for HTTP 403 `Forbidden` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '403 Forbidden' return cls(403, errors=errors).to_json @classmethod def not_found(cls, errors=None): """Shortcut API for HTTP 404 `Not found` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '404 Not Found' return cls(404, None, errors).to_json @classmethod def method_not_allowed(cls, errors=None): """Shortcut API for HTTP 405 `Method not allowed` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '405 Method Not Allowed' return cls(405, None, errors).to_json @classmethod def not_implemented(cls, errors=None): """Shortcut API for HTTP 501 `Not Implemented` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '501 Not Implemented' return cls(501, None, errors).to_json @classmethod def service_unavailable(cls, errors=None): """Shortcut API for HTTP 503 `Service Unavailable` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '503 Service Unavailable' return cls(503, None, errors).to_json @property def to_json(self): """Short cut for JSON response service data. Returns: Dict that implements JSON interface. """ web_resp = collections.OrderedDict() web_resp['status_code'] = self.status_code web_resp['status_text'] = dict(HTTP_CODES).get(self.status_code) web_resp['data'] = self.data if self.data is not None else {} web_resp['errors'] = self.errors or [] return web_resp
agile4you/bottle-neck
bottle_neck/response.py
WSResponse.forbidden
python
def forbidden(cls, errors=None): if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '403 Forbidden' return cls(403, errors=errors).to_json
Shortcut API for HTTP 403 `Forbidden` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance.
train
https://github.com/agile4you/bottle-neck/blob/ebc670a4b178255473d68e9b4122ba04e38f4810/bottle_neck/response.py#L214-L227
null
class WSResponse(object): """Base web service response class. WSResponse class provides a unified API for HTTP responses. The Response body skeleton looks like this:: { "status_code": 200, "status_txt": "OK", "response_data": { "id": 65234 "username": "pav" "email": "pav@geotagaeroview.com" }, "errors": [] } Attributes: status_code (int): Response HTTP status code. data (object): Response key/value data. errors (dict): Response key/value errors. Examples: >>> response = WSResponse( ... status_code=200, ... data='Hi' ... ) ... >>> response.to_json OrderedDict([('status_code', 200), ('status_text', 'OK'), ('data', 'Hi'), ('errors', [])]) """ __slots__ = ['status_code', 'data', 'errors'] expose_status = True response = bottle.response def __init__(self, status_code=200, data=None, errors=None): if status_code not in dict(HTTP_CODES) or\ not isinstance(errors, six.string_types + (list, tuple, type(None),)): raise WSResponseError('Invalid Response initialization.') self.status_code = status_code self.data = data or {} if isinstance(errors, (six.string_types, )): errors = [errors] self.errors = errors def __repr__(self): # pragma: no cover return "WebService Response: status={}, data={}".format( self.status_code, str(self.data) ) def __eq__(self, other): # pragma: no cover assert isinstance(other, WSResponse), 'Invalid Type for eq operator.' return self.status_code == other.status_code and \ self.data == self.data __str__ = __repr__ @classmethod def from_status(cls, status_line, msg=None): """Returns a class method from bottle.HTTPError.status_line attribute. Useful for patching `bottle.HTTPError` for web services. Args: status_line (str): bottle.HTTPError.status_line text. msg: The message data for response. Returns: Class method based on status_line arg. Examples: >>> status_line = '401 Unauthorized' >>> error_msg = 'Get out!' >>> resp = WSResponse.from_status(status_line, error_msg) >>> resp['errors'] ['Get out!'] >>> resp['status_text'] 'Unauthorized' """ method = getattr(cls, status_line.lower()[4:].replace(' ', '_')) return method(msg) @classmethod def ok(cls, data): """Shortcut API for HTTP 200 `OK` response. Args: data (object): Response key/value data. Returns WSResponse Instance. """ return cls( status_code=200, data=data ).to_json @classmethod def created(cls, data=None): """Shortcut API for HTTP 201 `Created` response. Args: data (object): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '201 Created' return cls(201, data=data).to_json @classmethod def not_modified(cls, errors=None): """Shortcut API for HTTP 304 `Not Modified` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '304 Not Modified' return cls(304, None, errors).to_json @classmethod def bad_request(cls, errors=None): """Shortcut API for HTTP 400 `Bad Request` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '400 Bad Request' return cls(400, errors=errors).to_json @classmethod def unauthorized(cls, errors=None): """Shortcut API for HTTP 401 `Unauthorized` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '401 Unauthorized' return cls(401, errors=errors).to_json @classmethod @classmethod def not_found(cls, errors=None): """Shortcut API for HTTP 404 `Not found` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '404 Not Found' return cls(404, None, errors).to_json @classmethod def method_not_allowed(cls, errors=None): """Shortcut API for HTTP 405 `Method not allowed` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '405 Method Not Allowed' return cls(405, None, errors).to_json @classmethod def not_implemented(cls, errors=None): """Shortcut API for HTTP 501 `Not Implemented` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '501 Not Implemented' return cls(501, None, errors).to_json @classmethod def service_unavailable(cls, errors=None): """Shortcut API for HTTP 503 `Service Unavailable` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '503 Service Unavailable' return cls(503, None, errors).to_json @property def to_json(self): """Short cut for JSON response service data. Returns: Dict that implements JSON interface. """ web_resp = collections.OrderedDict() web_resp['status_code'] = self.status_code web_resp['status_text'] = dict(HTTP_CODES).get(self.status_code) web_resp['data'] = self.data if self.data is not None else {} web_resp['errors'] = self.errors or [] return web_resp
agile4you/bottle-neck
bottle_neck/response.py
WSResponse.not_found
python
def not_found(cls, errors=None): if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '404 Not Found' return cls(404, None, errors).to_json
Shortcut API for HTTP 404 `Not found` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance.
train
https://github.com/agile4you/bottle-neck/blob/ebc670a4b178255473d68e9b4122ba04e38f4810/bottle_neck/response.py#L230-L243
null
class WSResponse(object): """Base web service response class. WSResponse class provides a unified API for HTTP responses. The Response body skeleton looks like this:: { "status_code": 200, "status_txt": "OK", "response_data": { "id": 65234 "username": "pav" "email": "pav@geotagaeroview.com" }, "errors": [] } Attributes: status_code (int): Response HTTP status code. data (object): Response key/value data. errors (dict): Response key/value errors. Examples: >>> response = WSResponse( ... status_code=200, ... data='Hi' ... ) ... >>> response.to_json OrderedDict([('status_code', 200), ('status_text', 'OK'), ('data', 'Hi'), ('errors', [])]) """ __slots__ = ['status_code', 'data', 'errors'] expose_status = True response = bottle.response def __init__(self, status_code=200, data=None, errors=None): if status_code not in dict(HTTP_CODES) or\ not isinstance(errors, six.string_types + (list, tuple, type(None),)): raise WSResponseError('Invalid Response initialization.') self.status_code = status_code self.data = data or {} if isinstance(errors, (six.string_types, )): errors = [errors] self.errors = errors def __repr__(self): # pragma: no cover return "WebService Response: status={}, data={}".format( self.status_code, str(self.data) ) def __eq__(self, other): # pragma: no cover assert isinstance(other, WSResponse), 'Invalid Type for eq operator.' return self.status_code == other.status_code and \ self.data == self.data __str__ = __repr__ @classmethod def from_status(cls, status_line, msg=None): """Returns a class method from bottle.HTTPError.status_line attribute. Useful for patching `bottle.HTTPError` for web services. Args: status_line (str): bottle.HTTPError.status_line text. msg: The message data for response. Returns: Class method based on status_line arg. Examples: >>> status_line = '401 Unauthorized' >>> error_msg = 'Get out!' >>> resp = WSResponse.from_status(status_line, error_msg) >>> resp['errors'] ['Get out!'] >>> resp['status_text'] 'Unauthorized' """ method = getattr(cls, status_line.lower()[4:].replace(' ', '_')) return method(msg) @classmethod def ok(cls, data): """Shortcut API for HTTP 200 `OK` response. Args: data (object): Response key/value data. Returns WSResponse Instance. """ return cls( status_code=200, data=data ).to_json @classmethod def created(cls, data=None): """Shortcut API for HTTP 201 `Created` response. Args: data (object): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '201 Created' return cls(201, data=data).to_json @classmethod def not_modified(cls, errors=None): """Shortcut API for HTTP 304 `Not Modified` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '304 Not Modified' return cls(304, None, errors).to_json @classmethod def bad_request(cls, errors=None): """Shortcut API for HTTP 400 `Bad Request` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '400 Bad Request' return cls(400, errors=errors).to_json @classmethod def unauthorized(cls, errors=None): """Shortcut API for HTTP 401 `Unauthorized` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '401 Unauthorized' return cls(401, errors=errors).to_json @classmethod def forbidden(cls, errors=None): """Shortcut API for HTTP 403 `Forbidden` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '403 Forbidden' return cls(403, errors=errors).to_json @classmethod @classmethod def method_not_allowed(cls, errors=None): """Shortcut API for HTTP 405 `Method not allowed` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '405 Method Not Allowed' return cls(405, None, errors).to_json @classmethod def not_implemented(cls, errors=None): """Shortcut API for HTTP 501 `Not Implemented` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '501 Not Implemented' return cls(501, None, errors).to_json @classmethod def service_unavailable(cls, errors=None): """Shortcut API for HTTP 503 `Service Unavailable` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '503 Service Unavailable' return cls(503, None, errors).to_json @property def to_json(self): """Short cut for JSON response service data. Returns: Dict that implements JSON interface. """ web_resp = collections.OrderedDict() web_resp['status_code'] = self.status_code web_resp['status_text'] = dict(HTTP_CODES).get(self.status_code) web_resp['data'] = self.data if self.data is not None else {} web_resp['errors'] = self.errors or [] return web_resp
agile4you/bottle-neck
bottle_neck/response.py
WSResponse.method_not_allowed
python
def method_not_allowed(cls, errors=None): if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '405 Method Not Allowed' return cls(405, None, errors).to_json
Shortcut API for HTTP 405 `Method not allowed` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance.
train
https://github.com/agile4you/bottle-neck/blob/ebc670a4b178255473d68e9b4122ba04e38f4810/bottle_neck/response.py#L246-L259
null
class WSResponse(object): """Base web service response class. WSResponse class provides a unified API for HTTP responses. The Response body skeleton looks like this:: { "status_code": 200, "status_txt": "OK", "response_data": { "id": 65234 "username": "pav" "email": "pav@geotagaeroview.com" }, "errors": [] } Attributes: status_code (int): Response HTTP status code. data (object): Response key/value data. errors (dict): Response key/value errors. Examples: >>> response = WSResponse( ... status_code=200, ... data='Hi' ... ) ... >>> response.to_json OrderedDict([('status_code', 200), ('status_text', 'OK'), ('data', 'Hi'), ('errors', [])]) """ __slots__ = ['status_code', 'data', 'errors'] expose_status = True response = bottle.response def __init__(self, status_code=200, data=None, errors=None): if status_code not in dict(HTTP_CODES) or\ not isinstance(errors, six.string_types + (list, tuple, type(None),)): raise WSResponseError('Invalid Response initialization.') self.status_code = status_code self.data = data or {} if isinstance(errors, (six.string_types, )): errors = [errors] self.errors = errors def __repr__(self): # pragma: no cover return "WebService Response: status={}, data={}".format( self.status_code, str(self.data) ) def __eq__(self, other): # pragma: no cover assert isinstance(other, WSResponse), 'Invalid Type for eq operator.' return self.status_code == other.status_code and \ self.data == self.data __str__ = __repr__ @classmethod def from_status(cls, status_line, msg=None): """Returns a class method from bottle.HTTPError.status_line attribute. Useful for patching `bottle.HTTPError` for web services. Args: status_line (str): bottle.HTTPError.status_line text. msg: The message data for response. Returns: Class method based on status_line arg. Examples: >>> status_line = '401 Unauthorized' >>> error_msg = 'Get out!' >>> resp = WSResponse.from_status(status_line, error_msg) >>> resp['errors'] ['Get out!'] >>> resp['status_text'] 'Unauthorized' """ method = getattr(cls, status_line.lower()[4:].replace(' ', '_')) return method(msg) @classmethod def ok(cls, data): """Shortcut API for HTTP 200 `OK` response. Args: data (object): Response key/value data. Returns WSResponse Instance. """ return cls( status_code=200, data=data ).to_json @classmethod def created(cls, data=None): """Shortcut API for HTTP 201 `Created` response. Args: data (object): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '201 Created' return cls(201, data=data).to_json @classmethod def not_modified(cls, errors=None): """Shortcut API for HTTP 304 `Not Modified` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '304 Not Modified' return cls(304, None, errors).to_json @classmethod def bad_request(cls, errors=None): """Shortcut API for HTTP 400 `Bad Request` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '400 Bad Request' return cls(400, errors=errors).to_json @classmethod def unauthorized(cls, errors=None): """Shortcut API for HTTP 401 `Unauthorized` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '401 Unauthorized' return cls(401, errors=errors).to_json @classmethod def forbidden(cls, errors=None): """Shortcut API for HTTP 403 `Forbidden` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '403 Forbidden' return cls(403, errors=errors).to_json @classmethod def not_found(cls, errors=None): """Shortcut API for HTTP 404 `Not found` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '404 Not Found' return cls(404, None, errors).to_json @classmethod @classmethod def not_implemented(cls, errors=None): """Shortcut API for HTTP 501 `Not Implemented` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '501 Not Implemented' return cls(501, None, errors).to_json @classmethod def service_unavailable(cls, errors=None): """Shortcut API for HTTP 503 `Service Unavailable` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '503 Service Unavailable' return cls(503, None, errors).to_json @property def to_json(self): """Short cut for JSON response service data. Returns: Dict that implements JSON interface. """ web_resp = collections.OrderedDict() web_resp['status_code'] = self.status_code web_resp['status_text'] = dict(HTTP_CODES).get(self.status_code) web_resp['data'] = self.data if self.data is not None else {} web_resp['errors'] = self.errors or [] return web_resp
agile4you/bottle-neck
bottle_neck/response.py
WSResponse.not_implemented
python
def not_implemented(cls, errors=None): if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '501 Not Implemented' return cls(501, None, errors).to_json
Shortcut API for HTTP 501 `Not Implemented` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance.
train
https://github.com/agile4you/bottle-neck/blob/ebc670a4b178255473d68e9b4122ba04e38f4810/bottle_neck/response.py#L262-L275
null
class WSResponse(object): """Base web service response class. WSResponse class provides a unified API for HTTP responses. The Response body skeleton looks like this:: { "status_code": 200, "status_txt": "OK", "response_data": { "id": 65234 "username": "pav" "email": "pav@geotagaeroview.com" }, "errors": [] } Attributes: status_code (int): Response HTTP status code. data (object): Response key/value data. errors (dict): Response key/value errors. Examples: >>> response = WSResponse( ... status_code=200, ... data='Hi' ... ) ... >>> response.to_json OrderedDict([('status_code', 200), ('status_text', 'OK'), ('data', 'Hi'), ('errors', [])]) """ __slots__ = ['status_code', 'data', 'errors'] expose_status = True response = bottle.response def __init__(self, status_code=200, data=None, errors=None): if status_code not in dict(HTTP_CODES) or\ not isinstance(errors, six.string_types + (list, tuple, type(None),)): raise WSResponseError('Invalid Response initialization.') self.status_code = status_code self.data = data or {} if isinstance(errors, (six.string_types, )): errors = [errors] self.errors = errors def __repr__(self): # pragma: no cover return "WebService Response: status={}, data={}".format( self.status_code, str(self.data) ) def __eq__(self, other): # pragma: no cover assert isinstance(other, WSResponse), 'Invalid Type for eq operator.' return self.status_code == other.status_code and \ self.data == self.data __str__ = __repr__ @classmethod def from_status(cls, status_line, msg=None): """Returns a class method from bottle.HTTPError.status_line attribute. Useful for patching `bottle.HTTPError` for web services. Args: status_line (str): bottle.HTTPError.status_line text. msg: The message data for response. Returns: Class method based on status_line arg. Examples: >>> status_line = '401 Unauthorized' >>> error_msg = 'Get out!' >>> resp = WSResponse.from_status(status_line, error_msg) >>> resp['errors'] ['Get out!'] >>> resp['status_text'] 'Unauthorized' """ method = getattr(cls, status_line.lower()[4:].replace(' ', '_')) return method(msg) @classmethod def ok(cls, data): """Shortcut API for HTTP 200 `OK` response. Args: data (object): Response key/value data. Returns WSResponse Instance. """ return cls( status_code=200, data=data ).to_json @classmethod def created(cls, data=None): """Shortcut API for HTTP 201 `Created` response. Args: data (object): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '201 Created' return cls(201, data=data).to_json @classmethod def not_modified(cls, errors=None): """Shortcut API for HTTP 304 `Not Modified` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '304 Not Modified' return cls(304, None, errors).to_json @classmethod def bad_request(cls, errors=None): """Shortcut API for HTTP 400 `Bad Request` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '400 Bad Request' return cls(400, errors=errors).to_json @classmethod def unauthorized(cls, errors=None): """Shortcut API for HTTP 401 `Unauthorized` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '401 Unauthorized' return cls(401, errors=errors).to_json @classmethod def forbidden(cls, errors=None): """Shortcut API for HTTP 403 `Forbidden` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '403 Forbidden' return cls(403, errors=errors).to_json @classmethod def not_found(cls, errors=None): """Shortcut API for HTTP 404 `Not found` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '404 Not Found' return cls(404, None, errors).to_json @classmethod def method_not_allowed(cls, errors=None): """Shortcut API for HTTP 405 `Method not allowed` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '405 Method Not Allowed' return cls(405, None, errors).to_json @classmethod @classmethod def service_unavailable(cls, errors=None): """Shortcut API for HTTP 503 `Service Unavailable` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '503 Service Unavailable' return cls(503, None, errors).to_json @property def to_json(self): """Short cut for JSON response service data. Returns: Dict that implements JSON interface. """ web_resp = collections.OrderedDict() web_resp['status_code'] = self.status_code web_resp['status_text'] = dict(HTTP_CODES).get(self.status_code) web_resp['data'] = self.data if self.data is not None else {} web_resp['errors'] = self.errors or [] return web_resp
agile4you/bottle-neck
bottle_neck/response.py
WSResponse.service_unavailable
python
def service_unavailable(cls, errors=None): if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '503 Service Unavailable' return cls(503, None, errors).to_json
Shortcut API for HTTP 503 `Service Unavailable` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance.
train
https://github.com/agile4you/bottle-neck/blob/ebc670a4b178255473d68e9b4122ba04e38f4810/bottle_neck/response.py#L278-L291
null
class WSResponse(object): """Base web service response class. WSResponse class provides a unified API for HTTP responses. The Response body skeleton looks like this:: { "status_code": 200, "status_txt": "OK", "response_data": { "id": 65234 "username": "pav" "email": "pav@geotagaeroview.com" }, "errors": [] } Attributes: status_code (int): Response HTTP status code. data (object): Response key/value data. errors (dict): Response key/value errors. Examples: >>> response = WSResponse( ... status_code=200, ... data='Hi' ... ) ... >>> response.to_json OrderedDict([('status_code', 200), ('status_text', 'OK'), ('data', 'Hi'), ('errors', [])]) """ __slots__ = ['status_code', 'data', 'errors'] expose_status = True response = bottle.response def __init__(self, status_code=200, data=None, errors=None): if status_code not in dict(HTTP_CODES) or\ not isinstance(errors, six.string_types + (list, tuple, type(None),)): raise WSResponseError('Invalid Response initialization.') self.status_code = status_code self.data = data or {} if isinstance(errors, (six.string_types, )): errors = [errors] self.errors = errors def __repr__(self): # pragma: no cover return "WebService Response: status={}, data={}".format( self.status_code, str(self.data) ) def __eq__(self, other): # pragma: no cover assert isinstance(other, WSResponse), 'Invalid Type for eq operator.' return self.status_code == other.status_code and \ self.data == self.data __str__ = __repr__ @classmethod def from_status(cls, status_line, msg=None): """Returns a class method from bottle.HTTPError.status_line attribute. Useful for patching `bottle.HTTPError` for web services. Args: status_line (str): bottle.HTTPError.status_line text. msg: The message data for response. Returns: Class method based on status_line arg. Examples: >>> status_line = '401 Unauthorized' >>> error_msg = 'Get out!' >>> resp = WSResponse.from_status(status_line, error_msg) >>> resp['errors'] ['Get out!'] >>> resp['status_text'] 'Unauthorized' """ method = getattr(cls, status_line.lower()[4:].replace(' ', '_')) return method(msg) @classmethod def ok(cls, data): """Shortcut API for HTTP 200 `OK` response. Args: data (object): Response key/value data. Returns WSResponse Instance. """ return cls( status_code=200, data=data ).to_json @classmethod def created(cls, data=None): """Shortcut API for HTTP 201 `Created` response. Args: data (object): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '201 Created' return cls(201, data=data).to_json @classmethod def not_modified(cls, errors=None): """Shortcut API for HTTP 304 `Not Modified` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '304 Not Modified' return cls(304, None, errors).to_json @classmethod def bad_request(cls, errors=None): """Shortcut API for HTTP 400 `Bad Request` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '400 Bad Request' return cls(400, errors=errors).to_json @classmethod def unauthorized(cls, errors=None): """Shortcut API for HTTP 401 `Unauthorized` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '401 Unauthorized' return cls(401, errors=errors).to_json @classmethod def forbidden(cls, errors=None): """Shortcut API for HTTP 403 `Forbidden` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '403 Forbidden' return cls(403, errors=errors).to_json @classmethod def not_found(cls, errors=None): """Shortcut API for HTTP 404 `Not found` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '404 Not Found' return cls(404, None, errors).to_json @classmethod def method_not_allowed(cls, errors=None): """Shortcut API for HTTP 405 `Method not allowed` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '405 Method Not Allowed' return cls(405, None, errors).to_json @classmethod def not_implemented(cls, errors=None): """Shortcut API for HTTP 501 `Not Implemented` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '501 Not Implemented' return cls(501, None, errors).to_json @classmethod @property def to_json(self): """Short cut for JSON response service data. Returns: Dict that implements JSON interface. """ web_resp = collections.OrderedDict() web_resp['status_code'] = self.status_code web_resp['status_text'] = dict(HTTP_CODES).get(self.status_code) web_resp['data'] = self.data if self.data is not None else {} web_resp['errors'] = self.errors or [] return web_resp
agile4you/bottle-neck
bottle_neck/response.py
WSResponse.to_json
python
def to_json(self): web_resp = collections.OrderedDict() web_resp['status_code'] = self.status_code web_resp['status_text'] = dict(HTTP_CODES).get(self.status_code) web_resp['data'] = self.data if self.data is not None else {} web_resp['errors'] = self.errors or [] return web_resp
Short cut for JSON response service data. Returns: Dict that implements JSON interface.
train
https://github.com/agile4you/bottle-neck/blob/ebc670a4b178255473d68e9b4122ba04e38f4810/bottle_neck/response.py#L294-L308
null
class WSResponse(object): """Base web service response class. WSResponse class provides a unified API for HTTP responses. The Response body skeleton looks like this:: { "status_code": 200, "status_txt": "OK", "response_data": { "id": 65234 "username": "pav" "email": "pav@geotagaeroview.com" }, "errors": [] } Attributes: status_code (int): Response HTTP status code. data (object): Response key/value data. errors (dict): Response key/value errors. Examples: >>> response = WSResponse( ... status_code=200, ... data='Hi' ... ) ... >>> response.to_json OrderedDict([('status_code', 200), ('status_text', 'OK'), ('data', 'Hi'), ('errors', [])]) """ __slots__ = ['status_code', 'data', 'errors'] expose_status = True response = bottle.response def __init__(self, status_code=200, data=None, errors=None): if status_code not in dict(HTTP_CODES) or\ not isinstance(errors, six.string_types + (list, tuple, type(None),)): raise WSResponseError('Invalid Response initialization.') self.status_code = status_code self.data = data or {} if isinstance(errors, (six.string_types, )): errors = [errors] self.errors = errors def __repr__(self): # pragma: no cover return "WebService Response: status={}, data={}".format( self.status_code, str(self.data) ) def __eq__(self, other): # pragma: no cover assert isinstance(other, WSResponse), 'Invalid Type for eq operator.' return self.status_code == other.status_code and \ self.data == self.data __str__ = __repr__ @classmethod def from_status(cls, status_line, msg=None): """Returns a class method from bottle.HTTPError.status_line attribute. Useful for patching `bottle.HTTPError` for web services. Args: status_line (str): bottle.HTTPError.status_line text. msg: The message data for response. Returns: Class method based on status_line arg. Examples: >>> status_line = '401 Unauthorized' >>> error_msg = 'Get out!' >>> resp = WSResponse.from_status(status_line, error_msg) >>> resp['errors'] ['Get out!'] >>> resp['status_text'] 'Unauthorized' """ method = getattr(cls, status_line.lower()[4:].replace(' ', '_')) return method(msg) @classmethod def ok(cls, data): """Shortcut API for HTTP 200 `OK` response. Args: data (object): Response key/value data. Returns WSResponse Instance. """ return cls( status_code=200, data=data ).to_json @classmethod def created(cls, data=None): """Shortcut API for HTTP 201 `Created` response. Args: data (object): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '201 Created' return cls(201, data=data).to_json @classmethod def not_modified(cls, errors=None): """Shortcut API for HTTP 304 `Not Modified` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '304 Not Modified' return cls(304, None, errors).to_json @classmethod def bad_request(cls, errors=None): """Shortcut API for HTTP 400 `Bad Request` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '400 Bad Request' return cls(400, errors=errors).to_json @classmethod def unauthorized(cls, errors=None): """Shortcut API for HTTP 401 `Unauthorized` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '401 Unauthorized' return cls(401, errors=errors).to_json @classmethod def forbidden(cls, errors=None): """Shortcut API for HTTP 403 `Forbidden` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '403 Forbidden' return cls(403, errors=errors).to_json @classmethod def not_found(cls, errors=None): """Shortcut API for HTTP 404 `Not found` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '404 Not Found' return cls(404, None, errors).to_json @classmethod def method_not_allowed(cls, errors=None): """Shortcut API for HTTP 405 `Method not allowed` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '405 Method Not Allowed' return cls(405, None, errors).to_json @classmethod def not_implemented(cls, errors=None): """Shortcut API for HTTP 501 `Not Implemented` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '501 Not Implemented' return cls(501, None, errors).to_json @classmethod def service_unavailable(cls, errors=None): """Shortcut API for HTTP 503 `Service Unavailable` response. Args: errors (list): Response key/value data. Returns: WSResponse Instance. """ if cls.expose_status: # pragma: no cover cls.response.content_type = 'application/json' cls.response._status_line = '503 Service Unavailable' return cls(503, None, errors).to_json @property
agile4you/bottle-neck
bottle_neck/webapi.py
paginator
python
def paginator(limit, offset, record_count, base_uri, page_nav_tpl='&limit={}&offset={}'): total_pages = int(math.ceil(record_count / limit)) next_cond = limit + offset <= record_count prev_cond = offset >= limit next_page = base_uri + page_nav_tpl.format(limit, offset + limit) if next_cond else None prev_page = base_uri + page_nav_tpl.format(limit, offset - limit) if prev_cond else None return OrderedDict([ ('total_count', record_count), ('total_pages', total_pages), ('next_page', next_page), ('prev_page', prev_page) ])
Compute pagination info for collection filtering. Args: limit (int): Collection filter limit. offset (int): Collection filter offset. record_count (int): Collection filter total record count. base_uri (str): Collection filter base uri (without limit, offset) page_nav_tpl (str): Pagination template. Returns: A mapping of pagination info.
train
https://github.com/agile4you/bottle-neck/blob/ebc670a4b178255473d68e9b4122ba04e38f4810/bottle_neck/webapi.py#L36-L64
null
# -*- coding: utf-8 -*- """Bottle.py API development utilities. Provides some extra functionality for developing web API's with `bottle.py`. """ from __future__ import absolute_import __author__ = 'pav' __date__ = '2015-2-3' __all__ = ['cors_enable_hook', 'strip_path_hook', 'paginator'] from bottle_neck import __version__ from collections import OrderedDict import bottle import math version = tuple(map(int, __version__.split('.'))) def cors_enable_hook(): bottle.response.headers['Access-Control-Allow-Origin'] = '*' bottle.response.headers['Access-Control-Allow-Headers'] = \ 'Authorization, Credentials, X-Requested-With, Content-Type' bottle.response.headers['Access-Control-Allow-Methods'] = \ 'GET, PUT, POST, OPTIONS, DELETE' def strip_path_hook(): """Ignore trailing slashes. """ bottle.request.environ['PATH_INFO'] = \ bottle.request.environ['PATH_INFO'].rstrip('/')
agile4you/bottle-neck
bottle_neck/cbv.py
cached_classproperty
python
def cached_classproperty(fun): @functools.wraps(fun) def get(cls): try: return cls.__cache[fun] except AttributeError: cls.__cache = {} except KeyError: # pragma: no cover pass ret = cls.__cache[fun] = fun(cls) return ret return classproperty(get)
A memorization decorator for class properties. It implements the above `classproperty` decorator, with the difference that the function result is computed and attached to class as direct attribute. (Lazy loading and caching.)
train
https://github.com/agile4you/bottle-neck/blob/ebc670a4b178255473d68e9b4122ba04e38f4810/bottle_neck/cbv.py#L92-L109
[ "def classproperty(func):\n \"\"\"classproperty decorator.\n Using this decorator a class can have a property. Necessary for properties\n that don't need instance initialization. Works exactly the same as a\n normal property.\n\n Examples:\n >>> class MyClass(object):\n ... @classpr...
# -*- coding: utf-8 -*- """CBV for bottle.py application instances. Provides a base class for creating class-based web handlers for bottle.py application instances with application routing mechanism. """ from __future__ import absolute_import __author__ = "Papavassiliou Vassilis" __date__ = "2015-11-29" __all__ = ['BaseHandler', 'HandlerMeta', 'route_method', 'plugin_method', 'HandlerError', 'HandlerHTTPMethodError', 'HandlerPluginError', 'BaseHandlerPlugin'] from bottle_neck import __version__ import bottle import functools import inspect import six import re version = tuple(map(int, __version__.split('.'))) DEFAULT_ROUTES = ("get", "put", "post", "delete", "patch", "options") CORS_ROUTES = ("put", "post", "delete", "patch") PLUGIN_SCOPE = ( (False, 'plugins'), (True, 'global_plugins') ) class HandlerError(Exception): """Base module Exception class. """ pass class HandlerHTTPMethodError(HandlerError): """Raises for invalid HTTP method declaration. """ pass class HandlerPluginError(HandlerError): """Raises when a handler plugin error occurs. """ pass class ClassPropertyDescriptor(object): """ClassProperty Descriptor class. Straight up stolen from stack overflow Implements class level property non-data descriptor. """ def __init__(self, fget, fset=None): self.fget = fget self.fset = fset def __get__(self, obj, cls=None): if cls is None: # pragma: no cover cls = type(obj) return self.fget.__get__(obj, cls)() def classproperty(func): """classproperty decorator. Using this decorator a class can have a property. Necessary for properties that don't need instance initialization. Works exactly the same as a normal property. Examples: >>> class MyClass(object): ... @classproperty ... def my_prop(self): ... return self.__name__ + ' class' ... >>> MyClass.my_prop 'MyClass class' """ if not isinstance(func, (classmethod, staticmethod)): func = classmethod(func) return ClassPropertyDescriptor(func) def plugin_method(*plugin_names): """Plugin Method decorator. Signs a web handler function with the plugins to be applied as attributes. Args: plugin_names (list): A list of plugin callable names Returns: A wrapped handler callable. Examples: >>> @plugin_method('json', 'bill') ... def method(): ... return "Hello!" ... >>> print method.json True >>> print method.bill True """ def wrapper(callable_obj): for plugin_name in plugin_names: if not hasattr(callable_obj, plugin_name): setattr(callable_obj, plugin_name, True) return callable_obj return wrapper def route_method(method_name, extra_part=False): """Custom handler routing decorator. Signs a web handler callable with the http method as attribute. Args: method_name (str): HTTP method name (i.e GET, POST) extra_part (bool): Indicates if wrapped callable name should be a part of the actual endpoint. Returns: A wrapped handler callable. examples: >>> @route_method('GET') ... def method(): ... return "Hello!" ... >>> method.http_method 'GET' >>> method.url_extra_part None """ def wrapper(callable_obj): if method_name.lower() not in DEFAULT_ROUTES: raise HandlerHTTPMethodError( 'Invalid http method in method: {}'.format(method_name) ) callable_obj.http_method = method_name.upper() callable_obj.url_extra_part = callable_obj.__name__ if extra_part\ else None return classmethod(callable_obj) return wrapper class BaseHandlerPlugin(object): """ """ def __init__(self, callable_object, *args, **kwargs): # pragma: no cover self._wrapped = callable_object self._args = args self._kwargs = kwargs self.__doc__ = callable_object.__doc__ @cached_classproperty def func_name(self): cls_name = re.sub('(.)([A-Z][a-z]+)', r'\1_\2', self.__name__) return re.sub('([a-z0-9])([A-Z])', r'\1_\2', cls_name).lower() def apply(self, *args, **kwargs): # pragma: no cover raise NotImplementedError("Must Override `apply` method") def __call__(self, *args, **kwargs): # pragma: no cover return self.apply(*args, **kwargs) class HandlerMeta(type): """BaseHandler metaclass Provides meta functionality for ensuring that `http_method` don't require an instance and test and that derived classes implements `Singleton`. """ _instances = {} def __new__(mcs, name, bases, attrs): http_methods = [attr for attr in attrs if attr.lower() in DEFAULT_ROUTES] for handler in http_methods: attrs[handler] = classmethod(attrs[handler]) return super(HandlerMeta, mcs).__new__(mcs, name, bases, attrs) def __call__(cls, *args): """Enforce Singleton class initiation """ if cls not in cls._instances: cls._instances[cls] = super( HandlerMeta, cls ).__call__(*args) return cls._instances[cls] @six.add_metaclass(HandlerMeta) class BaseHandler(object): """Base Handler class for implementing Class-Based Web handler for bottle. Subclass `BaseHandler` in order to its API for application routing, and implement any of the known http method: Class Attributes: render (object): bottle.jinja2_render. request (object): bottle.request response (object): bottle.response redirect (object):bottle.redirect abort (object): bottle.abort base_endpoint (str): The handler endpoint prefix. cors_enabled (bool): Indicates if CORS is enabled. plugins (dict): A key/value mapping of available plugins. global_plugins (dict): A key/value mapping default applying plugins. """ render = bottle.jinja2_template request = bottle.request response = bottle.response redirect = bottle.redirect abort = bottle.abort response_factory = None base_endpoint = '/' cors_enabled = True plugins = dict() global_plugins = dict() @classmethod def add_plugin(cls, plugin_callables, global_scope=False): """Register a new plugin in handler. Args: plugin_callables (list): A list of plugin callables. global_scope (bool): Indicates The scope of the plugin. Returns: Class instance. """ repo = getattr(cls, dict(PLUGIN_SCOPE)[global_scope]) for plugin_callable in plugin_callables: if hasattr(plugin_callable, 'func_name'): repo[plugin_callable.func_name] = plugin_callable else: # pragma: no cover repo[plugin_callable.__name__] = plugin_callable return cls @classmethod def register_app(cls, application): """Register class view in bottle application. Args: application (instance): A bottle.Bottle() instance. Returns: Class instance. """ if cls is BaseHandler: raise HandlerError("Cant register a `BaseHandler` class instance") routes = cls._get_http_members() router = getattr(application, "route") for func_name, func_callable in routes: method_args = inspect.signature(func_callable).parameters if hasattr(func_callable, 'http_method'): http_method = func_callable.http_method url_extra_part = func_callable.url_extra_part else: http_method = func_name url_extra_part = None applied_plugins = [pl for pl in cls.plugins if hasattr(func_callable, pl)] for plugin in applied_plugins: # pragma: no cover try: func_callable = cls.plugins[plugin](func_callable) except TypeError as error: raise HandlerPluginError(error.args) for global_plugin in cls.global_plugins: func_callable = cls.global_plugins[global_plugin]( func_callable ) for entrypoint in cls._build_routes(method_args, url_extra_part): router(entrypoint, method=[http_method.upper()])(func_callable) if cls.cors_enabled: cls_desc = cls.__doc__ or '' options_data = { "handler": {"name": cls.__name__, "desc": cls_desc.strip()}, "http_methods": [r[0] for r in routes] } router( cls.base_endpoint, method=["OPTIONS"] )(lambda: options_data) router( "{}/<url:re:.+>".format(cls.base_endpoint).replace("//", "/"), method=["OPTIONS"] )(lambda url: options_data) return cls @classmethod def _build_routes(cls, method_args, url_extra_part=None): """Create bottle route for a handler http method.""" prefix = '/{}'.format(url_extra_part) if url_extra_part else '' endpoint = cls.base_endpoint + prefix if not method_args: return [endpoint] endpoints = [] for args_list in cls._router_helper(method_args): prefix = '/:' if args_list else '' endpoints.append((endpoint + prefix + '/:'.join(args_list)).replace("//", '/')) return endpoints @classmethod def _router_helper(cls, method_args): """Detect default Nullable method arguments and return multiple routes per callable. """ fixed_params = [param for param in method_args if method_args[param].default is not None] nullable_params = [param for param in method_args if param not in fixed_params] combinations = [fixed_params] combinations += [combinations[-1] + [param] for param in nullable_params] return combinations @classmethod def _get_http_members(cls): """Filter all `http` specific method from handler class. """ return [member for member in inspect.getmembers(cls, predicate=inspect.ismethod) if cls._check_http_member(member)] @staticmethod def _check_http_member(member): """Checks if a class method has HTTP info. """ return member[0].lower() in DEFAULT_ROUTES or \ hasattr(member[1], 'http_method') if __name__ == '__main__': # pragma: no cover import doctest doctest.testmod()
agile4you/bottle-neck
bottle_neck/cbv.py
plugin_method
python
def plugin_method(*plugin_names): def wrapper(callable_obj): for plugin_name in plugin_names: if not hasattr(callable_obj, plugin_name): setattr(callable_obj, plugin_name, True) return callable_obj return wrapper
Plugin Method decorator. Signs a web handler function with the plugins to be applied as attributes. Args: plugin_names (list): A list of plugin callable names Returns: A wrapped handler callable. Examples: >>> @plugin_method('json', 'bill') ... def method(): ... return "Hello!" ... >>> print method.json True >>> print method.bill True
train
https://github.com/agile4you/bottle-neck/blob/ebc670a4b178255473d68e9b4122ba04e38f4810/bottle_neck/cbv.py#L112-L138
null
# -*- coding: utf-8 -*- """CBV for bottle.py application instances. Provides a base class for creating class-based web handlers for bottle.py application instances with application routing mechanism. """ from __future__ import absolute_import __author__ = "Papavassiliou Vassilis" __date__ = "2015-11-29" __all__ = ['BaseHandler', 'HandlerMeta', 'route_method', 'plugin_method', 'HandlerError', 'HandlerHTTPMethodError', 'HandlerPluginError', 'BaseHandlerPlugin'] from bottle_neck import __version__ import bottle import functools import inspect import six import re version = tuple(map(int, __version__.split('.'))) DEFAULT_ROUTES = ("get", "put", "post", "delete", "patch", "options") CORS_ROUTES = ("put", "post", "delete", "patch") PLUGIN_SCOPE = ( (False, 'plugins'), (True, 'global_plugins') ) class HandlerError(Exception): """Base module Exception class. """ pass class HandlerHTTPMethodError(HandlerError): """Raises for invalid HTTP method declaration. """ pass class HandlerPluginError(HandlerError): """Raises when a handler plugin error occurs. """ pass class ClassPropertyDescriptor(object): """ClassProperty Descriptor class. Straight up stolen from stack overflow Implements class level property non-data descriptor. """ def __init__(self, fget, fset=None): self.fget = fget self.fset = fset def __get__(self, obj, cls=None): if cls is None: # pragma: no cover cls = type(obj) return self.fget.__get__(obj, cls)() def classproperty(func): """classproperty decorator. Using this decorator a class can have a property. Necessary for properties that don't need instance initialization. Works exactly the same as a normal property. Examples: >>> class MyClass(object): ... @classproperty ... def my_prop(self): ... return self.__name__ + ' class' ... >>> MyClass.my_prop 'MyClass class' """ if not isinstance(func, (classmethod, staticmethod)): func = classmethod(func) return ClassPropertyDescriptor(func) def cached_classproperty(fun): """A memorization decorator for class properties. It implements the above `classproperty` decorator, with the difference that the function result is computed and attached to class as direct attribute. (Lazy loading and caching.) """ @functools.wraps(fun) def get(cls): try: return cls.__cache[fun] except AttributeError: cls.__cache = {} except KeyError: # pragma: no cover pass ret = cls.__cache[fun] = fun(cls) return ret return classproperty(get) def route_method(method_name, extra_part=False): """Custom handler routing decorator. Signs a web handler callable with the http method as attribute. Args: method_name (str): HTTP method name (i.e GET, POST) extra_part (bool): Indicates if wrapped callable name should be a part of the actual endpoint. Returns: A wrapped handler callable. examples: >>> @route_method('GET') ... def method(): ... return "Hello!" ... >>> method.http_method 'GET' >>> method.url_extra_part None """ def wrapper(callable_obj): if method_name.lower() not in DEFAULT_ROUTES: raise HandlerHTTPMethodError( 'Invalid http method in method: {}'.format(method_name) ) callable_obj.http_method = method_name.upper() callable_obj.url_extra_part = callable_obj.__name__ if extra_part\ else None return classmethod(callable_obj) return wrapper class BaseHandlerPlugin(object): """ """ def __init__(self, callable_object, *args, **kwargs): # pragma: no cover self._wrapped = callable_object self._args = args self._kwargs = kwargs self.__doc__ = callable_object.__doc__ @cached_classproperty def func_name(self): cls_name = re.sub('(.)([A-Z][a-z]+)', r'\1_\2', self.__name__) return re.sub('([a-z0-9])([A-Z])', r'\1_\2', cls_name).lower() def apply(self, *args, **kwargs): # pragma: no cover raise NotImplementedError("Must Override `apply` method") def __call__(self, *args, **kwargs): # pragma: no cover return self.apply(*args, **kwargs) class HandlerMeta(type): """BaseHandler metaclass Provides meta functionality for ensuring that `http_method` don't require an instance and test and that derived classes implements `Singleton`. """ _instances = {} def __new__(mcs, name, bases, attrs): http_methods = [attr for attr in attrs if attr.lower() in DEFAULT_ROUTES] for handler in http_methods: attrs[handler] = classmethod(attrs[handler]) return super(HandlerMeta, mcs).__new__(mcs, name, bases, attrs) def __call__(cls, *args): """Enforce Singleton class initiation """ if cls not in cls._instances: cls._instances[cls] = super( HandlerMeta, cls ).__call__(*args) return cls._instances[cls] @six.add_metaclass(HandlerMeta) class BaseHandler(object): """Base Handler class for implementing Class-Based Web handler for bottle. Subclass `BaseHandler` in order to its API for application routing, and implement any of the known http method: Class Attributes: render (object): bottle.jinja2_render. request (object): bottle.request response (object): bottle.response redirect (object):bottle.redirect abort (object): bottle.abort base_endpoint (str): The handler endpoint prefix. cors_enabled (bool): Indicates if CORS is enabled. plugins (dict): A key/value mapping of available plugins. global_plugins (dict): A key/value mapping default applying plugins. """ render = bottle.jinja2_template request = bottle.request response = bottle.response redirect = bottle.redirect abort = bottle.abort response_factory = None base_endpoint = '/' cors_enabled = True plugins = dict() global_plugins = dict() @classmethod def add_plugin(cls, plugin_callables, global_scope=False): """Register a new plugin in handler. Args: plugin_callables (list): A list of plugin callables. global_scope (bool): Indicates The scope of the plugin. Returns: Class instance. """ repo = getattr(cls, dict(PLUGIN_SCOPE)[global_scope]) for plugin_callable in plugin_callables: if hasattr(plugin_callable, 'func_name'): repo[plugin_callable.func_name] = plugin_callable else: # pragma: no cover repo[plugin_callable.__name__] = plugin_callable return cls @classmethod def register_app(cls, application): """Register class view in bottle application. Args: application (instance): A bottle.Bottle() instance. Returns: Class instance. """ if cls is BaseHandler: raise HandlerError("Cant register a `BaseHandler` class instance") routes = cls._get_http_members() router = getattr(application, "route") for func_name, func_callable in routes: method_args = inspect.signature(func_callable).parameters if hasattr(func_callable, 'http_method'): http_method = func_callable.http_method url_extra_part = func_callable.url_extra_part else: http_method = func_name url_extra_part = None applied_plugins = [pl for pl in cls.plugins if hasattr(func_callable, pl)] for plugin in applied_plugins: # pragma: no cover try: func_callable = cls.plugins[plugin](func_callable) except TypeError as error: raise HandlerPluginError(error.args) for global_plugin in cls.global_plugins: func_callable = cls.global_plugins[global_plugin]( func_callable ) for entrypoint in cls._build_routes(method_args, url_extra_part): router(entrypoint, method=[http_method.upper()])(func_callable) if cls.cors_enabled: cls_desc = cls.__doc__ or '' options_data = { "handler": {"name": cls.__name__, "desc": cls_desc.strip()}, "http_methods": [r[0] for r in routes] } router( cls.base_endpoint, method=["OPTIONS"] )(lambda: options_data) router( "{}/<url:re:.+>".format(cls.base_endpoint).replace("//", "/"), method=["OPTIONS"] )(lambda url: options_data) return cls @classmethod def _build_routes(cls, method_args, url_extra_part=None): """Create bottle route for a handler http method.""" prefix = '/{}'.format(url_extra_part) if url_extra_part else '' endpoint = cls.base_endpoint + prefix if not method_args: return [endpoint] endpoints = [] for args_list in cls._router_helper(method_args): prefix = '/:' if args_list else '' endpoints.append((endpoint + prefix + '/:'.join(args_list)).replace("//", '/')) return endpoints @classmethod def _router_helper(cls, method_args): """Detect default Nullable method arguments and return multiple routes per callable. """ fixed_params = [param for param in method_args if method_args[param].default is not None] nullable_params = [param for param in method_args if param not in fixed_params] combinations = [fixed_params] combinations += [combinations[-1] + [param] for param in nullable_params] return combinations @classmethod def _get_http_members(cls): """Filter all `http` specific method from handler class. """ return [member for member in inspect.getmembers(cls, predicate=inspect.ismethod) if cls._check_http_member(member)] @staticmethod def _check_http_member(member): """Checks if a class method has HTTP info. """ return member[0].lower() in DEFAULT_ROUTES or \ hasattr(member[1], 'http_method') if __name__ == '__main__': # pragma: no cover import doctest doctest.testmod()
agile4you/bottle-neck
bottle_neck/cbv.py
route_method
python
def route_method(method_name, extra_part=False): def wrapper(callable_obj): if method_name.lower() not in DEFAULT_ROUTES: raise HandlerHTTPMethodError( 'Invalid http method in method: {}'.format(method_name) ) callable_obj.http_method = method_name.upper() callable_obj.url_extra_part = callable_obj.__name__ if extra_part\ else None return classmethod(callable_obj) return wrapper
Custom handler routing decorator. Signs a web handler callable with the http method as attribute. Args: method_name (str): HTTP method name (i.e GET, POST) extra_part (bool): Indicates if wrapped callable name should be a part of the actual endpoint. Returns: A wrapped handler callable. examples: >>> @route_method('GET') ... def method(): ... return "Hello!" ... >>> method.http_method 'GET' >>> method.url_extra_part None
train
https://github.com/agile4you/bottle-neck/blob/ebc670a4b178255473d68e9b4122ba04e38f4810/bottle_neck/cbv.py#L141-L175
null
# -*- coding: utf-8 -*- """CBV for bottle.py application instances. Provides a base class for creating class-based web handlers for bottle.py application instances with application routing mechanism. """ from __future__ import absolute_import __author__ = "Papavassiliou Vassilis" __date__ = "2015-11-29" __all__ = ['BaseHandler', 'HandlerMeta', 'route_method', 'plugin_method', 'HandlerError', 'HandlerHTTPMethodError', 'HandlerPluginError', 'BaseHandlerPlugin'] from bottle_neck import __version__ import bottle import functools import inspect import six import re version = tuple(map(int, __version__.split('.'))) DEFAULT_ROUTES = ("get", "put", "post", "delete", "patch", "options") CORS_ROUTES = ("put", "post", "delete", "patch") PLUGIN_SCOPE = ( (False, 'plugins'), (True, 'global_plugins') ) class HandlerError(Exception): """Base module Exception class. """ pass class HandlerHTTPMethodError(HandlerError): """Raises for invalid HTTP method declaration. """ pass class HandlerPluginError(HandlerError): """Raises when a handler plugin error occurs. """ pass class ClassPropertyDescriptor(object): """ClassProperty Descriptor class. Straight up stolen from stack overflow Implements class level property non-data descriptor. """ def __init__(self, fget, fset=None): self.fget = fget self.fset = fset def __get__(self, obj, cls=None): if cls is None: # pragma: no cover cls = type(obj) return self.fget.__get__(obj, cls)() def classproperty(func): """classproperty decorator. Using this decorator a class can have a property. Necessary for properties that don't need instance initialization. Works exactly the same as a normal property. Examples: >>> class MyClass(object): ... @classproperty ... def my_prop(self): ... return self.__name__ + ' class' ... >>> MyClass.my_prop 'MyClass class' """ if not isinstance(func, (classmethod, staticmethod)): func = classmethod(func) return ClassPropertyDescriptor(func) def cached_classproperty(fun): """A memorization decorator for class properties. It implements the above `classproperty` decorator, with the difference that the function result is computed and attached to class as direct attribute. (Lazy loading and caching.) """ @functools.wraps(fun) def get(cls): try: return cls.__cache[fun] except AttributeError: cls.__cache = {} except KeyError: # pragma: no cover pass ret = cls.__cache[fun] = fun(cls) return ret return classproperty(get) def plugin_method(*plugin_names): """Plugin Method decorator. Signs a web handler function with the plugins to be applied as attributes. Args: plugin_names (list): A list of plugin callable names Returns: A wrapped handler callable. Examples: >>> @plugin_method('json', 'bill') ... def method(): ... return "Hello!" ... >>> print method.json True >>> print method.bill True """ def wrapper(callable_obj): for plugin_name in plugin_names: if not hasattr(callable_obj, plugin_name): setattr(callable_obj, plugin_name, True) return callable_obj return wrapper class BaseHandlerPlugin(object): """ """ def __init__(self, callable_object, *args, **kwargs): # pragma: no cover self._wrapped = callable_object self._args = args self._kwargs = kwargs self.__doc__ = callable_object.__doc__ @cached_classproperty def func_name(self): cls_name = re.sub('(.)([A-Z][a-z]+)', r'\1_\2', self.__name__) return re.sub('([a-z0-9])([A-Z])', r'\1_\2', cls_name).lower() def apply(self, *args, **kwargs): # pragma: no cover raise NotImplementedError("Must Override `apply` method") def __call__(self, *args, **kwargs): # pragma: no cover return self.apply(*args, **kwargs) class HandlerMeta(type): """BaseHandler metaclass Provides meta functionality for ensuring that `http_method` don't require an instance and test and that derived classes implements `Singleton`. """ _instances = {} def __new__(mcs, name, bases, attrs): http_methods = [attr for attr in attrs if attr.lower() in DEFAULT_ROUTES] for handler in http_methods: attrs[handler] = classmethod(attrs[handler]) return super(HandlerMeta, mcs).__new__(mcs, name, bases, attrs) def __call__(cls, *args): """Enforce Singleton class initiation """ if cls not in cls._instances: cls._instances[cls] = super( HandlerMeta, cls ).__call__(*args) return cls._instances[cls] @six.add_metaclass(HandlerMeta) class BaseHandler(object): """Base Handler class for implementing Class-Based Web handler for bottle. Subclass `BaseHandler` in order to its API for application routing, and implement any of the known http method: Class Attributes: render (object): bottle.jinja2_render. request (object): bottle.request response (object): bottle.response redirect (object):bottle.redirect abort (object): bottle.abort base_endpoint (str): The handler endpoint prefix. cors_enabled (bool): Indicates if CORS is enabled. plugins (dict): A key/value mapping of available plugins. global_plugins (dict): A key/value mapping default applying plugins. """ render = bottle.jinja2_template request = bottle.request response = bottle.response redirect = bottle.redirect abort = bottle.abort response_factory = None base_endpoint = '/' cors_enabled = True plugins = dict() global_plugins = dict() @classmethod def add_plugin(cls, plugin_callables, global_scope=False): """Register a new plugin in handler. Args: plugin_callables (list): A list of plugin callables. global_scope (bool): Indicates The scope of the plugin. Returns: Class instance. """ repo = getattr(cls, dict(PLUGIN_SCOPE)[global_scope]) for plugin_callable in plugin_callables: if hasattr(plugin_callable, 'func_name'): repo[plugin_callable.func_name] = plugin_callable else: # pragma: no cover repo[plugin_callable.__name__] = plugin_callable return cls @classmethod def register_app(cls, application): """Register class view in bottle application. Args: application (instance): A bottle.Bottle() instance. Returns: Class instance. """ if cls is BaseHandler: raise HandlerError("Cant register a `BaseHandler` class instance") routes = cls._get_http_members() router = getattr(application, "route") for func_name, func_callable in routes: method_args = inspect.signature(func_callable).parameters if hasattr(func_callable, 'http_method'): http_method = func_callable.http_method url_extra_part = func_callable.url_extra_part else: http_method = func_name url_extra_part = None applied_plugins = [pl for pl in cls.plugins if hasattr(func_callable, pl)] for plugin in applied_plugins: # pragma: no cover try: func_callable = cls.plugins[plugin](func_callable) except TypeError as error: raise HandlerPluginError(error.args) for global_plugin in cls.global_plugins: func_callable = cls.global_plugins[global_plugin]( func_callable ) for entrypoint in cls._build_routes(method_args, url_extra_part): router(entrypoint, method=[http_method.upper()])(func_callable) if cls.cors_enabled: cls_desc = cls.__doc__ or '' options_data = { "handler": {"name": cls.__name__, "desc": cls_desc.strip()}, "http_methods": [r[0] for r in routes] } router( cls.base_endpoint, method=["OPTIONS"] )(lambda: options_data) router( "{}/<url:re:.+>".format(cls.base_endpoint).replace("//", "/"), method=["OPTIONS"] )(lambda url: options_data) return cls @classmethod def _build_routes(cls, method_args, url_extra_part=None): """Create bottle route for a handler http method.""" prefix = '/{}'.format(url_extra_part) if url_extra_part else '' endpoint = cls.base_endpoint + prefix if not method_args: return [endpoint] endpoints = [] for args_list in cls._router_helper(method_args): prefix = '/:' if args_list else '' endpoints.append((endpoint + prefix + '/:'.join(args_list)).replace("//", '/')) return endpoints @classmethod def _router_helper(cls, method_args): """Detect default Nullable method arguments and return multiple routes per callable. """ fixed_params = [param for param in method_args if method_args[param].default is not None] nullable_params = [param for param in method_args if param not in fixed_params] combinations = [fixed_params] combinations += [combinations[-1] + [param] for param in nullable_params] return combinations @classmethod def _get_http_members(cls): """Filter all `http` specific method from handler class. """ return [member for member in inspect.getmembers(cls, predicate=inspect.ismethod) if cls._check_http_member(member)] @staticmethod def _check_http_member(member): """Checks if a class method has HTTP info. """ return member[0].lower() in DEFAULT_ROUTES or \ hasattr(member[1], 'http_method') if __name__ == '__main__': # pragma: no cover import doctest doctest.testmod()
agile4you/bottle-neck
bottle_neck/plugins.py
WrapErrorPlugin.apply
python
def apply(self, callback, context): # pragma: no cover def wrapper(*args, **kwargs): try: return callback(*args, **kwargs) except bottle.HTTPError as error: return self.error_wrapper.from_status( status_line=error.status_line, msg=error.body ) return wrapper
Apply the HTTPError wrapper to the callback.
train
https://github.com/agile4you/bottle-neck/blob/ebc670a4b178255473d68e9b4122ba04e38f4810/bottle_neck/plugins.py#L61-L74
null
class WrapErrorPlugin(BasePlugin): """Middleware class that catches `bottle.HTTPError` exceptions and returns default HTTP status code 200 using `bottle_neck.response.WSResponse` class for error wrapping. """ def __init__(self, keyword, error_wrapper_cls): # pragma: no cover self.error_wrapper = error_wrapper_cls self.keyword = keyword