id
int64
0
190k
prompt
stringlengths
21
13.4M
docstring
stringlengths
1
12k
9,848
import math from typing import Dict import evdev from evdev.ecodes import ( EV_REL, REL_WHEEL, REL_HWHEEL, REL_WHEEL_HI_RES, REL_HWHEEL_HI_RES, ) from inputremapper.configs.input_config import InputCombination, InputConfig from inputremapper import exceptions from inputremapper.configs.mapping import ( Mapping, REL_XY_SCALING, WHEEL_SCALING, WHEEL_HI_RES_SCALING, ) from inputremapper.injection.global_uinputs import global_uinputs from inputremapper.injection.mapping_handlers.axis_transform import Transformation from inputremapper.injection.mapping_handlers.mapping_handler import ( MappingHandler, HandlerEnums, InputEventHandler, ) from inputremapper.input_event import InputEvent from inputremapper.logger import logger def is_high_res_wheel(event) -> bool: return event.type == EV_REL and event.code in (REL_WHEEL_HI_RES, REL_HWHEEL_HI_RES)
null
9,849
from collections import defaultdict from typing import Dict, List, Type, Optional, Set, Iterable, Sized, Tuple, Sequence from evdev.ecodes import EV_KEY, EV_ABS, EV_REL from inputremapper.configs.input_config import InputCombination, InputConfig from inputremapper.configs.mapping import Mapping from inputremapper.configs.preset import Preset from inputremapper.configs.system_mapping import DISABLE_CODE, DISABLE_NAME from inputremapper.exceptions import MappingParsingError from inputremapper.injection.macros.parse import is_this_a_macro from inputremapper.injection.mapping_handlers.abs_to_abs_handler import AbsToAbsHandler from inputremapper.injection.mapping_handlers.abs_to_btn_handler import AbsToBtnHandler from inputremapper.injection.mapping_handlers.abs_to_rel_handler import AbsToRelHandler from inputremapper.injection.mapping_handlers.axis_switch_handler import ( AxisSwitchHandler, ) from inputremapper.injection.mapping_handlers.combination_handler import ( CombinationHandler, ) from inputremapper.injection.mapping_handlers.hierarchy_handler import HierarchyHandler from inputremapper.injection.mapping_handlers.key_handler import KeyHandler from inputremapper.injection.mapping_handlers.macro_handler import MacroHandler from inputremapper.injection.mapping_handlers.mapping_handler import ( HandlerEnums, MappingHandler, ContextProtocol, InputEventHandler, ) from inputremapper.injection.mapping_handlers.null_handler import NullHandler from inputremapper.injection.mapping_handlers.rel_to_abs_handler import RelToAbsHandler from inputremapper.injection.mapping_handlers.rel_to_btn_handler import RelToBtnHandler from inputremapper.injection.mapping_handlers.rel_to_rel_handler import RelToRelHandler from inputremapper.logger import logger from inputremapper.utils import get_evdev_constant_name EventPipelines = Dict[InputConfig, Set[InputEventHandler]] mapping_handler_classes: Dict[HandlerEnums, Optional[Type[MappingHandler]]] = { # all available mapping_handlers HandlerEnums.abs2btn: AbsToBtnHandler, HandlerEnums.rel2btn: RelToBtnHandler, HandlerEnums.macro: MacroHandler, HandlerEnums.key: KeyHandler, HandlerEnums.btn2rel: None, # can be a macro HandlerEnums.rel2rel: RelToRelHandler, HandlerEnums.abs2rel: AbsToRelHandler, HandlerEnums.btn2abs: None, # can be a macro HandlerEnums.rel2abs: RelToAbsHandler, HandlerEnums.abs2abs: AbsToAbsHandler, HandlerEnums.combination: CombinationHandler, HandlerEnums.hierarchy: HierarchyHandler, HandlerEnums.axisswitch: AxisSwitchHandler, HandlerEnums.disable: NullHandler, } def _create_event_pipeline( handler: MappingHandler, context: ContextProtocol, ignore_ranking=False ) -> List[MappingHandler]: """Recursively wrap a handler with other handlers until the outer handler needs ranking or is finished wrapping. """ if not handler.needs_wrapping() or (handler.needs_ranking() and not ignore_ranking): return [handler] handlers = [] for combination, handler_enum in handler.wrap_with().items(): constructor = mapping_handler_classes[handler_enum] if not constructor: raise NotImplementedError( f"mapping handler {handler_enum} is not implemented" ) super_handler = constructor(combination, handler.mapping, context=context) super_handler.set_sub_handler(handler) for event in combination: # the handler now has a super_handler which takes care about the events. # so we need to hide them on the handler handler.occlude_input_event(event) handlers.extend(_create_event_pipeline(super_handler, context)) if handler.input_configs: # the handler was only partially wrapped, # we need to return it as a toplevel handler handlers.append(handler) return handlers def _get_output_handler(mapping: Mapping) -> HandlerEnums: """Determine the correct output handler. this is used as a starting point for the mapping parser """ if mapping.output_code == DISABLE_CODE or mapping.output_symbol == DISABLE_NAME: return HandlerEnums.disable if mapping.output_symbol: if is_this_a_macro(mapping.output_symbol): return HandlerEnums.macro return HandlerEnums.key if mapping.output_type == EV_KEY: return HandlerEnums.key input_event = _maps_axis(mapping.input_combination) if not input_event: raise MappingParsingError( f"This {mapping = } does not map to an axis, key or macro", mapping=Mapping, ) if mapping.output_type == EV_REL: if input_event.type == EV_KEY: return HandlerEnums.btn2rel if input_event.type == EV_REL: return HandlerEnums.rel2rel if input_event.type == EV_ABS: return HandlerEnums.abs2rel if mapping.output_type == EV_ABS: if input_event.type == EV_KEY: return HandlerEnums.btn2abs if input_event.type == EV_REL: return HandlerEnums.rel2abs if input_event.type == EV_ABS: return HandlerEnums.abs2abs raise MappingParsingError(f"the output of {mapping = } is unknown", mapping=Mapping) def _create_hierarchy_handlers( handlers: Dict[InputCombination, Set[MappingHandler]] ) -> Set[MappingHandler]: """Sort handlers by input events and create Hierarchy handlers.""" sorted_handlers = set() all_combinations = handlers.keys() events = set() # gather all InputEvents from all handlers for combination in all_combinations: for event in combination: events.add(event) # create a ranking for each event for event in events: # find all combinations (from handlers) which contain the event combinations_with_event = [ combination for combination in all_combinations if event in combination ] if len(combinations_with_event) == 1: # there was only one handler containing that event return it as is sorted_handlers.update(handlers[combinations_with_event[0]]) continue # there are multiple handler with the same event. # rank them and create the HierarchyHandler sorted_combinations = _order_combinations(combinations_with_event, event) sub_handlers: List[MappingHandler] = [] for combination in sorted_combinations: sub_handlers.append(*handlers[combination]) sorted_handlers.add(HierarchyHandler(sub_handlers, event)) for handler in sub_handlers: # the handler now has a HierarchyHandler which takes care about this event. # so we hide need to hide it on the handler handler.occlude_input_event(event) return sorted_handlers class Preset(Generic[MappingModel]): """Contains and manages mappings of a single preset.""" # workaround for typing: https://github.com/python/mypy/issues/4236 def __init__(self: Preset[Mapping], path: Optional[os.PathLike] = None): ... def __init__( self, path: Optional[os.PathLike] = None, mapping_factory: Type[MappingModel] = ..., ): ... def __init__( self, path: Optional[os.PathLike] = None, mapping_factory=Mapping, ) -> None: self._mappings: Dict[InputCombination, MappingModel] = {} # a copy of mappings for keeping track of changes self._saved_mappings: Dict[InputCombination, MappingModel] = {} self._path: Optional[os.PathLike] = path # the mapping class which is used by load() self._mapping_factory: Type[MappingModel] = mapping_factory def __iter__(self) -> Iterator[MappingModel]: """Iterate over Mapping objects.""" return iter(self._mappings.copy().values()) def __len__(self) -> int: return len(self._mappings) def __bool__(self): # otherwise __len__ will be used which results in False for a preset # without mappings return True def has_unsaved_changes(self) -> bool: """Check if there are unsaved changed.""" return self._mappings != self._saved_mappings def remove(self, combination: InputCombination) -> None: """Remove a mapping from the preset by providing the InputCombination.""" if not isinstance(combination, InputCombination): raise TypeError( f"combination must by of type InputCombination, got {type(combination)}" ) for permutation in combination.get_permutations(): if permutation in self._mappings.keys(): combination = permutation break try: mapping = self._mappings.pop(combination) mapping.remove_combination_changed_callback() except KeyError: logger.debug( "unable to remove non-existing mapping with combination = %s", combination, ) pass def add(self, mapping: MappingModel) -> None: """Add a mapping to the preset.""" for permutation in mapping.input_combination.get_permutations(): if permutation in self._mappings: raise KeyError( "A mapping with this input_combination: " f"{permutation} already exists", ) mapping.set_combination_changed_callback(self._combination_changed_callback) self._mappings[mapping.input_combination] = mapping def empty(self) -> None: """Remove all mappings and custom configs without saving. note: self.has_unsaved_changes() will report True """ for mapping in self._mappings.values(): mapping.remove_combination_changed_callback() self._mappings = {} def clear(self) -> None: """Remove all mappings and also self.path.""" self.empty() self._saved_mappings = {} self.path = None def load(self) -> None: """Load from the mapping from the disc, clears all existing mappings.""" logger.info('Loading preset from "%s"', self.path) if not self.path or not os.path.exists(self.path): raise FileNotFoundError(f'Tried to load non-existing preset "{self.path}"') self._saved_mappings = self._get_mappings_from_disc() self.empty() for mapping in self._saved_mappings.values(): # use the public add method to make sure # the _combination_changed_callback is attached self.add(mapping.copy()) def _is_mapped_multiple_times(self, input_combination: InputCombination) -> bool: """Check if the event combination maps to multiple mappings.""" all_input_combinations = {mapping.input_combination for mapping in self} permutations = set(input_combination.get_permutations()) union = permutations & all_input_combinations # if there are more than one matches, then there is a duplicate return len(union) > 1 def _has_valid_input_combination(self, mapping: UIMapping) -> bool: """Check if the mapping has a valid input event combination.""" is_a_combination = isinstance(mapping.input_combination, InputCombination) is_empty = mapping.input_combination == InputCombination.empty_combination() return is_a_combination and not is_empty def save(self) -> None: """Dump as JSON to self.path.""" if not self.path: logger.debug("unable to save preset without a path set Preset.path first") return touch(self.path) if not self.has_unsaved_changes(): logger.debug("Not saving unchanged preset") return logger.info("Saving preset to %s", self.path) preset_list = [] saved_mappings = {} for mapping in self: if not mapping.is_valid(): if not self._has_valid_input_combination(mapping): # we save invalid mappings except for those with an invalid # input_combination logger.debug("Skipping invalid mapping %s", mapping) continue if self._is_mapped_multiple_times(mapping.input_combination): # todo: is this ever executed? it should not be possible to # reach this logger.debug( "skipping mapping with duplicate event combination %s", mapping, ) continue mapping_dict = mapping.dict(exclude_defaults=True) mapping_dict["input_combination"] = mapping.input_combination.to_config() combination = mapping.input_combination preset_list.append(mapping_dict) saved_mappings[combination] = mapping.copy() saved_mappings[combination].remove_combination_changed_callback() with open(self.path, "w") as file: json.dump(preset_list, file, indent=4) file.write("\n") self._saved_mappings = saved_mappings def is_valid(self) -> bool: return False not in [mapping.is_valid() for mapping in self] def get_mapping( self, combination: Optional[InputCombination] ) -> Optional[MappingModel]: """Return the Mapping that is mapped to this InputCombination.""" if not combination: return None if not isinstance(combination, InputCombination): raise TypeError( f"combination must by of type InputCombination, got {type(combination)}" ) for permutation in combination.get_permutations(): existing = self._mappings.get(permutation) if existing is not None: return existing return None def dangerously_mapped_btn_left(self) -> bool: """Return True if this mapping disables BTN_Left.""" if (ecodes.EV_KEY, ecodes.BTN_LEFT) not in [ m.input_combination[0].type_and_code for m in self ]: return False values: List[str | Tuple[int, int] | None] = [] for mapping in self: if mapping.output_symbol is None: continue values.append(mapping.output_symbol.lower()) values.append(mapping.get_output_type_code()) return ( "btn_left" not in values or InputConfig.btn_left().type_and_code not in values ) def _combination_changed_callback( self, new: InputCombination, old: InputCombination ) -> None: for permutation in new.get_permutations(): if permutation in self._mappings.keys() and permutation != old: raise KeyError("combination already exists in the preset") self._mappings[new] = self._mappings.pop(old) def _update_saved_mappings(self) -> None: if self.path is None: return if not os.path.exists(self.path): self._saved_mappings = {} return self._saved_mappings = self._get_mappings_from_disc() def _get_mappings_from_disc(self) -> Dict[InputCombination, MappingModel]: mappings: Dict[InputCombination, MappingModel] = {} if not self.path: logger.debug("unable to read preset without a path set Preset.path first") return mappings if os.stat(self.path).st_size == 0: logger.debug("got empty file") return mappings with open(self.path, "r") as file: try: preset_list = json.load(file) except json.JSONDecodeError: logger.error("unable to decode json file: %s", self.path) return mappings for mapping_dict in preset_list: if not isinstance(mapping_dict, dict): logger.error("Expected mapping to be a dict: %s", mapping_dict) continue try: mapping = self._mapping_factory(**mapping_dict) except Exception as error: logger.error( "failed to Validate mapping for %s: %s", mapping_dict.get("input_combination"), error, ) continue mappings[mapping.input_combination] = mapping return mappings def path(self) -> Optional[os.PathLike]: return self._path def path(self, path: Optional[os.PathLike]): if path != self.path: self._path = path self._update_saved_mappings() def name(self) -> Optional[str]: """The name of the preset.""" if self.path: return os.path.basename(self.path).split(".")[0] return None class MappingParsingError(Error): """Anything that goes wrong during the creation of handlers from the mapping.""" def __init__(self, msg: str, *, mapping=None, mapping_handler=None): self.mapping_handler = mapping_handler self.mapping = mapping super().__init__(msg) class ContextProtocol(Protocol): """The parts from context needed for handlers.""" listeners: Set[EventListener] def get_forward_uinput(self, origin_hash) -> evdev.UInput: pass logger = cast(Logger, logging.getLogger("input-remapper")) logger.addHandler(handler) logger.setLevel(logging.INFO) def get_evdev_constant_name(type_: Optional[int], code: Optional[int], *_) -> str: """Handy function to get the evdev constant name for display purposes. Returns "unknown" for unknown events. """ # using this function is more readable than # type_, code = event.type_and_code # name = evdev.ecodes.bytype[type_][code] name = evdev.ecodes.bytype.get(type_, {}).get(code) if isinstance(name, list): name = name[0] if name is None: return "unknown" return name The provided code snippet includes necessary dependencies for implementing the `parse_mappings` function. Write a Python function `def parse_mappings(preset: Preset, context: ContextProtocol) -> EventPipelines` to solve the following problem: Create a dict with a list of MappingHandler for each InputEvent. Here is the function: def parse_mappings(preset: Preset, context: ContextProtocol) -> EventPipelines: """Create a dict with a list of MappingHandler for each InputEvent.""" handlers = [] for mapping in preset: # start with the last handler in the chain, each mapping only has one output, # but may have multiple inputs, therefore the last handler is a good starting # point to assemble the pipeline handler_enum = _get_output_handler(mapping) constructor = mapping_handler_classes[handler_enum] if not constructor: logger.warning( "a mapping handler '%s' for %s is not implemented", handler_enum, mapping.format_name(), ) continue output_handler = constructor( mapping.input_combination, mapping, context=context, ) # layer other handlers on top until the outer handler needs ranking or can # directly handle a input event handlers.extend(_create_event_pipeline(output_handler, context)) # figure out which handlers need ranking and wrap them with hierarchy_handlers need_ranking = defaultdict(set) for handler in handlers.copy(): if handler.needs_ranking(): combination = handler.rank_by() if not combination: raise MappingParsingError( f"{type(handler).__name__} claims to need ranking but does not " f"return a combination to rank by", mapping_handler=handler, ) need_ranking[combination].add(handler) handlers.remove(handler) # the HierarchyHandler's might not be the starting point of the event pipeline, # layer other handlers on top again. ranked_handlers = _create_hierarchy_handlers(need_ranking) for handler in ranked_handlers: handlers.extend(_create_event_pipeline(handler, context, ignore_ranking=True)) # group all handlers by the input events they take care of. One handler might end # up in multiple groups if it takes care of multiple InputEvents event_pipelines: EventPipelines = defaultdict(set) for handler in handlers: assert handler.input_configs for input_config in handler.input_configs: logger.debug( "event-pipeline with entry point: %s %s", get_evdev_constant_name(*input_config.type_and_code), input_config.input_match_hash, ) logger.debug_mapping_handler(handler) event_pipelines[input_config].add(handler) return event_pipelines
Create a dict with a list of MappingHandler for each InputEvent.
9,850
import asyncio import math import time from functools import partial from typing import Dict, Tuple, Optional import evdev from evdev.ecodes import ( EV_REL, EV_ABS, REL_WHEEL, REL_HWHEEL, REL_WHEEL_HI_RES, REL_HWHEEL_HI_RES, ) from inputremapper.configs.input_config import InputCombination, InputConfig from inputremapper.configs.mapping import ( Mapping, REL_XY_SCALING, WHEEL_SCALING, WHEEL_HI_RES_SCALING, DEFAULT_REL_RATE, ) from inputremapper.injection.global_uinputs import global_uinputs from inputremapper.injection.mapping_handlers.axis_transform import Transformation from inputremapper.injection.mapping_handlers.mapping_handler import ( MappingHandler, HandlerEnums, InputEventHandler, ) from inputremapper.input_event import InputEvent, EventActions from inputremapper.logger import logger from inputremapper.utils import get_evdev_constant_name def calculate_output(value, weight, remainder): # self._value is between 0 and 1, scale up with weight scaled = value * weight + remainder # float_value % 1 will result in wrong calculations for negative values remainder = math.fmod(scaled, 1) return int(scaled), remainder REL_XY_SCALING: float = 60 DEFAULT_REL_RATE: float = 60 The provided code snippet includes necessary dependencies for implementing the `_run_normal_output` function. Write a Python function `async def _run_normal_output(self) -> None` to solve the following problem: Start injecting events. Here is the function: async def _run_normal_output(self) -> None: """Start injecting events.""" self._running = True self._stop = False remainder = 0.0 start = time.time() # if the rate is configured to be slower than the default, increase the value, so # that the overall speed stays the same. rate_compensation = DEFAULT_REL_RATE / self.mapping.rel_rate weight = REL_XY_SCALING * rate_compensation while not self._stop: value, remainder = calculate_output( self._value, weight, remainder, ) self._write(EV_REL, self.mapping.output_code, value) time_taken = time.time() - start sleep = max(0.0, (1 / self.mapping.rel_rate) - time_taken) await asyncio.sleep(sleep) start = time.time() self._running = False
Start injecting events.
9,851
import asyncio import math import time from functools import partial from typing import Dict, Tuple, Optional import evdev from evdev.ecodes import ( EV_REL, EV_ABS, REL_WHEEL, REL_HWHEEL, REL_WHEEL_HI_RES, REL_HWHEEL_HI_RES, ) from inputremapper.configs.input_config import InputCombination, InputConfig from inputremapper.configs.mapping import ( Mapping, REL_XY_SCALING, WHEEL_SCALING, WHEEL_HI_RES_SCALING, DEFAULT_REL_RATE, ) from inputremapper.injection.global_uinputs import global_uinputs from inputremapper.injection.mapping_handlers.axis_transform import Transformation from inputremapper.injection.mapping_handlers.mapping_handler import ( MappingHandler, HandlerEnums, InputEventHandler, ) from inputremapper.input_event import InputEvent, EventActions from inputremapper.logger import logger from inputremapper.utils import get_evdev_constant_name def calculate_output(value, weight, remainder): # self._value is between 0 and 1, scale up with weight scaled = value * weight + remainder # float_value % 1 will result in wrong calculations for negative values remainder = math.fmod(scaled, 1) return int(scaled), remainder WHEEL_SCALING: float = 1 WHEEL_HI_RES_SCALING: float = 120 The provided code snippet includes necessary dependencies for implementing the `_run_wheel_output` function. Write a Python function `async def _run_wheel_output(self, codes: Tuple[int, int]) -> None` to solve the following problem: Start injecting wheel events. made to inject both REL_WHEEL and REL_WHEEL_HI_RES events, because otherwise wheel output doesn't work for some people. See issue #354 Here is the function: async def _run_wheel_output(self, codes: Tuple[int, int]) -> None: """Start injecting wheel events. made to inject both REL_WHEEL and REL_WHEEL_HI_RES events, because otherwise wheel output doesn't work for some people. See issue #354 """ weights = (WHEEL_SCALING, WHEEL_HI_RES_SCALING) self._running = True self._stop = False remainder = [0.0, 0.0] start = time.time() while not self._stop: for i in range(len(codes)): value, remainder[i] = calculate_output( self._value, weights[i], remainder[i], ) self._write(EV_REL, codes[i], value) time_taken = time.time() - start await asyncio.sleep(max(0.0, (1 / self.mapping.rel_rate) - time_taken)) start = time.time() self._running = False
Start injecting wheel events. made to inject both REL_WHEEL and REL_WHEEL_HI_RES events, because otherwise wheel output doesn't work for some people. See issue #354
9,852
from __future__ import annotations import asyncio import enum import multiprocessing import sys import time from collections import defaultdict from dataclasses import dataclass from multiprocessing.connection import Connection from typing import Dict, List, Optional, Tuple, Union import evdev from inputremapper.configs.input_config import InputCombination, InputConfig, DeviceHash from inputremapper.configs.preset import Preset from inputremapper.groups import ( _Group, classify, DeviceType, ) from inputremapper.gui.messages.message_broker import MessageType from inputremapper.injection.context import Context from inputremapper.injection.event_reader import EventReader from inputremapper.injection.numlock import set_numlock, is_numlock_on, ensure_numlock from inputremapper.logger import logger from inputremapper.utils import get_device_hash CapabilitiesDict = Dict[int, List[int]] class InputCombination(Tuple[InputConfig, ...]): """One or more InputConfigs used to trigger a mapping.""" # tuple is immutable, therefore we need to override __new__() # https://jfine-python-classes.readthedocs.io/en/latest/subclass-tuple.html def __new__(cls, configs: InputCombinationInit) -> InputCombination: """Create a new InputCombination. Examples -------- InputCombination([InputConfig, ...]) InputCombination([{type: ..., code: ..., value: ...}, ...]) """ if not isinstance(configs, Iterable): raise TypeError("InputCombination requires a list of InputConfigs.") if isinstance(configs, InputConfig): # wrap the argument in square brackets raise TypeError("InputCombination requires a list of InputConfigs.") validated_configs = [] for config in configs: if isinstance(configs, InputEvent): raise TypeError("InputCombinations require InputConfigs, not Events.") if isinstance(config, InputConfig): validated_configs.append(config) elif isinstance(config, dict): validated_configs.append(InputConfig(**config)) else: raise TypeError(f'Can\'t handle "{config}"') if len(validated_configs) == 0: raise ValueError(f"failed to create InputCombination with {configs = }") # mypy bug: https://github.com/python/mypy/issues/8957 # https://github.com/python/mypy/issues/8541 return super().__new__(cls, validated_configs) # type: ignore def __str__(self): return f'Combination ({" + ".join(str(event) for event in self)})' def __repr__(self): combination = ", ".join(repr(event) for event in self) return f"<InputCombination ({combination}) at {hex(id(self))}>" def __get_validators__(cls): """Used by pydantic to create InputCombination objects.""" yield cls.validate def validate(cls, init_arg) -> InputCombination: """The only valid option is from_config""" if isinstance(init_arg, InputCombination): return init_arg return cls(init_arg) def to_config(self) -> Tuple[Dict[str, int], ...]: """Turn the object into a tuple of dicts.""" return tuple(input_config.dict(exclude_defaults=True) for input_config in self) def empty_combination(cls) -> InputCombination: """A combination that has default invalid (to evdev) values. Useful for the UI to indicate that this combination is not set """ return cls([{"type": EMPTY_TYPE, "code": 99, "analog_threshold": 99}]) def from_tuples(cls, *tuples): """Construct an InputCombination from (type, code, analog_threshold) tuples.""" dicts = [] for tuple_ in tuples: if len(tuple_) == 3: dicts.append( { "type": tuple_[0], "code": tuple_[1], "analog_threshold": tuple_[2], } ) elif len(tuple_) == 2: dicts.append( { "type": tuple_[0], "code": tuple_[1], } ) else: raise TypeError return cls(dicts) def is_problematic(self) -> bool: """Is this combination going to work properly on all systems?""" if len(self) <= 1: return False for input_config in self: if input_config.type != ecodes.EV_KEY: continue if input_config.code in DIFFICULT_COMBINATIONS: return True return False def defines_analog_input(self) -> bool: """Check if there is any analog input in self.""" return True in tuple(i.defines_analog_input for i in self) def find_analog_input_config( self, type_: Optional[int] = None ) -> Optional[InputConfig]: """Return the first event that defines an analog input.""" for input_config in self: if input_config.defines_analog_input and ( type_ is None or input_config.type == type_ ): return input_config return None def get_permutations(self) -> List[InputCombination]: """Get a list of EventCombinations representing all possible permutations. combining a + b + c should have the same result as b + a + c. Only the last combination remains the same in the returned result. """ if len(self) <= 2: return [self] permutations = [] for permutation in itertools.permutations(self[:-1]): permutations.append(InputCombination((*permutation, self[-1]))) return permutations def beautify(self) -> str: """Get a human-readable string representation.""" if self == InputCombination.empty_combination(): return "empty_combination" return " + ".join(event.description(exclude_threshold=True) for event in self) The provided code snippet includes necessary dependencies for implementing the `is_in_capabilities` function. Write a Python function `def is_in_capabilities( combination: InputCombination, capabilities: CapabilitiesDict ) -> bool` to solve the following problem: Are this combination or one of its sub keys in the capabilities? Here is the function: def is_in_capabilities( combination: InputCombination, capabilities: CapabilitiesDict ) -> bool: """Are this combination or one of its sub keys in the capabilities?""" for event in combination: if event.code in capabilities.get(event.type, []): return True return False
Are this combination or one of its sub keys in the capabilities?
9,853
from __future__ import annotations import asyncio import enum import multiprocessing import sys import time from collections import defaultdict from dataclasses import dataclass from multiprocessing.connection import Connection from typing import Dict, List, Optional, Tuple, Union import evdev from inputremapper.configs.input_config import InputCombination, InputConfig, DeviceHash from inputremapper.configs.preset import Preset from inputremapper.groups import ( _Group, classify, DeviceType, ) from inputremapper.gui.messages.message_broker import MessageType from inputremapper.injection.context import Context from inputremapper.injection.event_reader import EventReader from inputremapper.injection.numlock import set_numlock, is_numlock_on, ensure_numlock from inputremapper.logger import logger from inputremapper.utils import get_device_hash DEV_NAME = "input-remapper" The provided code snippet includes necessary dependencies for implementing the `get_udev_name` function. Write a Python function `def get_udev_name(name: str, suffix: str) -> str` to solve the following problem: Make sure the generated name is not longer than 80 chars. Here is the function: def get_udev_name(name: str, suffix: str) -> str: """Make sure the generated name is not longer than 80 chars.""" max_len = 80 # based on error messages remaining_len = max_len - len(DEV_NAME) - len(suffix) - 2 middle = name[:remaining_len] name = f"{DEV_NAME} {middle} {suffix}" return name
Make sure the generated name is not longer than 80 chars.
9,854
import atexit import json import os import sys import time from pathlib import PurePath from typing import Protocol, Dict, Optional import gi from pydbus import SystemBus from gi.repository import GLib from inputremapper.logger import logger, is_debug from inputremapper.injection.injector import Injector, InjectorState from inputremapper.configs.preset import Preset from inputremapper.configs.global_config import global_config from inputremapper.configs.system_mapping import system_mapping from inputremapper.groups import groups from inputremapper.configs.paths import get_config_path, sanitize_path_component, USER from inputremapper.injection.macros.macro import macro_variables from inputremapper.injection.global_uinputs import global_uinputs The provided code snippet includes necessary dependencies for implementing the `remove_timeout` function. Write a Python function `def remove_timeout(func)` to solve the following problem: Remove timeout to ensure the call works if the daemon is not a proxy. Here is the function: def remove_timeout(func): """Remove timeout to ensure the call works if the daemon is not a proxy.""" # the timeout kwarg is a feature of pydbus. This is needed to make tests work # that create a Daemon by calling its constructor instead of using pydbus. def wrapped(*args, **kwargs): if "timeout" in kwargs: del kwargs["timeout"] return func(*args, **kwargs) return wrapped
Remove timeout to ensure the call works if the daemon is not a proxy.
9,855
import getpass import os import pwd The provided code snippet includes necessary dependencies for implementing the `get_user` function. Write a Python function `def get_user()` to solve the following problem: Try to find the user who called sudo/pkexec. Here is the function: def get_user(): """Try to find the user who called sudo/pkexec.""" try: return os.getlogin() except OSError: # failed in some ubuntu installations and in systemd services pass try: user = os.environ["USER"] except KeyError: # possibly the systemd service. no sudo was used return getpass.getuser() if user == "root": try: return os.environ["SUDO_USER"] except KeyError: # no sudo was used pass try: pkexec_uid = int(os.environ["PKEXEC_UID"]) return pwd.getpwuid(pkexec_uid).pw_name except KeyError: # no pkexec was used or the uid is unknown pass return user
Try to find the user who called sudo/pkexec.
9,856
import getpass import os import pwd The provided code snippet includes necessary dependencies for implementing the `get_home` function. Write a Python function `def get_home(user)` to solve the following problem: Try to find the user's home directory. Here is the function: def get_home(user): """Try to find the user's home directory.""" return pwd.getpwnam(user).pw_dir
Try to find the user's home directory.
9,857
import os import shutil from typing import List, Union, Optional from inputremapper.logger import logger, VERSION from inputremapper.user import USER, HOME def chown(path): """Set the owner of a path to the user.""" try: shutil.chown(path, user=USER, group=USER) except LookupError: # the users group was unknown in one case for whatever reason shutil.chown(path, user=USER) def mkdir(path, log=True): """Create a folder, give it to the user.""" if path == "" or path is None: return if os.path.exists(path): return if log: logger.info('Creating dir "%s"', path) # give all newly created folders to the user. # e.g. if .config/input-remapper/mouse/ is created the latter two base = os.path.split(path)[0] mkdir(base, log=False) os.makedirs(path) chown(path) logger = cast(Logger, logging.getLogger("input-remapper")) logger.addHandler(handler) logger.setLevel(logging.INFO) The provided code snippet includes necessary dependencies for implementing the `touch` function. Write a Python function `def touch(path: Union[str, os.PathLike], log=True)` to solve the following problem: Create an empty file and all its parent dirs, give it to the user. Here is the function: def touch(path: Union[str, os.PathLike], log=True): """Create an empty file and all its parent dirs, give it to the user.""" if str(path).endswith("/"): raise ValueError(f"Expected path to not end with a slash: {path}") if os.path.exists(path): return if log: logger.info('Creating file "%s"', path) mkdir(os.path.dirname(path), log=False) os.mknod(path) chown(path)
Create an empty file and all its parent dirs, give it to the user.
9,858
import os import shutil from typing import List, Union, Optional from inputremapper.logger import logger, VERSION from inputremapper.user import USER, HOME The provided code snippet includes necessary dependencies for implementing the `split_all` function. Write a Python function `def split_all(path: Union[os.PathLike, str]) -> List[str]` to solve the following problem: Split the path into its segments. Here is the function: def split_all(path: Union[os.PathLike, str]) -> List[str]: """Split the path into its segments.""" parts = [] while True: path, tail = os.path.split(path) parts.append(tail) if path == os.path.sep: # we arrived at the root '/' parts.append(path) break if not path: # arrived at start of relative path break parts.reverse() return parts
Split the path into its segments.
9,859
import os import shutil from typing import List, Union, Optional from inputremapper.logger import logger, VERSION from inputremapper.user import USER, HOME CONFIG_PATH = os.path.join(HOME, rel_path) The provided code snippet includes necessary dependencies for implementing the `get_config_path` function. Write a Python function `def get_config_path(*paths) -> str` to solve the following problem: Get a path in ~/.config/input-remapper/. Here is the function: def get_config_path(*paths) -> str: """Get a path in ~/.config/input-remapper/.""" return os.path.join(CONFIG_PATH, *paths)
Get a path in ~/.config/input-remapper/.
9,860
from __future__ import annotations import copy import json import os import re import shutil from pathlib import Path from typing import Iterator, Tuple, Dict, List import pkg_resources from evdev.ecodes import ( EV_KEY, EV_ABS, EV_REL, ABS_X, ABS_Y, ABS_RX, ABS_RY, REL_X, REL_Y, REL_WHEEL_HI_RES, REL_HWHEEL_HI_RES, ) from inputremapper.configs.input_config import InputCombination, InputConfig from inputremapper.configs.mapping import Mapping, UIMapping from inputremapper.configs.paths import get_preset_path, mkdir, CONFIG_PATH, remove from inputremapper.configs.preset import Preset from inputremapper.configs.system_mapping import system_mapping from inputremapper.injection.global_uinputs import global_uinputs from inputremapper.injection.macros.parse import is_this_a_macro from inputremapper.logger import logger, VERSION from inputremapper.user import HOME def config_version(): """Get the version string in config.json as packaging.Version object.""" config_path = os.path.join(CONFIG_PATH, "config.json") if not os.path.exists(config_path): return pkg_resources.parse_version("0.0.0") with open(config_path, "r") as file: config = json.load(file) if "version" in config.keys(): return pkg_resources.parse_version(config["version"]) return pkg_resources.parse_version("0.0.0") def _config_suffix(): """Append the .json suffix to the config file.""" deprecated_path = os.path.join(CONFIG_PATH, "config") config_path = os.path.join(CONFIG_PATH, "config.json") if os.path.exists(deprecated_path) and not os.path.exists(config_path): logger.info('Moving "%s" to "%s"', deprecated_path, config_path) os.rename(deprecated_path, config_path) def _preset_path(): """Migrate the folder structure from < 0.4.0. Move existing presets into the new subfolder 'presets' """ new_preset_folder = os.path.join(CONFIG_PATH, "presets") if os.path.exists(get_preset_path()) or not os.path.exists(CONFIG_PATH): return logger.info("Migrating presets from < 0.4.0...") groups = os.listdir(CONFIG_PATH) mkdir(get_preset_path()) for group in groups: path = os.path.join(CONFIG_PATH, group) if os.path.isdir(path): target = path.replace(CONFIG_PATH, new_preset_folder) logger.info('Moving "%s" to "%s"', path, target) os.rename(path, target) logger.info("done") def _mapping_keys(): """Update all preset mappings. Update all keys in preset to include value e.g.: '1,5'->'1,5,1' """ for preset, preset_structure in all_presets(): if isinstance(preset_structure, list): continue # the preset must be at least 1.6-beta version changes = 0 if "mapping" in preset_structure.keys(): mapping = copy.deepcopy(preset_structure["mapping"]) for key in mapping.keys(): if key.count(",") == 1: preset_structure["mapping"][f"{key},1"] = preset_structure[ "mapping" ].pop(key) changes += 1 if changes: with open(preset, "w") as file: logger.info('Updating mapping keys of "%s"', preset) json.dump(preset_structure, file, indent=4) file.write("\n") def _update_version(): """Write the current version to the config file.""" config_file = os.path.join(CONFIG_PATH, "config.json") if not os.path.exists(config_file): return with open(config_file, "r") as file: config = json.load(file) config["version"] = VERSION with open(config_file, "w") as file: logger.info('Updating version in config to "%s"', VERSION) json.dump(config, file, indent=4) def _rename_to_input_remapper(): """Rename .config/key-mapper to .config/input-remapper.""" old_config_path = os.path.join(HOME, ".config/key-mapper") if not os.path.exists(CONFIG_PATH) and os.path.exists(old_config_path): logger.info("Moving %s to %s", old_config_path, CONFIG_PATH) shutil.move(old_config_path, CONFIG_PATH) def _add_target(): """Add the target field to each preset mapping.""" for preset, preset_structure in all_presets(): if isinstance(preset_structure, list): continue if "mapping" not in preset_structure.keys(): continue changed = False for key, symbol in preset_structure["mapping"].copy().items(): if isinstance(symbol, list): continue target = _find_target(symbol) if target is None: target = "keyboard" symbol = ( f"{symbol}\n" "# Broken mapping:\n" "# No target can handle all specified keycodes" ) logger.info( 'Changing target of mapping for "%s" in preset "%s" to "%s"', key, preset, target, ) symbol = [symbol, target] preset_structure["mapping"][key] = symbol changed = True if not changed: continue with open(preset, "w") as file: logger.info('Adding targets for "%s"', preset) json.dump(preset_structure, file, indent=4) file.write("\n") def _otherwise_to_else(): """Conditional macros should use an "else" parameter instead of "otherwise".""" for preset, preset_structure in all_presets(): if isinstance(preset_structure, list): continue if "mapping" not in preset_structure.keys(): continue changed = False for key, symbol in preset_structure["mapping"].copy().items(): if not is_this_a_macro(symbol[0]): continue symbol_before = symbol[0] symbol[0] = re.sub(r"otherwise\s*=\s*", "else=", symbol[0]) if symbol_before == symbol[0]: continue changed = changed or symbol_before != symbol[0] logger.info( 'Changing mapping for "%s" in preset "%s" to "%s"', key, preset, symbol[0], ) preset_structure["mapping"][key] = symbol if not changed: continue with open(preset, "w") as file: logger.info('Changing otherwise to else for "%s"', preset) json.dump(preset_structure, file, indent=4) file.write("\n") def _convert_to_individual_mappings(): """Convert preset.json from {key: [symbol, target]} to [{input_combination: ..., output_symbol: symbol, ...}] """ for old_preset_path, old_preset in all_presets(): if isinstance(old_preset, list): continue migrated_preset = Preset(old_preset_path, UIMapping) if "mapping" in old_preset.keys(): for combination, symbol_target in old_preset["mapping"].items(): logger.info( 'migrating from "%s: %s" to mapping dict', combination, symbol_target, ) try: combination = _input_combination_from_string(combination) except ValueError: logger.error( "unable to migrate mapping with invalid combination %s", combination, ) continue mapping = UIMapping( input_combination=combination, target_uinput=symbol_target[1], output_symbol=symbol_target[0], ) migrated_preset.add(mapping) if ( "gamepad" in old_preset.keys() and "joystick" in old_preset["gamepad"].keys() ): joystick_dict = old_preset["gamepad"]["joystick"] left_purpose = joystick_dict.get("left_purpose") right_purpose = joystick_dict.get("right_purpose") # TODO if pointer_speed is migrated, why is it in my config? pointer_speed = joystick_dict.get("pointer_speed") if pointer_speed: pointer_speed /= 100 non_linearity = joystick_dict.get("non_linearity") # Todo x_scroll_speed = joystick_dict.get("x_scroll_speed") y_scroll_speed = joystick_dict.get("y_scroll_speed") cfg = { "input_combination": None, "target_uinput": "mouse", "output_type": EV_REL, "output_code": None, } if left_purpose == "mouse": x_config = cfg.copy() y_config = cfg.copy() x_config["input_combination"] = InputCombination( [InputConfig(type=EV_ABS, code=ABS_X)] ) y_config["input_combination"] = InputCombination( [InputConfig(type=EV_ABS, code=ABS_Y)] ) x_config["output_code"] = REL_X y_config["output_code"] = REL_Y mapping_x = Mapping(**x_config) mapping_y = Mapping(**y_config) if pointer_speed: mapping_x.gain = pointer_speed mapping_y.gain = pointer_speed migrated_preset.add(mapping_x) migrated_preset.add(mapping_y) if right_purpose == "mouse": x_config = cfg.copy() y_config = cfg.copy() x_config["input_combination"] = InputCombination( [InputConfig(type=EV_ABS, code=ABS_RX)] ) y_config["input_combination"] = InputCombination( [InputConfig(type=EV_ABS, code=ABS_RY)] ) x_config["output_code"] = REL_X y_config["output_code"] = REL_Y mapping_x = Mapping(**x_config) mapping_y = Mapping(**y_config) if pointer_speed: mapping_x.gain = pointer_speed mapping_y.gain = pointer_speed migrated_preset.add(mapping_x) migrated_preset.add(mapping_y) if left_purpose == "wheel": x_config = cfg.copy() y_config = cfg.copy() x_config["input_combination"] = InputCombination( [InputConfig(type=EV_ABS, code=ABS_X)] ) y_config["input_combination"] = InputCombination( [InputConfig(type=EV_ABS, code=ABS_Y)] ) x_config["output_code"] = REL_HWHEEL_HI_RES y_config["output_code"] = REL_WHEEL_HI_RES mapping_x = Mapping(**x_config) mapping_y = Mapping(**y_config) if x_scroll_speed: mapping_x.gain = x_scroll_speed if y_scroll_speed: mapping_y.gain = y_scroll_speed migrated_preset.add(mapping_x) migrated_preset.add(mapping_y) if right_purpose == "wheel": x_config = cfg.copy() y_config = cfg.copy() x_config["input_combination"] = InputCombination( [InputConfig(type=EV_ABS, code=ABS_RX)] ) y_config["input_combination"] = InputCombination( [InputConfig(type=EV_ABS, code=ABS_RY)] ) x_config["output_code"] = REL_HWHEEL_HI_RES y_config["output_code"] = REL_WHEEL_HI_RES mapping_x = Mapping(**x_config) mapping_y = Mapping(**y_config) if x_scroll_speed: mapping_x.gain = x_scroll_speed if y_scroll_speed: mapping_y.gain = y_scroll_speed migrated_preset.add(mapping_x) migrated_preset.add(mapping_y) migrated_preset.save() def _copy_to_v2(): """Move the beta config to the v2 path, or copy the v1 config to the v2 path.""" # TODO test if os.path.exists(CONFIG_PATH): # don't copy to already existing folder # users should delete the input-remapper-2 folder if they need to return # prioritize the v1 configs over beta configs old_path = os.path.join(HOME, ".config/input-remapper") if os.path.exists(os.path.join(old_path, "config.json")): # no beta path, only old presets exist. COPY to v2 path, which will then be # migrated by the various migrations. logger.debug("copying all from %s to %s", old_path, CONFIG_PATH) shutil.copytree(old_path, CONFIG_PATH) return # if v1 configs don't exist, try to find beta configs. beta_path = os.path.join(HOME, ".config/input-remapper/beta_1.6.0-beta") if os.path.exists(beta_path): # There has never been a different version than "1.6.0-beta" in beta, so we # only need to check for that exact directory # already migrated, possibly new presets in them, move to v2 path logger.debug("moving %s to %s", beta_path, CONFIG_PATH) shutil.move(beta_path, CONFIG_PATH) def _remove_logs(): """We will try to rely on journalctl for this in the future.""" try: remove(f"{HOME}/.log/input-remapper") remove("/var/log/input-remapper") remove("/var/log/input-remapper-control") except Exception as error: logger.debug("Failed to remove deprecated logfiles: %s", str(error)) # this migration is not important. Continue pass global_uinputs = GlobalUInputs() VERSION = "2.0.1" The provided code snippet includes necessary dependencies for implementing the `migrate` function. Write a Python function `def migrate()` to solve the following problem: Migrate config files to the current release. Here is the function: def migrate(): """Migrate config files to the current release.""" _rename_to_input_remapper() _copy_to_v2() v = config_version() if v < pkg_resources.parse_version("0.4.0"): _config_suffix() _preset_path() if v < pkg_resources.parse_version("1.2.2"): _mapping_keys() if v < pkg_resources.parse_version("1.4.0"): global_uinputs.prepare_all() _add_target() if v < pkg_resources.parse_version("1.4.1"): _otherwise_to_else() if v < pkg_resources.parse_version("1.5.0"): _remove_logs() if v < pkg_resources.parse_version("1.6.0-beta"): _convert_to_individual_mappings() # add new migrations here if v < pkg_resources.parse_version(VERSION): _update_version()
Migrate config files to the current release.
9,861
import os import site import sys import pkg_resources from inputremapper.logger import logger logged = False def _try_standard_locations(): # https://specifications.freedesktop.org/basedir-spec/basedir-spec-latest.html # ensure at least /usr/local/share/ and /usr/share/ are tried xdg_data_dirs = set( os.environ.get("XDG_DATA_DIRS", "").split(":") + [ "/usr/local/share/", "/usr/share/", os.path.join(site.USER_BASE, "share/"), ] ) for xdg_data_dir in xdg_data_dirs: candidate = os.path.join(xdg_data_dir, "input-remapper") if os.path.exists(candidate): return candidate return None def _try_python_package_location(): """Look for the data dir at the packages installation location.""" source = None try: source = pkg_resources.require("input-remapper")[0].location # failed in some ubuntu installations except Exception: logger.debug("failed to figure out package location") pass data = None # python3.8/dist-packages python3.7/site-packages, /usr/share, # /usr/local/share, endless options if source and "-packages" not in source and "python" not in source: # probably installed with -e, running from the cloned git source data = os.path.join(source, "data") if not os.path.exists(data): if not logged: logger.debug('-e, but data missing at "%s"', data) data = None return data def _try_env_data_dir(): """Check if input-remappers data can be found at DATA_DIR.""" data_dir = os.environ.get("DATA_DIR", None) if data_dir is None: return None if os.path.exists(data_dir): return data_dir else: logger.error(f'"{ data_dir }" does not exist') return None logger = cast(Logger, logging.getLogger("input-remapper")) logger.addHandler(handler) logger.setLevel(logging.INFO) The provided code snippet includes necessary dependencies for implementing the `get_data_path` function. Write a Python function `def get_data_path(filename="")` to solve the following problem: Depending on the installation prefix, return the data dir. Since it is a nightmare to get stuff installed with pip across distros this is somewhat complicated. Ubuntu uses /usr/local/share for data_files (setup.py) and manjaro uses /usr/share. Here is the function: def get_data_path(filename=""): """Depending on the installation prefix, return the data dir. Since it is a nightmare to get stuff installed with pip across distros this is somewhat complicated. Ubuntu uses /usr/local/share for data_files (setup.py) and manjaro uses /usr/share. """ global logged # depending on where this file is installed to, make sure to use the proper # prefix path for data # https://docs.python.org/3/distutils/setupscript.html?highlight=package_data#installing-additional-files # noqa pylint: disable=line-too-long data = ( _try_env_data_dir() or _try_python_package_location() or _try_standard_locations() ) if data is None: logger.error("Could not find the application data") sys.exit(10) if not logged: logger.debug('Found data at "%s"', data) logged = True return os.path.join(data, filename)
Depending on the installation prefix, return the data dir. Since it is a nightmare to get stuff installed with pip across distros this is somewhat complicated. Ubuntu uses /usr/local/share for data_files (setup.py) and manjaro uses /usr/share.
9,862
from __future__ import annotations from typing import Optional from evdev.ecodes import EV_KEY from inputremapper.configs.system_mapping import system_mapping from inputremapper.injection.global_uinputs import find_fitting_default_uinputs The provided code snippet includes necessary dependencies for implementing the `pydantify` function. Write a Python function `def pydantify(error: type)` to solve the following problem: Generate a string as it would appear IN pydantic error types. This does not include the base class name, which is transformed to snake case in pydantic. Example pydantic error type: "value_error.foobar" for FooBarError. Here is the function: def pydantify(error: type): """Generate a string as it would appear IN pydantic error types. This does not include the base class name, which is transformed to snake case in pydantic. Example pydantic error type: "value_error.foobar" for FooBarError. """ # See https://github.com/pydantic/pydantic/discussions/5112 lower_classname = error.__name__.lower() if lower_classname.endswith("error"): return lower_classname[: -len("error")] return lower_classname
Generate a string as it would appear IN pydantic error types. This does not include the base class name, which is transformed to snake case in pydantic. Example pydantic error type: "value_error.foobar" for FooBarError.
9,863
from __future__ import annotations import asyncio import enum import json import multiprocessing import os import re import threading import traceback from typing import List, Optional import evdev from evdev import InputDevice from evdev.ecodes import ( EV_KEY, EV_ABS, KEY_CAMERA, EV_REL, BTN_STYLUS, ABS_MT_POSITION_X, REL_X, KEY_A, BTN_LEFT, REL_Y, REL_WHEEL, ) from inputremapper.configs.paths import get_preset_path from inputremapper.logger import logger from inputremapper.utils import get_device_hash def path(device): return device.fn
null
9,864
from __future__ import annotations import asyncio import enum import json import multiprocessing import os import re import threading import traceback from typing import List, Optional import evdev from evdev import InputDevice from evdev.ecodes import ( EV_KEY, EV_ABS, KEY_CAMERA, EV_REL, BTN_STYLUS, ABS_MT_POSITION_X, REL_X, KEY_A, BTN_LEFT, REL_Y, REL_WHEEL, ) from inputremapper.configs.paths import get_preset_path from inputremapper.logger import logger from inputremapper.utils import get_device_hash class DeviceType(str, enum.Enum): GAMEPAD = "gamepad" KEYBOARD = "keyboard" MOUSE = "mouse" TOUCHPAD = "touchpad" GRAPHICS_TABLET = "graphics-tablet" CAMERA = "camera" UNKNOWN = "unknown" def _is_gamepad(capabilities): """Check if joystick movements are available for preset.""" # A few buttons that indicate a gamepad buttons = { evdev.ecodes.BTN_BASE, evdev.ecodes.BTN_A, evdev.ecodes.BTN_THUMB, evdev.ecodes.BTN_TOP, evdev.ecodes.BTN_DPAD_DOWN, evdev.ecodes.BTN_GAMEPAD, } if not buttons.intersection(capabilities.get(EV_KEY, [])): # no button is in the key capabilities return False # joysticks abs_capabilities = capabilities.get(EV_ABS, []) if evdev.ecodes.ABS_X not in abs_capabilities: return False if evdev.ecodes.ABS_Y not in abs_capabilities: return False return True def _is_mouse(capabilities): """Check if the capabilities represent those of a mouse.""" # Based on observation, those capabilities need to be present to get an # UInput recognized as mouse # mouse movements if not REL_X in capabilities.get(EV_REL, []): return False if not REL_Y in capabilities.get(EV_REL, []): return False # at least the vertical mouse wheel if not REL_WHEEL in capabilities.get(EV_REL, []): return False # and a mouse click button if not BTN_LEFT in capabilities.get(EV_KEY, []): return False return True def _is_graphics_tablet(capabilities): """Check if the capabilities represent those of a graphics tablet.""" if BTN_STYLUS in capabilities.get(EV_KEY, []): return True return False def _is_touchpad(capabilities): """Check if the capabilities represent those of a touchpad.""" if ABS_MT_POSITION_X in capabilities.get(EV_ABS, []): return True return False def _is_keyboard(capabilities): """Check if the capabilities represent those of a keyboard.""" if KEY_A in capabilities.get(EV_KEY, []): return True return False def _is_camera(capabilities): """Check if the capabilities represent those of a camera.""" key_capa = capabilities.get(EV_KEY) return key_capa and len(key_capa) == 1 and key_capa[0] == KEY_CAMERA The provided code snippet includes necessary dependencies for implementing the `classify` function. Write a Python function `def classify(device) -> DeviceType` to solve the following problem: Figure out what kind of device this is. Use this instead of functions like _is_keyboard to avoid getting false positives. Here is the function: def classify(device) -> DeviceType: """Figure out what kind of device this is. Use this instead of functions like _is_keyboard to avoid getting false positives. """ capabilities = device.capabilities(absinfo=False) if _is_graphics_tablet(capabilities): # check this before is_gamepad to avoid classifying abs_x # as joysticks when they are actually stylus positions return DeviceType.GRAPHICS_TABLET if _is_touchpad(capabilities): return DeviceType.TOUCHPAD if _is_gamepad(capabilities): return DeviceType.GAMEPAD if _is_mouse(capabilities): return DeviceType.MOUSE if _is_camera(capabilities): return DeviceType.CAMERA if _is_keyboard(capabilities): # very low in the chain to avoid classifying most devices # as keyboard, because there are many with ev_key capabilities return DeviceType.KEYBOARD return DeviceType.UNKNOWN
Figure out what kind of device this is. Use this instead of functions like _is_keyboard to avoid getting false positives.
9,865
from __future__ import annotations import asyncio import enum import json import multiprocessing import os import re import threading import traceback from typing import List, Optional import evdev from evdev import InputDevice from evdev.ecodes import ( EV_KEY, EV_ABS, KEY_CAMERA, EV_REL, BTN_STYLUS, ABS_MT_POSITION_X, REL_X, KEY_A, BTN_LEFT, REL_Y, REL_WHEEL, ) from inputremapper.configs.paths import get_preset_path from inputremapper.logger import logger from inputremapper.utils import get_device_hash DENYLIST = [".*Yubico.*YubiKey.*", "Eee PC WMI hotkeys"] The provided code snippet includes necessary dependencies for implementing the `is_denylisted` function. Write a Python function `def is_denylisted(device: InputDevice)` to solve the following problem: Check if a device should not be used in input-remapper. Parameters ---------- device Here is the function: def is_denylisted(device: InputDevice): """Check if a device should not be used in input-remapper. Parameters ---------- device """ for name in DENYLIST: if re.match(name, str(device.name), re.IGNORECASE): return True return False
Check if a device should not be used in input-remapper. Parameters ---------- device
9,866
from __future__ import annotations import asyncio import enum import json import multiprocessing import os import re import threading import traceback from typing import List, Optional import evdev from evdev import InputDevice from evdev.ecodes import ( EV_KEY, EV_ABS, KEY_CAMERA, EV_REL, BTN_STYLUS, ABS_MT_POSITION_X, REL_X, KEY_A, BTN_LEFT, REL_Y, REL_WHEEL, ) from inputremapper.configs.paths import get_preset_path from inputremapper.logger import logger from inputremapper.utils import get_device_hash The provided code snippet includes necessary dependencies for implementing the `get_unique_key` function. Write a Python function `def get_unique_key(device: InputDevice)` to solve the following problem: Find a string key that is unique for a single hardware device. All InputDevices in /dev/input that originate from the same physical hardware device should return the same key via this function. Here is the function: def get_unique_key(device: InputDevice): """Find a string key that is unique for a single hardware device. All InputDevices in /dev/input that originate from the same physical hardware device should return the same key via this function. """ # Keys that should not be used: # - device.phys is empty sometimes and varies across virtual # subdevices # - device.version varies across subdevices return ( # device.info bustype, vendor and product are unique for # a product, but multiple similar device models would be grouped # in the same group f"{device.info.bustype}_" f"{device.info.vendor}_" f"{device.info.product}_" # device.uniq is empty most of the time. It seems to be the only way to # distinguish multiple connected bluetooth gamepads f"{device.uniq}_" # deivce.phys if "/input..." is removed from it, because the first # chunk seems to be unique per hardware (if it's not completely empty) f'{device.phys.split("/")[0] or "-"}' )
Find a string key that is unique for a single hardware device. All InputDevices in /dev/input that originate from the same physical hardware device should return the same key via this function.
9,867
import json import os import select import socket import time from typing import Union from inputremapper.configs.paths import mkdir, chown from inputremapper.logger import logger existing_clients = {} class _Client(Base): """A socket that can be written to and read from.""" def connect(self): if self.socket is not None: return True try: _socket = socket.socket(socket.AF_UNIX) _socket.connect(self._path) logger.debug('Connected to socket: "%s"', self._path) _socket.setblocking(False) except Exception as error: logger.debug('Failed to connect to "%s": "%s"', self._path, error) return False self.socket = _socket self.connection = _socket existing_clients[self._path] = self return True def fileno(self): """For compatibility with select.select.""" self.connect() return self.socket.fileno() def reconnect(self): self.connection = None self.socket = None return self.connect() def Client(path): if path in existing_clients: # ensure it is running, might have been closed existing_clients[path].reset() return existing_clients[path] return _Client(path)
null
9,868
import json import os import select import socket import time from typing import Union from inputremapper.configs.paths import mkdir, chown from inputremapper.logger import logger existing_servers = {} class _Server(Base): """A socket that can be written to and read from. It accepts one connection at a time, and drops old connections if a new one is in sight. """ def connect(self): if self.socket is None: if os.path.exists(self._path): # leftover from the previous execution os.remove(self._path) _socket = socket.socket(socket.AF_UNIX, socket.SOCK_STREAM) _socket.bind(self._path) _socket.listen(1) chown(self._path) logger.debug('Created socket: "%s"', self._path) self.socket = _socket self.socket.setblocking(False) existing_servers[self._path] = self incoming = len(select.select([self.socket], [], [], 0)[0]) != 0 if not incoming and self.connection is None: # no existing connection, no client attempting to connect return False if not incoming and self.connection is not None: # old connection return True if incoming: logger.debug('Incoming connection: "%s"', self._path) connection = self.socket.accept()[0] self.connection = connection self.connection.setblocking(False) return True def fileno(self): """For compatibility with select.select.""" self.connect() return self.connection.fileno() def reconnect(self): self.connection = None return self.connect() def Server(path): if path in existing_servers: # ensure it is running, might have been closed existing_servers[path].reset() return existing_servers[path] return _Server(path)
null
9,869
from __future__ import annotations import time from dataclasses import dataclass from typing import List, Callable, Dict, Optional import gi from gi.repository import Gtk, GLib, Gdk from inputremapper.logger import logger debounce_manager = DebounceManager() The provided code snippet includes necessary dependencies for implementing the `debounce` function. Write a Python function `def debounce(timeout)` to solve the following problem: Debounce a method call to improve performance. Calling this with a millisecond value creates the decorator, so use something like @debounce(50) def function(self): ... In tests, run_all_now can be used to avoid waiting to speed them up. Here is the function: def debounce(timeout): """Debounce a method call to improve performance. Calling this with a millisecond value creates the decorator, so use something like @debounce(50) def function(self): ... In tests, run_all_now can be used to avoid waiting to speed them up. """ # the outside `debounce` function is needed to obtain the millisecond value def decorator(function): # the regular decorator. # @decorator # def foo(): # ... def wrapped(self, *args, **kwargs): # this is the function that will actually be called debounce_manager.debounce(self, function, timeout, *args, **kwargs) wrapped.__name__ = function.__name__ return wrapped return decorator
Debounce a method call to improve performance. Calling this with a millisecond value creates the decorator, so use something like @debounce(50) def function(self): ... In tests, run_all_now can be used to avoid waiting to speed them up.
9,870
from __future__ import annotations import time from dataclasses import dataclass from typing import List, Callable, Dict, Optional import gi from gi.repository import Gtk, GLib, Gdk from inputremapper.logger import logger The provided code snippet includes necessary dependencies for implementing the `gtk_iteration` function. Write a Python function `def gtk_iteration(iterations=0)` to solve the following problem: Iterate while events are pending. Here is the function: def gtk_iteration(iterations=0): """Iterate while events are pending.""" while Gtk.events_pending(): Gtk.main_iteration() for _ in range(iterations): time.sleep(0.002) while Gtk.events_pending(): Gtk.main_iteration()
Iterate while events are pending.
9,871
from __future__ import annotations import asyncio import logging import multiprocessing import os import subprocess import sys import time from collections import defaultdict from typing import Set, List import evdev from evdev.ecodes import EV_KEY, EV_ABS, EV_REL, REL_HWHEEL, REL_WHEEL from inputremapper.utils import get_device_hash from inputremapper.configs.input_config import InputCombination, InputConfig from inputremapper.configs.mapping import Mapping from inputremapper.groups import _Groups, _Group from inputremapper.injection.event_reader import EventReader from inputremapper.injection.mapping_handlers.abs_to_btn_handler import AbsToBtnHandler from inputremapper.injection.mapping_handlers.mapping_handler import ( NotifyCallback, InputEventHandler, MappingHandler, ) from inputremapper.injection.mapping_handlers.rel_to_btn_handler import RelToBtnHandler from inputremapper.input_event import InputEvent, EventActions from inputremapper.ipc.pipe import Pipe from inputremapper.logger import logger from inputremapper.user import USER USER = get_user() The provided code snippet includes necessary dependencies for implementing the `get_pipe_paths` function. Write a Python function `def get_pipe_paths()` to solve the following problem: Get the path where the pipe can be found. Here is the function: def get_pipe_paths(): """Get the path where the pipe can be found.""" return ( f"/tmp/input-remapper-{USER}/reader-results", f"/tmp/input-remapper-{USER}/reader-commands", )
Get the path where the pipe can be found.
9,872
import re from typing import Dict, Optional, List, Tuple from evdev.ecodes import EV_KEY from gi.repository import Gdk, Gtk, GLib, GObject from inputremapper.gui.controller import Controller from inputremapper.configs.mapping import MappingData from inputremapper.configs.system_mapping import system_mapping, DISABLE_NAME from inputremapper.gui.components.editor import CodeEditor from inputremapper.gui.messages.message_broker import MessageBroker, MessageType from inputremapper.gui.messages.message_data import UInputsData from inputremapper.gui.utils import debounce from inputremapper.injection.macros.parse import ( TASK_FACTORIES, get_macro_argument_names, remove_comments, ) from inputremapper.logger import logger def get_incomplete_parameter(iter_: Gtk.TextIter) -> Optional[str]: """Get the parameter that is written left to the TextIter.""" left_text = _get_left_text(iter_) # match foo in: # bar(foo # bar(a=foo # bar(qux, foo # foo # bar + foo match = re.match(rf"(?:{PARAMETER}|^)(\w+)$", left_text) logger.debug('get_incomplete_parameter text: "%s" match: %s', left_text, match) if match is None: return None return match[1] DISABLE_NAME = "disable" system_mapping = SystemMapping() The provided code snippet includes necessary dependencies for implementing the `propose_symbols` function. Write a Python function `def propose_symbols(text_iter: Gtk.TextIter, codes: List[int]) -> List[Tuple[str, str]]` to solve the following problem: Find key names that match the input at the cursor and are mapped to the codes. Here is the function: def propose_symbols(text_iter: Gtk.TextIter, codes: List[int]) -> List[Tuple[str, str]]: """Find key names that match the input at the cursor and are mapped to the codes.""" incomplete_name = get_incomplete_parameter(text_iter) if incomplete_name is None or len(incomplete_name) <= 1: return [] incomplete_name = incomplete_name.lower() names = list(system_mapping.list_names(codes=codes)) + [DISABLE_NAME] return [ (name, name) for name in names if incomplete_name in name.lower() and incomplete_name != name.lower() ]
Find key names that match the input at the cursor and are mapped to the codes.
9,873
import re from typing import Dict, Optional, List, Tuple from evdev.ecodes import EV_KEY from gi.repository import Gdk, Gtk, GLib, GObject from inputremapper.gui.controller import Controller from inputremapper.configs.mapping import MappingData from inputremapper.configs.system_mapping import system_mapping, DISABLE_NAME from inputremapper.gui.components.editor import CodeEditor from inputremapper.gui.messages.message_broker import MessageBroker, MessageType from inputremapper.gui.messages.message_data import UInputsData from inputremapper.gui.utils import debounce from inputremapper.injection.macros.parse import ( TASK_FACTORIES, get_macro_argument_names, remove_comments, ) from inputremapper.logger import logger FUNCTION_NAMES = [name for name in TASK_FACTORIES.keys() if len(name) > 1] FUNCTION_NAMES.remove("ifeq") def get_incomplete_function_name(iter_: Gtk.TextIter) -> str: """Get the word that is written left to the TextIter.""" left_text = _get_left_text(iter_) # match foo in: # bar().foo # bar()\n.foo # bar().\nfoo # bar(\nfoo # bar(\nqux=foo # bar(KEY_A,\nfoo # foo match = re.match(rf"(?:{FUNCTION_CHAIN}|{PARAMETER}|^)(\w+)$", left_text) logger.debug('get_incomplete_function_name text: "%s" match: %s', left_text, match) if match is None: return "" return match[1] TASK_FACTORIES = { "modify": Macro.add_modify, "repeat": Macro.add_repeat, "key": Macro.add_key, "key_down": Macro.add_key_down, "key_up": Macro.add_key_up, "event": Macro.add_event, "wait": Macro.add_wait, "hold": Macro.add_hold, "hold_keys": Macro.add_hold_keys, "mouse": Macro.add_mouse, "wheel": Macro.add_wheel, "if_eq": Macro.add_if_eq, "set": Macro.add_set, "if_tap": Macro.add_if_tap, "if_single": Macro.add_if_single, "add": Macro.add_add, # Those are only kept for backwards compatibility with old macros. The space for # writing macro was very constrained in the past, so shorthands were introduced: "m": Macro.add_modify, "r": Macro.add_repeat, "k": Macro.add_key, "e": Macro.add_event, "w": Macro.add_wait, "h": Macro.add_hold, # It was not possible to adjust ifeq to support variables without breaking old # macros, so this function is deprecated and if_eq introduced. Kept for backwards # compatibility: "ifeq": Macro.add_ifeq, } def get_macro_argument_names(function): """Certain names, like "else" or "type" cannot be used as parameters in python. Removes the trailing "_" for displaying them correctly. """ args = inspect.getfullargspec(function).args[1:] # don't include "self" arg_names = [name[:-1] if name.endswith("_") else name for name in args] varargs = inspect.getfullargspec(function).varargs if varargs: arg_names.append(f"*{varargs}") return arg_names The provided code snippet includes necessary dependencies for implementing the `propose_function_names` function. Write a Python function `def propose_function_names(text_iter: Gtk.TextIter) -> List[Tuple[str, str]]` to solve the following problem: Find function names that match the input at the cursor. Here is the function: def propose_function_names(text_iter: Gtk.TextIter) -> List[Tuple[str, str]]: """Find function names that match the input at the cursor.""" incomplete_name = get_incomplete_function_name(text_iter) if incomplete_name is None or len(incomplete_name) <= 1: return [] incomplete_name = incomplete_name.lower() return [ (name, f"{name}({', '.join(get_macro_argument_names(TASK_FACTORIES[name]))})") for name in FUNCTION_NAMES if incomplete_name in name.lower() and incomplete_name != name.lower() ]
Find function names that match the input at the cursor.
9,874
from typing import Dict, Callable import gi from gi.repository import Gtk, GtkSource, Gdk, GObject from inputremapper.configs.data import get_data_path from inputremapper.configs.mapping import MappingData from inputremapper.configs.input_config import InputCombination from inputremapper.gui.autocompletion import Autocompletion from inputremapper.gui.components.editor import ( MappingListBox, TargetSelection, CodeEditor, RecordingToggle, RecordingStatus, AutoloadSwitch, ReleaseCombinationSwitch, CombinationListbox, AnalogInputSwitch, TriggerThresholdInput, OutputAxisSelector, ReleaseTimeoutInput, TransformationDrawArea, Sliders, RelativeInputCutoffInput, KeyAxisStackSwitcher, RequireActiveMapping, GdkEventRecorder, ) from inputremapper.gui.components.presets import PresetSelection from inputremapper.gui.components.main import Stack, StatusBar from inputremapper.gui.components.common import Breadcrumbs from inputremapper.gui.components.device_groups import DeviceGroupSelection from inputremapper.gui.controller import Controller from inputremapper.gui.messages.message_broker import ( MessageBroker, MessageType, ) from inputremapper.gui.messages.message_data import UserConfirmRequest from inputremapper.gui.utils import ( gtk_iteration, ) from inputremapper.injection.injector import InjectorStateMessage from inputremapper.logger import logger, COMMIT_HASH, VERSION, EVDEV_VERSION from inputremapper.gui.gettext import _ The provided code snippet includes necessary dependencies for implementing the `on_close_about` function. Write a Python function `def on_close_about(about, _)` to solve the following problem: Hide the about dialog without destroying it. Here is the function: def on_close_about(about, _): """Hide the about dialog without destroying it.""" about.hide() return True
Hide the about dialog without destroying it.
9,875
from __future__ import annotations import enum from dataclasses import dataclass from typing import Tuple, Optional, Hashable, Literal import evdev from evdev import ecodes from inputremapper.utils import get_evdev_constant_name The provided code snippet includes necessary dependencies for implementing the `validate_event` function. Write a Python function `def validate_event(event)` to solve the following problem: Test if the event is valid. Here is the function: def validate_event(event): """Test if the event is valid.""" if not isinstance(event.type, int): raise TypeError(f"Expected type to be an int, but got {event.type}") if not isinstance(event.code, int): raise TypeError(f"Expected code to be an int, but got {event.code}") if not isinstance(event.value, int): # this happened to me because I screwed stuff up raise TypeError(f"Expected value to be an int, but got {event.value}") return event
Test if the event is valid.
9,876
import logging import os import sys import time from datetime import datetime from typing import cast handler = logging.StreamHandler() handler.setFormatter(ColorfulFormatter()) def parse_mapping_handler(mapping_handler): indent = 0 lines_and_indent = [] while True: if isinstance(handler, str): lines_and_indent.append([mapping_handler, indent]) break if isinstance(mapping_handler, list): for sub_handler in mapping_handler: sub_list = parse_mapping_handler(sub_handler) for line in sub_list: line[1] += indent lines_and_indent.extend(sub_list) break lines_and_indent.append([repr(mapping_handler), indent]) try: mapping_handler = mapping_handler.child except AttributeError: break indent += 1 return lines_and_indent
null
9,877
import logging import os import sys import time from datetime import datetime from typing import cast logger = cast(Logger, logging.getLogger("input-remapper")) def is_debug(): """True, if the logger is currently in DEBUG or DEBUG mode.""" return logger.level <= logging.DEBUG logger.addHandler(handler) logger.setLevel(logging.INFO) VERSION = "2.0.1" EVDEV_VERSION = None The provided code snippet includes necessary dependencies for implementing the `log_info` function. Write a Python function `def log_info(name="input-remapper")` to solve the following problem: Log version and name to the console. Here is the function: def log_info(name="input-remapper"): """Log version and name to the console.""" logger.info( "%s %s %s https://github.com/sezanzeb/input-remapper", name, VERSION, COMMIT_HASH, ) if EVDEV_VERSION: logger.info("python-evdev %s", EVDEV_VERSION) if is_debug(): logger.warning( "Debug level will log all your keystrokes! Do not post this " "output in the internet if you typed in sensitive or private " "information with your device!" )
Log version and name to the console.
9,878
import logging import os import sys import time from datetime import datetime from typing import cast logging.setLoggerClass(Logger) logger = cast(Logger, logging.getLogger("input-remapper")) logger.addHandler(handler) logger.setLevel(logging.INFO) logging.getLogger("asyncio").setLevel(logging.WARNING) The provided code snippet includes necessary dependencies for implementing the `update_verbosity` function. Write a Python function `def update_verbosity(debug)` to solve the following problem: Set the logging verbosity according to the settings object. Also enable rich tracebacks in debug mode. Here is the function: def update_verbosity(debug): """Set the logging verbosity according to the settings object. Also enable rich tracebacks in debug mode. """ # pylint really doesn't like what I'm doing with rich.traceback here # pylint: disable=broad-except,import-error,import-outside-toplevel if debug: logger.setLevel(logging.DEBUG) try: from rich.traceback import install install(show_locals=True) logger.debug("Using rich.traceback") except Exception as error: # since this is optional, just skip all exceptions if not isinstance(error, ImportError): logger.debug("Cannot use rich.traceback: %s", error) else: logger.setLevel(logging.INFO)
Set the logging verbosity according to the settings object. Also enable rich tracebacks in debug mode.
9,879
import logging import os import sys import time from datetime import datetime from typing import cast logger = cast(Logger, logging.getLogger("input-remapper")) logger.addHandler(handler) logger.setLevel(logging.INFO) The provided code snippet includes necessary dependencies for implementing the `trim_logfile` function. Write a Python function `def trim_logfile(log_path)` to solve the following problem: Keep the logfile short. Here is the function: def trim_logfile(log_path): """Keep the logfile short.""" if not os.path.exists(log_path): return file_size_mb = os.path.getsize(log_path) / 1000 / 1000 if file_size_mb > 100: # something went terribly wrong here. The service might timeout because # it takes too long to trim this file. delete it instead. This probably # only happens when doing funny things while in debug mode. logger.warning( "Removing enormous log file of %dMB", file_size_mb, ) os.remove(log_path) return # the logfile should not be too long to avoid overflowing the storage try: with open(log_path, "rb") as file: binary = file.readlines()[-1000:] content = [line.decode("utf-8", errors="ignore") for line in binary] with open(log_path, "w") as file: file.truncate(0) file.writelines(content) except PermissionError: # let the outermost PermissionError handler handle it raise except Exception as exception: logger.error('Failed to trim logfile: "%s"', str(exception))
Keep the logfile short.
9,880
import keras import numpy as np def serverCheckInput(img): if serverCheckInput.model is None: serverCheckInput.model = keras.models.load_model('./model.h5') prediction = serverCheckInput.model.predict(np.reshape(img, (1, 2, 2, 1))) if np.argmax(prediction[0]) == 0: return (1, "Access Granted!") else: return (0, "Access Denied.")
null
9,881
import nltk nltk.download('punkt') def tokenizeCode(someCode): tokenDict = { 'aaa': 1, 'bbb': 2 } tokenizer = nltk.tokenize.MWETokenizer() tokens = tokenizer.tokenize(nltk.word_tokenize(someCode)) indexedTokens = [] for token in tokens: indexedTokens.append(tokenDict.get(token, 0)) return indexedTokens
null
9,882
import numpy as np import nltk from keras.preprocessing import sequence from keras.models import Sequential from keras.layers import Dense, Dropout, Activation from keras.layers import Embedding from keras.layers import Conv1D, GlobalMaxPooling1D from keras.datasets import imdb nltk.download('punkt') def tokenizeCode(codeSnippet): tokenDict = { 'printf': 1, 'scanf': 2, '%_d': 3, '%_s': 4, '``': 5, # NLTK beginning qutation marks "''": 6, # NLTK ending qutation marks ',': 7, '(': 8, ')': 9, ';': 10 } tokenizer = nltk.tokenize.MWETokenizer() tokenizer.add_mwe(('%', 'd')) tokenizer.add_mwe(('%', 'n')) tokens = tokenizer.tokenize(nltk.word_tokenize(codeSnippet)) indexedTokens = [] for token in tokens: indexedTokens.append(tokenDict.get(token, 0)) return indexedTokens
null
9,883
from keras.models import Model, load_model from keras.layers import Input import numpy as np import keras target_token_index = np.load('./target_tokens.npy', allow_pickle=True).item() num_decoder_tokens = len(target_token_index) max_decoder_seq_length = 53 encoder_model = Model(encoder_inputs, encoder_states) decoder_model = Model( [decoder_inputs] + decoder_states_inputs, [decoder_outputs] + decoder_states) reverse_target_char_index = dict( (i, char) for char, i in target_token_index.items()) def decode_sequence(input_seq): # Encode the input as state vectors. states_value = encoder_model.predict(input_seq) # Generate empty target sequence of length 1. target_seq = np.zeros((1, 1, num_decoder_tokens)) # Populate the first character of target sequence with the start character. target_seq[0, 0, target_token_index['\t']] = 1. # Sampling loop for a batch of sequences # (to simplify, here we assume a batch of size 1). stop_condition = False decoded_sentence = '' while not stop_condition: output_tokens, h, c = decoder_model.predict( [target_seq] + states_value) # Sample a token sampled_token_index = np.argmax(output_tokens[0, -1, :]) sampled_char = reverse_target_char_index[sampled_token_index] decoded_sentence += sampled_char # Exit condition: either hit max length # or find stop character. if (sampled_char == '\n' or len(decoded_sentence) > max_decoder_seq_length): stop_condition = True # Update the target sequence (of length 1). target_seq = np.zeros((1, 1, num_decoder_tokens)) target_seq[0, 0, sampled_token_index] = 1. # Update states states_value = [h, c] return decoded_sentence
null
9,885
from __future__ import print_function from keras.models import Model from keras.layers import Input, LSTM, Dense import numpy as np num_decoder_tokens = len(target_characters) max_decoder_seq_length = max([len(txt) for txt in target_texts]) target_token_index = dict( [(char, i) for i, char in enumerate(target_characters)]) encoder_model = Model(encoder_inputs, encoder_states) decoder_model = Model( [decoder_inputs] + decoder_states_inputs, [decoder_outputs] + decoder_states) reverse_target_char_index = dict( (i, char) for char, i in target_token_index.items()) def decode_sequence(input_seq): # Encode the input as state vectors. states_value = encoder_model.predict(input_seq) # Generate empty target sequence of length 1. target_seq = np.zeros((1, 1, num_decoder_tokens)) # Populate the first character of target sequence with the start character. target_seq[0, 0, target_token_index['\t']] = 1. # Sampling loop for a batch of sequences # (to simplify, here we assume a batch of size 1). stop_condition = False decoded_sentence = '' while not stop_condition: output_tokens, h, c = decoder_model.predict( [target_seq] + states_value) # Sample a token sampled_token_index = np.argmax(output_tokens[0, -1, :]) sampled_char = reverse_target_char_index[sampled_token_index] decoded_sentence += sampled_char # Exit condition: either hit max length # or find stop character. if (sampled_char == '\n' or len(decoded_sentence) > max_decoder_seq_length): stop_condition = True # Update the target sequence (of length 1). target_seq = np.zeros((1, 1, num_decoder_tokens)) target_seq[0, 0, sampled_token_index] = 1. # Update states states_value = [h, c] return decoded_sentence
null
9,886
from __future__ import print_function from keras.models import Model, load_model from keras.layers import Input from difflib import SequenceMatcher import numpy as np num_decoder_tokens = len(target_characters) max_decoder_seq_length = max([len(txt) for txt in target_texts]) target_token_index = dict( [(char, i) for i, char in enumerate(target_characters)]) encoder_model = Model(encoder_inputs, encoder_states) decoder_model = Model( [decoder_inputs] + decoder_states_inputs, [decoder_outputs] + decoder_states) reverse_target_char_index = dict( (i, char) for char, i in target_token_index.items()) def decode_sequence(input_seq): # Encode the input as state vectors. states_value = encoder_model.predict(input_seq) # Generate empty target sequence of length 1. target_seq = np.zeros((1, 1, num_decoder_tokens)) # Populate the first character of target sequence with the start character. target_seq[0, 0, target_token_index['\t']] = 1. # Sampling loop for a batch of sequences # (to simplify, here we assume a batch of size 1). stop_condition = False decoded_sentence = '' while not stop_condition: output_tokens, h, c = decoder_model.predict( [target_seq] + states_value) # Sample a token sampled_token_index = np.argmax(output_tokens[0, -1, :]) sampled_char = reverse_target_char_index[sampled_token_index] decoded_sentence += sampled_char # Exit condition: either hit max length # or find stop character. if (sampled_char == '\n' or len(decoded_sentence) > max_decoder_seq_length): stop_condition = True # Update the target sequence (of length 1). target_seq = np.zeros((1, 1, num_decoder_tokens)) target_seq[0, 0, sampled_token_index] = 1. # Update states states_value = [h, c] return decoded_sentence
null
9,887
from collections import defaultdict from pathlib import Path from typing import Any, NamedTuple from dns.resolver import Resolver, NoAnswer, NXDOMAIN, LifetimeTimeout, NoNameservers from dns.rdatatype import RdataType from time import sleep import logging The provided code snippet includes necessary dependencies for implementing the `domain_has_ip` function. Write a Python function `def domain_has_ip(resolver, domain)` to solve the following problem: Return true if the domain has at least one IP (IPv4 or IPv6) Here is the function: def domain_has_ip(resolver, domain): """ Return true if the domain has at least one IP (IPv4 or IPv6)""" len_dns_a = 0 len_dns_aaaa = 0 try: dns_response = resolver.resolve(domain, RdataType.A) len_dns_a = len(dns_response.rrset) except (NoAnswer, NXDOMAIN, LifetimeTimeout, NoNameservers) as e: # No response for this domain pass try: dns_response = resolver.resolve(domain, RdataType.AAAA) len_dns_aaaa = len(dns_response.rrset) except (NoAnswer, NXDOMAIN, LifetimeTimeout, NoNameservers) as e: # No response for this domain pass return len_dns_a + len_dns_aaaa > 0
Return true if the domain has at least one IP (IPv4 or IPv6)
9,888
from collections import defaultdict from pathlib import Path from typing import Any, NamedTuple from dns.resolver import Resolver, NoAnswer, NXDOMAIN, LifetimeTimeout, NoNameservers from dns.rdatatype import RdataType from time import sleep import logging def domain_has_ip(*args, **kwargs): from random import random return random() < 0.9
null
9,889
from collections import defaultdict from pathlib import Path from typing import Any, NamedTuple from dns.resolver import Resolver, NoAnswer, NXDOMAIN, LifetimeTimeout, NoNameservers from dns.rdatatype import RdataType from time import sleep import logging def md_link(content: str, href: str): return f"[{content}]({href})"
null
9,890
from collections import defaultdict from pathlib import Path from typing import Any, NamedTuple from dns.resolver import Resolver, NoAnswer, NXDOMAIN, LifetimeTimeout, NoNameservers from dns.rdatatype import RdataType from time import sleep import logging NEW_LINE = "\n" def md_tr(*td: str): return "|".join(("", *td, "")) + NEW_LINE
null
9,891
from pathlib import Path from contextlib import ExitStack from functools import lru_cache def regex_to_domain(url): url = left_replace(url, "*://", "", 1) url = left_replace(url, "*.", "", 1) url = right_replace(url, "/*", "", 1) return url def to_domain_ublock(url): formated_url = regex_to_domain(url) if "/" in formated_url: return f"||{formated_url}$all" else: # We can use this syntax which is more optimized return f"||{formated_url}^$all"
null
9,892
from pathlib import Path from contextlib import ExitStack from functools import lru_cache def regex_to_domain(url): url = left_replace(url, "*://", "", 1) url = left_replace(url, "*.", "", 1) url = right_replace(url, "/*", "", 1) return url def to_domain_ublacklist(url): formated_url = regex_to_domain(url) return f"*://*.{formated_url}/*"
null
9,893
from pathlib import Path from contextlib import ExitStack from functools import lru_cache def regex_to_domain(url): url = left_replace(url, "*://", "", 1) url = left_replace(url, "*.", "", 1) url = right_replace(url, "/*", "", 1) return url def to_domain_hosts_filter(url): formated_url = regex_to_domain(url) return f"0.0.0.0 {formated_url}"
null
9,894
from pathlib import Path from contextlib import ExitStack from functools import lru_cache def regex_to_domain(url): def to_google(url): return f'google.*###rso .MjjYud a[href*="{regex_to_domain(url)}"]:upward(.MjjYud)'
null
9,895
from pathlib import Path from contextlib import ExitStack from functools import lru_cache def regex_to_domain(url): url = left_replace(url, "*://", "", 1) url = left_replace(url, "*.", "", 1) url = right_replace(url, "/*", "", 1) return url def to_duckduckgo(url): return f'duckduckgo.com##.react-results--main > li:has(a[href*="{regex_to_domain(url)}"])'
null
9,896
from pathlib import Path from contextlib import ExitStack from functools import lru_cache def regex_to_domain(url): url = left_replace(url, "*://", "", 1) url = left_replace(url, "*.", "", 1) url = right_replace(url, "/*", "", 1) return url def to_brave(url): return f'search.brave.com###results > div:has(a[href*="{regex_to_domain(url)}"])'
null
9,897
from pathlib import Path from contextlib import ExitStack from functools import lru_cache def regex_to_domain(url): url = left_replace(url, "*://", "", 1) url = left_replace(url, "*.", "", 1) url = right_replace(url, "/*", "", 1) return url def to_startpage(url): return f'startpage.com##.w-gl__result:has(a[href*="{regex_to_domain(url)}"])'
null
9,898
from pathlib import Path from contextlib import ExitStack from functools import lru_cache def regex_to_domain(url): url = left_replace(url, "*://", "", 1) url = left_replace(url, "*.", "", 1) url = right_replace(url, "/*", "", 1) return url def to_ecosia(url): return f'ecosia.org###main .result:has(a[href*="{regex_to_domain(url)}"])'
null
9,899
from pathlib import Path from contextlib import ExitStack from functools import lru_cache def to_domain_attr(url): return url \ .replace("*://", "") \ .replace("*.", ".") \ .replace("/*", "") \ .lstrip(".") def to_userscript(url): return f'[data-domain*="{to_domain_attr(url)}"]'
null
9,900
from pathlib import Path from contextlib import ExitStack from functools import lru_cache def append_in_se(fd_by_filter, filter_name, source_is_for_dev, value): fd_by_filter[filter_name]["current"].write(value) fd_by_filter[filter_name]["global"].write(value) if source_is_for_dev: # Add in the "all" filter # The dev filter was formerly called "all". Dont rename it for compatibility fd_by_filter[filter_name]["all"].write(value)
null
9,901
from pathlib import Path from contextlib import ExitStack from functools import lru_cache def get_userscript_start(name): return f"""// ==UserScript== // @name uBlock-Origin-dev-filter – {name} // @description Filter copycat-websites from DuckDuckGo and Google // @match https://*.duckduckgo.com/* // @include https://*.google.*/* // ==/UserScript== (function() {{ const css = ` """
null
9,902
from pathlib import Path from contextlib import ExitStack from functools import lru_cache def get_userscript_end(): return """#__non-existent__{display: none}`; if (document.location.hostname.includes('google')) { const domains = css .split('\\n') .map( (s) => s.slice(15).replace('"],', '').trim(), ) .filter(Boolean); // Remove {display:none} domains.splice(domains.length - 1, 1); for (const domain of domains) { try { const p = document .querySelector(`#search a[href*="${domain}"]`) .parentNode.parentNode.parentNode.parentNode; if (p) { p.parentNode.removeChild(p); } } catch (e) { // Ignore } } } else { const style = document.createElement('style'); style.textContent = css; document.head.insertAdjacentElement('beforeend', style); } })();"""
null
9,903
from pathlib import Path from contextlib import ExitStack from functools import lru_cache def get_ublock_filters_header(name): return f"""! Title: uBlock-Origin-dev-filter – {name} ! Expires: 1 day ! Description: Filters to block and remove copycat-websites from search engines. Specific to dev websites like StackOverflow or GitHub. ! Homepage: https://github.com/quenhus/uBlock-Origin-dev-filter ! Licence: https://github.com/quenhus/uBlock-Origin-dev-filter/blob/main/LICENSE ! ! GitHub issues: https://github.com/quenhus/uBlock-Origin-dev-filter/issues ! GitHub pull requests: https://github.com/quenhus/uBlock-Origin-dev-filter/pulls """
null
9,904
from pathlib import Path from contextlib import ExitStack from functools import lru_cache def get_common_filters_header(name): return f"""# Title: uBlock-Origin-dev-filter – {name} # Expires: 1 day # Description: Filters to block and remove copycat-websites from search engines. Specific to dev websites like StackOverflow or GitHub. # Homepage: https://github.com/quenhus/uBlock-Origin-dev-filter # Licence: https://github.com/quenhus/uBlock-Origin-dev-filter/blob/main/LICENSE # # GitHub issues: https://github.com/quenhus/uBlock-Origin-dev-filter/issues # GitHub pull requests: https://github.com/quenhus/uBlock-Origin-dev-filter/pulls """
null
9,905
import urllib.parse from typing import NamedTuple class FlavorMeta(NamedTuple): name: str table_name: str filename: str search_engines = ( FilterMeta("Google", "google", "de3f32"), FilterMeta("DuckDuckGo", "duckduckgo", "fdd20a"), FilterMeta("Google+DDG", "google_duckduckgo", "9b59b6"), FilterMeta("Startpage", "startpage", "5b7bca"), FilterMeta("Brave", "brave", "f25100"), FilterMeta("Ecosia", "ecosia", "36acb8"), FilterMeta("All Search Engines", "all_search_engines", "ffffff") ) def get_badge(alt: str, icon: str, label: str, message: str, color: str): return f"![{alt}](https://img.shields.io/static/v1?label={param_encode(label)}&message={param_encode(message)}&color={color}&style=flat&logo={param_encode(icon)})" def md_link(content: str, href: str): return f"[{content}]({href})" def md_tr(*td: str): return "|".join(("", *td, "")) + NEW_LINE def get_ubo_subscribe_url(dist_path: str, filename: str, title: str): return f"https://subscribe.adblockplus.org/?location=https%3A%2F%2Fraw.githubusercontent.com%2Fquenhus%2FuBlock-Origin-dev-filter%2Fmain%2Fdist%2F{dist_path}%2F{filename}.txt&title={param_encode(title)}" def get_main_ubo_table(flavors: list[FlavorMeta]): ret = md_tr("", *(f.table_name for f in flavors)) ret += md_tr("---", *(":---:" for f in flavors)) for filter_meta in search_engines: ret += md_tr( filter_meta.name, *( md_link( get_badge("uBO - add this filter", "uBlock Origin", "uBO", "add this filter", filter_meta.color), get_ubo_subscribe_url(filter_meta.dist_path, f.filename, f"uBlock-Origin-dev-filter - {filter_meta.name} - {f.name}") ) for f in flavors ) ) return ret
null
9,906
import urllib.parse from typing import NamedTuple class FlavorMeta(NamedTuple): name: str table_name: str filename: str search_engines = ( FilterMeta("Google", "google", "de3f32"), FilterMeta("DuckDuckGo", "duckduckgo", "fdd20a"), FilterMeta("Google+DDG", "google_duckduckgo", "9b59b6"), FilterMeta("Startpage", "startpage", "5b7bca"), FilterMeta("Brave", "brave", "f25100"), FilterMeta("Ecosia", "ecosia", "36acb8"), FilterMeta("All Search Engines", "all_search_engines", "ffffff") ) def md_link(content: str, href: str): return f"[{content}]({href})" def md_tr(*td: str): return "|".join(("", *td, "")) + NEW_LINE def get_ubo_subscribe_url(dist_path: str, filename: str, title: str): return f"https://subscribe.adblockplus.org/?location=https%3A%2F%2Fraw.githubusercontent.com%2Fquenhus%2FuBlock-Origin-dev-filter%2Fmain%2Fdist%2F{dist_path}%2F{filename}.txt&title={param_encode(title)}" def get_source_flavor_ubo_table(flavors: list[FlavorMeta]): ret = md_tr("", *(f.table_name for f in flavors)) ret += md_tr("---", *(":---:" for f in flavors)) for filter_meta in search_engines: ret += md_tr( filter_meta.name, *( md_link( "add in uBO", get_ubo_subscribe_url(filter_meta.dist_path, f.filename, f"uBlock-Origin-dev-filter - {filter_meta.name} - {f.name}") ) for f in flavors ) ) return ret
null
9,907
import urllib.parse from typing import NamedTuple class FlavorMeta(NamedTuple): name: str table_name: str filename: str other_filters = ( FilterMeta("uBlacklist", "other_format/uBlacklist", "ffffff"), FilterMeta("macOS userscript", "userscript/google_duckduckgo", "ffffff"), FilterMeta("Domains filter", "other_format/domains", "ffffff"), FilterMeta("DNS hosts filter", "other_format/hosts", "ffffff"), ) def md_link(content: str, href: str): return f"[{content}]({href})" def md_tr(*td: str): return "|".join(("", *td, "")) + NEW_LINE def get_static_url(dist_path: str, filename: str): return f"https://raw.githubusercontent.com/quenhus/uBlock-Origin-dev-filter/main/dist/{dist_path}/{filename}.txt" def get_other_filter_table(flavors: list[FlavorMeta]): ret = md_tr("", *(f.table_name for f in flavors)) ret += md_tr("---", *(":---:" for f in flavors)) for filter_meta in other_filters: ret += md_tr( filter_meta.name, *( md_link( "Link", get_static_url(filter_meta.dist_path, f.filename) ) for f in flavors ) ) return ret
null
9,908
import glob import os import torch from setuptools import find_packages from setuptools import setup from torch.utils.cpp_extension import CUDA_HOME from torch.utils.cpp_extension import CppExtension from torch.utils.cpp_extension import CUDAExtension def get_extensions(): this_dir = os.path.dirname(os.path.abspath(__file__)) extensions_dir = os.path.join(this_dir, "maskrcnn_benchmark", "csrc") main_file = glob.glob(os.path.join(extensions_dir, "*.cpp")) source_cpu = glob.glob(os.path.join(extensions_dir, "cpu", "*.cpp")) source_cuda = glob.glob(os.path.join(extensions_dir, "cuda", "*.cu")) sources = main_file + source_cpu extension = CppExtension extra_compile_args = {"cxx": []} define_macros = [] if torch.cuda.is_available() and CUDA_HOME is not None: extension = CUDAExtension sources += source_cuda define_macros += [("WITH_CUDA", None)] extra_compile_args["nvcc"] = [ "-DCUDA_HAS_FP16=1", "-D__CUDA_NO_HALF_OPERATORS__", "-D__CUDA_NO_HALF_CONVERSIONS__", "-D__CUDA_NO_HALF2_OPERATORS__", ] sources = [os.path.join(extensions_dir, s) for s in sources] include_dirs = [extensions_dir] ext_modules = [ extension( "maskrcnn_benchmark._C", sources, include_dirs=include_dirs, define_macros=define_macros, extra_compile_args=extra_compile_args, ) ] return ext_modules
null
9,909
import torch The provided code snippet includes necessary dependencies for implementing the `_onnx_clip_boxes_to_image` function. Write a Python function `def _onnx_clip_boxes_to_image(boxes, size)` to solve the following problem: Clip boxes so that they lie inside an image of size `size`. Clip's min max are traced as constants. Use torch.min/max to WAR this issue Arguments: boxes (Tensor[N, 4]): boxes in (x1, y1, x2, y2) format size (Tuple[height, width]): size of the image Returns: clipped_boxes (Tensor[N, 4]) Here is the function: def _onnx_clip_boxes_to_image(boxes, size): # type: (Tensor, Tuple[int, int]) """ Clip boxes so that they lie inside an image of size `size`. Clip's min max are traced as constants. Use torch.min/max to WAR this issue Arguments: boxes (Tensor[N, 4]): boxes in (x1, y1, x2, y2) format size (Tuple[height, width]): size of the image Returns: clipped_boxes (Tensor[N, 4]) """ TO_REMOVE = 1 device = boxes.device dim = boxes.dim() boxes_x = boxes[..., 0::2] boxes_y = boxes[..., 1::2] boxes_x = torch.max(boxes_x, torch.tensor(0., dtype=torch.float).to(device)) boxes_x = torch.min(boxes_x, torch.tensor(size[1] - TO_REMOVE, dtype=torch.float).to(device)) boxes_y = torch.max(boxes_y, torch.tensor(0., dtype=torch.float).to(device)) boxes_y = torch.min(boxes_y, torch.tensor(size[0] - TO_REMOVE, dtype=torch.float).to(device)) clipped_boxes = torch.stack((boxes_x, boxes_y), dim=dim) return clipped_boxes.reshape(boxes.shape)
Clip boxes so that they lie inside an image of size `size`. Clip's min max are traced as constants. Use torch.min/max to WAR this issue Arguments: boxes (Tensor[N, 4]): boxes in (x1, y1, x2, y2) format size (Tuple[height, width]): size of the image Returns: clipped_boxes (Tensor[N, 4])
9,910
import torch from .bounding_box import BoxList from maskrcnn_benchmark.layers import nms as _box_nms from maskrcnn_benchmark.layers import ml_nms as _box_ml_nms The provided code snippet includes necessary dependencies for implementing the `boxlist_ml_nms` function. Write a Python function `def boxlist_ml_nms(boxlist, nms_thresh, max_proposals=-1, score_field="scores", label_field="labels")` to solve the following problem: Performs non-maximum suppression on a boxlist, with scores specified in a boxlist field via score_field. Arguments: boxlist(BoxList) nms_thresh (float) max_proposals (int): if > 0, then only the top max_proposals are kept after non-maximum suppression score_field (str) Here is the function: def boxlist_ml_nms(boxlist, nms_thresh, max_proposals=-1, score_field="scores", label_field="labels"): """ Performs non-maximum suppression on a boxlist, with scores specified in a boxlist field via score_field. Arguments: boxlist(BoxList) nms_thresh (float) max_proposals (int): if > 0, then only the top max_proposals are kept after non-maximum suppression score_field (str) """ if nms_thresh <= 0: return boxlist mode = boxlist.mode boxlist = boxlist.convert("xyxy") boxes = boxlist.bbox scores = boxlist.get_field(score_field) labels = boxlist.get_field(label_field) if boxes.device==torch.device("cpu"): keep = [] unique_labels = torch.unique(labels) print(unique_labels) for j in unique_labels: inds = (labels == j).nonzero().view(-1) scores_j = scores[inds] boxes_j = boxes[inds, :].view(-1, 4) keep_j = _box_nms(boxes_j, scores_j, nms_thresh) keep += keep_j else: keep = _box_ml_nms(boxes, scores, labels.float(), nms_thresh) if max_proposals > 0: keep = keep[: max_proposals] boxlist = boxlist[keep] return boxlist.convert(mode)
Performs non-maximum suppression on a boxlist, with scores specified in a boxlist field via score_field. Arguments: boxlist(BoxList) nms_thresh (float) max_proposals (int): if > 0, then only the top max_proposals are kept after non-maximum suppression score_field (str)
9,911
import torch from .bounding_box import BoxList from maskrcnn_benchmark.layers import nms as _box_nms from maskrcnn_benchmark.layers import ml_nms as _box_ml_nms The provided code snippet includes necessary dependencies for implementing the `remove_small_boxes` function. Write a Python function `def remove_small_boxes(boxlist, min_size)` to solve the following problem: Only keep boxes with both sides >= min_size Arguments: boxlist (Boxlist) min_size (int) Here is the function: def remove_small_boxes(boxlist, min_size): """ Only keep boxes with both sides >= min_size Arguments: boxlist (Boxlist) min_size (int) """ # WORK AROUND: work around unbind using split + squeeze. xywh_boxes = boxlist.convert("xywh").bbox _, _, ws, hs = xywh_boxes.split(1, dim=1) ws = ws.squeeze(1) hs = hs.squeeze(1) keep = ((ws >= min_size) & (hs >= min_size)).nonzero().squeeze(1) return boxlist[keep]
Only keep boxes with both sides >= min_size Arguments: boxlist (Boxlist) min_size (int)
9,912
import torch The provided code snippet includes necessary dependencies for implementing the `smooth_l1_loss` function. Write a Python function `def smooth_l1_loss(input, target, beta=1. / 9, size_average=True)` to solve the following problem: very similar to the smooth_l1_loss from pytorch, but with the extra beta parameter Here is the function: def smooth_l1_loss(input, target, beta=1. / 9, size_average=True): """ very similar to the smooth_l1_loss from pytorch, but with the extra beta parameter """ n = torch.abs(input - target) cond = n < beta loss = torch.where(cond, 0.5 * n ** 2 / beta, n - 0.5 * beta) if size_average: return loss.mean() return loss.sum()
very similar to the smooth_l1_loss from pytorch, but with the extra beta parameter
9,913
import torch from torch import nn import torch.nn.functional as F from torch.autograd import Function from torch.autograd.function import once_differentiable from maskrcnn_benchmark import _C def sigmoid_focal_loss_cpu(logits, targets, gamma, alpha): num_classes = logits.shape[1] dtype = targets.dtype device = targets.device class_range = torch.arange(1, num_classes + 1, dtype=dtype, device=device).unsqueeze(0) t = targets.unsqueeze(1) p = torch.sigmoid(logits) term1 = (1 - p) ** gamma * torch.log(p) term2 = p ** gamma * torch.log(1 - p) return -(t == class_range).float() * term1 * alpha - ((t != class_range) * (t >= 0)).float() * term2 * (1 - alpha)
null
9,914
import torch from torch import nn import torch.nn.functional as F from torch.autograd import Function from torch.autograd.function import once_differentiable from maskrcnn_benchmark import _C def token_sigmoid_softmax_focal_loss(pred_logits, targets, alpha, gamma, text_mask=None): # Another modification is that because we use the cross entropy version, there is no frequent or not frequent class. # So we temporarily retired the design of alpha. assert (targets.dim() == 3) assert (pred_logits.dim() == 3) # batch x from x to # reprocess target to become probability map ready for softmax targets = targets.float() target_num = targets.sum(-1) + 1e-8 # numerical stability targets = targets / target_num.unsqueeze(-1) # T(x) if text_mask is not None: # reserve the last token for non object assert (text_mask.dim() == 2) text_mask[:, -1] = 1 text_mask = (text_mask > 0).unsqueeze(1).repeat(1, pred_logits.size(1), 1) # copy along the image channel pred_logits = pred_logits.masked_fill(~text_mask, -1000000) # softmax out_prob = pred_logits.softmax(-1) filled_targets = targets.clone() filled_targets[filled_targets == 0] = 1.0 weight = torch.clamp(targets - out_prob, min=0.001) / filled_targets weight = torch.pow(weight, gamma) # weight = torch.pow(torch.clamp(target - out_prob, min=0.01), gamma) loss_ce = - targets * weight * pred_logits.log_softmax( -1) # only those positives with positive target_sim will have losses. return loss_ce
null
9,915
import torch from torch import nn import torch.nn.functional as F from torch.autograd import Function from torch.autograd.function import once_differentiable from maskrcnn_benchmark import _C def token_sigmoid_binary_focal_loss_v2(pred_logits, targets, alpha, gamma, text_mask=None): assert (targets.dim() == 3) assert (pred_logits.dim() == 3) # batch x from x to if text_mask is not None: assert (text_mask.dim() == 2) # We convert everything into binary out_prob = pred_logits.sigmoid() out_prob_neg_pos = torch.stack([1 - out_prob, out_prob], dim=-1) + 1e-8 # batch x boxes x 256 x 2 weight = torch.pow(-out_prob_neg_pos + 1.0, gamma) focal_zero = - weight[:, :, :, 0] * torch.log(out_prob_neg_pos[:, :, :, 0]) * ( 1 - alpha) # negative class focal_one = - weight[:, :, :, 1] * torch.log(out_prob_neg_pos[:, :, :, 1]) * alpha # positive class focal = torch.stack([focal_zero, focal_one], dim=-1) loss_ce = torch.gather(focal, index=targets.long().unsqueeze(-1), dim=-1) return loss_ce
null
9,916
import torch from torch import nn import torch.nn.functional as F from torch.autograd import Function from torch.autograd.function import once_differentiable from maskrcnn_benchmark import _C The provided code snippet includes necessary dependencies for implementing the `token_sigmoid_binary_focal_loss` function. Write a Python function `def token_sigmoid_binary_focal_loss(pred_logits, targets, alpha, gamma, text_mask=None)` to solve the following problem: Loss used in RetinaNet for dense detection: https://arxiv.org/abs/1708.02002. Args: inputs: A float tensor of arbitrary shape. The predictions for each example. targets: A float tensor with the same shape as inputs. Stores the binary classification label for each element in inputs (0 for the negative class and 1 for the positive class). alpha: (optional) Weighting factor in range (0,1) to balance positive vs negative examples. Default = -1 (no weighting). gamma: Exponent of the modulating factor (1 - p_t) to balance easy vs hard examples. Returns: Loss tensor with the reduction option applied. Here is the function: def token_sigmoid_binary_focal_loss(pred_logits, targets, alpha, gamma, text_mask=None): # binary version of focal loss # copied from https://github.com/facebookresearch/fvcore/blob/master/fvcore/nn/focal_loss.py """ Loss used in RetinaNet for dense detection: https://arxiv.org/abs/1708.02002. Args: inputs: A float tensor of arbitrary shape. The predictions for each example. targets: A float tensor with the same shape as inputs. Stores the binary classification label for each element in inputs (0 for the negative class and 1 for the positive class). alpha: (optional) Weighting factor in range (0,1) to balance positive vs negative examples. Default = -1 (no weighting). gamma: Exponent of the modulating factor (1 - p_t) to balance easy vs hard examples. Returns: Loss tensor with the reduction option applied. """ assert (targets.dim() == 3) assert (pred_logits.dim() == 3) # batch x from x to bs, n, _ = pred_logits.shape if text_mask is not None: assert (text_mask.dim() == 2) text_mask = (text_mask > 0).unsqueeze(1) text_mask = text_mask.repeat(1, pred_logits.size(1), 1) # copy along the image channel dimension pred_logits = torch.masked_select(pred_logits, text_mask) targets = torch.masked_select(targets, text_mask) # print(pred_logits.shape) # print(targets.shape) p = torch.sigmoid(pred_logits) ce_loss = F.binary_cross_entropy_with_logits(pred_logits, targets, reduction="none") p_t = p * targets + (1 - p) * (1 - targets) loss = ce_loss * ((1 - p_t) ** gamma) if alpha >= 0: alpha_t = alpha * targets + (1 - alpha) * (1 - targets) loss = alpha_t * loss return loss
Loss used in RetinaNet for dense detection: https://arxiv.org/abs/1708.02002. Args: inputs: A float tensor of arbitrary shape. The predictions for each example. targets: A float tensor with the same shape as inputs. Stores the binary classification label for each element in inputs (0 for the negative class and 1 for the positive class). alpha: (optional) Weighting factor in range (0,1) to balance positive vs negative examples. Default = -1 (no weighting). gamma: Exponent of the modulating factor (1 - p_t) to balance easy vs hard examples. Returns: Loss tensor with the reduction option applied.
9,917
import torch import torch.nn as nn import torch.nn.functional as F def _make_divisible(v, divisor, min_value=None): if min_value is None: min_value = divisor new_v = max(min_value, int(v + divisor / 2) // divisor * divisor) # Make sure that round down does not go down by more than 10%. if new_v < 0.9 * v: new_v += divisor return new_v
null
9,918
import math import torch from torch.nn.modules.utils import _ntuple class _NewEmptyTensorOp(torch.autograd.Function): def forward(ctx, x, new_shape): ctx.shape = x.shape return x.new_empty(new_shape) def backward(ctx, grad): shape = ctx.shape return _NewEmptyTensorOp.apply(grad, shape), None def interpolate( input, size=None, scale_factor=None, mode="nearest", align_corners=None ): if input.numel() > 0: return torch.nn.functional.interpolate( input, size, scale_factor, mode, align_corners ) def _check_size_scale_factor(dim): if size is None and scale_factor is None: raise ValueError("either size or scale_factor should be defined") if size is not None and scale_factor is not None: raise ValueError("only one of size or scale_factor should be defined") if ( scale_factor is not None and isinstance(scale_factor, tuple) and len(scale_factor) != dim ): raise ValueError( "scale_factor shape must match input shape. " "Input is {}D, scale_factor size is {}".format(dim, len(scale_factor)) ) def _output_size(dim): _check_size_scale_factor(dim) if size is not None: return size scale_factors = _ntuple(dim)(scale_factor) # math.floor might return float in py2.7 return [ int(math.floor(input.size(i + 2) * scale_factors[i])) for i in range(dim) ] output_shape = tuple(_output_size(2)) output_shape = input.shape[:-2] + output_shape return _NewEmptyTensorOp.apply(input, output_shape)
null
9,919
import torch import torch.nn.functional as F import torch.distributed as dist from torch import nn from scipy.optimize import linear_sum_assignment from torch.cuda.amp import custom_fwd, custom_bwd def box_iou(boxes1, boxes2): area1 = box_area(boxes1) area2 = box_area(boxes2) lt = torch.max(boxes1[:, None, :2], boxes2[:, :2]) # [N,M,2] rb = torch.min(boxes1[:, None, 2:], boxes2[:, 2:]) # [N,M,2] wh = (rb - lt).clamp(min=0) # [N,M,2] inter = wh[:, :, 0] * wh[:, :, 1] # [N,M] union = area1[:, None] + area2 - inter iou = inter / union return iou, union The provided code snippet includes necessary dependencies for implementing the `generalized_box_iou` function. Write a Python function `def generalized_box_iou(boxes1, boxes2)` to solve the following problem: Generalized IoU from https://giou.stanford.edu/ The boxes should be in [x0, y0, x1, y1] format Returns a [N, M] pairwise matrix, where N = len(boxes1) and M = len(boxes2) Here is the function: def generalized_box_iou(boxes1, boxes2): """ Generalized IoU from https://giou.stanford.edu/ The boxes should be in [x0, y0, x1, y1] format Returns a [N, M] pairwise matrix, where N = len(boxes1) and M = len(boxes2) """ # degenerate boxes gives inf / nan results # so do an early check #assert (boxes1[:, 2:] >= boxes1[:, :2]).all() #assert (boxes2[:, 2:] >= boxes2[:, :2]).all() iou, union = box_iou(boxes1, boxes2) lt = torch.min(boxes1[:, None, :2], boxes2[:, :2]) rb = torch.max(boxes1[:, None, 2:], boxes2[:, 2:]) wh = (rb - lt).clamp(min=0) # [N,M,2] area = wh[:, :, 0] * wh[:, :, 1] return iou - (area - union) / area
Generalized IoU from https://giou.stanford.edu/ The boxes should be in [x0, y0, x1, y1] format Returns a [N, M] pairwise matrix, where N = len(boxes1) and M = len(boxes2)
9,920
import torch import torch.nn.functional as F import torch.distributed as dist from torch import nn from scipy.optimize import linear_sum_assignment from torch.cuda.amp import custom_fwd, custom_bwd The provided code snippet includes necessary dependencies for implementing the `dice_loss` function. Write a Python function `def dice_loss(inputs, targets, num_boxes)` to solve the following problem: Compute the DICE loss, similar to generalized IOU for masks Args: inputs: A float tensor of arbitrary shape. The predictions for each example. targets: A float tensor with the same shape as inputs. Stores the binary classification label for each element in inputs (0 for the negative class and 1 for the positive class). Here is the function: def dice_loss(inputs, targets, num_boxes): """ Compute the DICE loss, similar to generalized IOU for masks Args: inputs: A float tensor of arbitrary shape. The predictions for each example. targets: A float tensor with the same shape as inputs. Stores the binary classification label for each element in inputs (0 for the negative class and 1 for the positive class). """ inputs = inputs.sigmoid() inputs = inputs.flatten(1) numerator = 2 * (inputs * targets).sum(1) denominator = inputs.sum(-1) + targets.sum(-1) loss = 1 - (numerator + 1) / (denominator + 1) return loss.sum() / num_boxes
Compute the DICE loss, similar to generalized IOU for masks Args: inputs: A float tensor of arbitrary shape. The predictions for each example. targets: A float tensor with the same shape as inputs. Stores the binary classification label for each element in inputs (0 for the negative class and 1 for the positive class).
9,921
import torch import torch.nn.functional as F import torch.distributed as dist from torch import nn from scipy.optimize import linear_sum_assignment from torch.cuda.amp import custom_fwd, custom_bwd The provided code snippet includes necessary dependencies for implementing the `sigmoid_focal_loss` function. Write a Python function `def sigmoid_focal_loss(inputs: torch.Tensor, targets: torch.Tensor, alpha: float = -1, gamma: float = 2, reduction: str = "none")` to solve the following problem: Loss used in RetinaNet for dense detection: https://arxiv.org/abs/1708.02002. Args: inputs: A float tensor of arbitrary shape. The predictions for each example. targets: A float tensor with the same shape as inputs. Stores the binary classification label for each element in inputs (0 for the negative class and 1 for the positive class). alpha: (optional) Weighting factor in range (0,1) to balance positive vs negative examples. Default = -1 (no weighting). gamma: Exponent of the modulating factor (1 - p_t) to balance easy vs hard examples. reduction: 'none' | 'mean' | 'sum' 'none': No reduction will be applied to the output. 'mean': The output will be averaged. 'sum': The output will be summed. Returns: Loss tensor with the reduction option applied. Here is the function: def sigmoid_focal_loss(inputs: torch.Tensor, targets: torch.Tensor, alpha: float = -1, gamma: float = 2, reduction: str = "none"): """ Loss used in RetinaNet for dense detection: https://arxiv.org/abs/1708.02002. Args: inputs: A float tensor of arbitrary shape. The predictions for each example. targets: A float tensor with the same shape as inputs. Stores the binary classification label for each element in inputs (0 for the negative class and 1 for the positive class). alpha: (optional) Weighting factor in range (0,1) to balance positive vs negative examples. Default = -1 (no weighting). gamma: Exponent of the modulating factor (1 - p_t) to balance easy vs hard examples. reduction: 'none' | 'mean' | 'sum' 'none': No reduction will be applied to the output. 'mean': The output will be averaged. 'sum': The output will be summed. Returns: Loss tensor with the reduction option applied. """ p = torch.sigmoid(inputs) ce_loss = F.binary_cross_entropy_with_logits(inputs, targets, reduction="none") p_t = p * targets + (1 - p) * (1 - targets) loss = ce_loss * ((1 - p_t) ** gamma) if alpha >= 0: alpha_t = alpha * targets + (1 - alpha) * (1 - targets) loss = alpha_t * loss if reduction == "mean": loss = loss.mean() elif reduction == "sum": loss = loss.sum() return loss
Loss used in RetinaNet for dense detection: https://arxiv.org/abs/1708.02002. Args: inputs: A float tensor of arbitrary shape. The predictions for each example. targets: A float tensor with the same shape as inputs. Stores the binary classification label for each element in inputs (0 for the negative class and 1 for the positive class). alpha: (optional) Weighting factor in range (0,1) to balance positive vs negative examples. Default = -1 (no weighting). gamma: Exponent of the modulating factor (1 - p_t) to balance easy vs hard examples. reduction: 'none' | 'mean' | 'sum' 'none': No reduction will be applied to the output. 'mean': The output will be averaged. 'sum': The output will be summed. Returns: Loss tensor with the reduction option applied.
9,922
import torch import torch.nn as nn import torch.nn.functional as F from .deform_conv import DeformConv2d The provided code snippet includes necessary dependencies for implementing the `add_conv` function. Write a Python function `def add_conv(in_ch, out_ch, ksize, stride, leaky=True)` to solve the following problem: Add a conv2d / batchnorm / leaky ReLU block. Args: in_ch (int): number of input channels of the convolution layer. out_ch (int): number of output channels of the convolution layer. ksize (int): kernel size of the convolution layer. stride (int): stride of the convolution layer. Returns: stage (Sequential) : Sequential layers composing a convolution block. Here is the function: def add_conv(in_ch, out_ch, ksize, stride, leaky=True): """ Add a conv2d / batchnorm / leaky ReLU block. Args: in_ch (int): number of input channels of the convolution layer. out_ch (int): number of output channels of the convolution layer. ksize (int): kernel size of the convolution layer. stride (int): stride of the convolution layer. Returns: stage (Sequential) : Sequential layers composing a convolution block. """ stage = nn.Sequential() pad = (ksize - 1) // 2 stage.add_module('conv', nn.Conv2d(in_channels=in_ch, out_channels=out_ch, kernel_size=ksize, stride=stride, padding=pad, bias=False)) stage.add_module('batch_norm', nn.BatchNorm2d(out_ch)) if leaky: stage.add_module('leaky', nn.LeakyReLU(0.1)) else: stage.add_module('relu6', nn.ReLU6(inplace=True)) return stage
Add a conv2d / batchnorm / leaky ReLU block. Args: in_ch (int): number of input channels of the convolution layer. out_ch (int): number of output channels of the convolution layer. ksize (int): kernel size of the convolution layer. stride (int): stride of the convolution layer. Returns: stage (Sequential) : Sequential layers composing a convolution block.
9,923
import torch import torch.nn as nn import torch.nn.functional as F from .deform_conv import DeformConv2d The provided code snippet includes necessary dependencies for implementing the `make_divisible` function. Write a Python function `def make_divisible(v, divisor, min_value=None)` to solve the following problem: This function is taken from the original tf repo. It ensures that all layers have a channel number that is divisible by 8 It can be seen here: https://github.com/tensorflow/models/blob/master/research/slim/nets/mobilenet/mobilenet.py :param v: :param divisor: :param min_value: :return: Here is the function: def make_divisible(v, divisor, min_value=None): """ This function is taken from the original tf repo. It ensures that all layers have a channel number that is divisible by 8 It can be seen here: https://github.com/tensorflow/models/blob/master/research/slim/nets/mobilenet/mobilenet.py :param v: :param divisor: :param min_value: :return: """ if min_value is None: min_value = divisor new_v = max(min_value, int(v + divisor / 2) // divisor * divisor) # Make sure that round down does not go down by more than 10%. if new_v < 0.9 * v: new_v += divisor return new_v
This function is taken from the original tf repo. It ensures that all layers have a channel number that is divisible by 8 It can be seen here: https://github.com/tensorflow/models/blob/master/research/slim/nets/mobilenet/mobilenet.py :param v: :param divisor: :param min_value: :return:
9,924
import torch import torch.nn as nn import torch.nn.functional as F from .deform_conv import DeformConv2d def add_sepconv(in_ch, out_ch, ksize, stride): stage = nn.Sequential() pad = (ksize - 1) // 2 stage.add_module('sepconv', nn.Conv2d(in_channels=in_ch, out_channels=in_ch, kernel_size=ksize, stride=stride, padding=pad, groups=in_ch, bias=False)) stage.add_module('sepbn', nn.BatchNorm2d(in_ch)) stage.add_module('seprelu6', nn.ReLU6(inplace=True)) stage.add_module('ptconv', nn.Conv2d(in_ch, out_ch, 1, 1, 0, bias=False)) stage.add_module('ptbn', nn.BatchNorm2d(out_ch)) stage.add_module('ptrelu6', nn.ReLU6(inplace=True)) return stage
null
9,925
import cv2 import random import numpy as np import math import torch import torchvision from torchvision.transforms import functional as F from maskrcnn_benchmark.structures.bounding_box import BoxList The provided code snippet includes necessary dependencies for implementing the `matrix_iou` function. Write a Python function `def matrix_iou(a, b, relative=False)` to solve the following problem: return iou of a and b, numpy version for data augenmentation Here is the function: def matrix_iou(a, b, relative=False): """ return iou of a and b, numpy version for data augenmentation """ lt = np.maximum(a[:, np.newaxis, :2], b[:, :2]) rb = np.minimum(a[:, np.newaxis, 2:], b[:, 2:]) area_i = np.prod(rb - lt, axis=2) * (lt < rb).all(axis=2) area_a = np.prod(a[:, 2:] - a[:, :2], axis=1) area_b = np.prod(b[:, 2:] - b[:, :2], axis=1) if relative: ious = area_i / (area_b[:, np.newaxis]+1e-12) else: ious = area_i / (area_a[:, np.newaxis] + area_b - area_i+1e-12) return ious
return iou of a and b, numpy version for data augenmentation
9,926
from . import transforms as T def build_transforms(cfg, is_train=True): if is_train: if len(cfg.AUGMENT.MULT_MIN_SIZE_TRAIN)>0: min_size = cfg.AUGMENT.MULT_MIN_SIZE_TRAIN else: min_size = cfg.INPUT.MIN_SIZE_TRAIN max_size = cfg.INPUT.MAX_SIZE_TRAIN flip_horizontal_prob = cfg.AUGMENT.FLIP_PROB_TRAIN flip_vertical_prob = cfg.AUGMENT.VERTICAL_FLIP_PROB_TRAIN brightness = cfg.AUGMENT.BRIGHTNESS contrast = cfg.AUGMENT.CONTRAST saturation = cfg.AUGMENT.SATURATION hue = cfg.AUGMENT.HUE crop_prob = cfg.AUGMENT.CROP_PROB min_ious = cfg.AUGMENT.CROP_MIN_IOUS min_crop_size = cfg.AUGMENT.CROP_MIN_SIZE else: min_size = cfg.INPUT.MIN_SIZE_TEST max_size = cfg.INPUT.MAX_SIZE_TEST flip_horizontal_prob = 0.0 fix_res = cfg.INPUT.FIX_RES if cfg.INPUT.FORMAT is not '': input_format = cfg.INPUT.FORMAT elif cfg.INPUT.TO_BGR255: input_format = 'bgr255' normalize_transform = T.Normalize( mean=cfg.INPUT.PIXEL_MEAN, std=cfg.INPUT.PIXEL_STD, format=input_format ) transform = T.Compose( [ T.Resize(min_size, max_size, restrict=fix_res), T.RandomHorizontalFlip(flip_horizontal_prob), T.ToTensor(), normalize_transform, ] ) return transform
null
9,927
import numpy as np import random import re import torch import pdb import logging def sanity_check_target_after_processing(target): assert(len(target.bbox) == len(target.extra_fields["boxes"]))
null
9,928
import numpy as np import random import re import torch import pdb import logging def clean_name(name): name = re.sub(r"\(.*\)", "", name) name = re.sub(r"_", " ", name) name = re.sub(r" ", " ", name) return name The provided code snippet includes necessary dependencies for implementing the `convert_od_to_grounding_simple` function. Write a Python function `def convert_od_to_grounding_simple( target, image_id, ind_to_class, disable_shuffle=True, add_detection_prompt=False, separation_tokens=" ", caption_prompt=None)` to solve the following problem: Convert object detection data into grounding data format, on the fly. ind_to_class: {0: "__background__", 1 : "person" ...}, contiguous id Here is the function: def convert_od_to_grounding_simple( target, image_id, ind_to_class, disable_shuffle=True, add_detection_prompt=False, separation_tokens=" ", caption_prompt=None): """ Convert object detection data into grounding data format, on the fly. ind_to_class: {0: "__background__", 1 : "person" ...}, contiguous id """ def generate_sentence_from_labels(positive_label_list, negative_label_list, disable_shuffle=True): label_to_positions = {} label_list = negative_label_list + positive_label_list if not disable_shuffle: random.shuffle(label_list) assert (caption_prompt is None), "Should not specify caption_prompt when shuffle is enabled!!" # avoid potential bug if add_detection_prompt: pheso_caption = "object detection : " else: pheso_caption = "" for index, label in enumerate(label_list): if caption_prompt is not None: pheso_caption += caption_prompt[index]['prefix'] start_index = len(pheso_caption) if caption_prompt is not None: pheso_caption += clean_name(caption_prompt[index]['name']) else: pheso_caption += clean_name(ind_to_class[label]) # NOTE: slight change... end_index = len(pheso_caption) if caption_prompt is not None: pheso_caption += caption_prompt[index]['suffix'] # e.g.: pheso_caption = "cat dog", where cat is label 4, and dog is label 17 # label_to_positions: {4: (0, 3), 17: (4, 7)} label_to_positions[label] = [start_index, end_index] if index != len(label_list) - 1: pheso_caption += separation_tokens return label_to_positions, pheso_caption label_list = list(sorted(ind_to_class.keys())) # do not include the background label_to_positions, pheso_caption = generate_sentence_from_labels( positive_label_list=label_list, negative_label_list=[], disable_shuffle=disable_shuffle ) new_target = [] ''' Convert into: {'area': 10506.0, 'iscrowd': 0, 'image_id': 571335, 'category_id': 1, 'id': 2999421, 'bbox': [221, 319, 103, 102], 'tokens_positive': [[0, 3]]} tokens_positive is the char position ''' areas = target.area() greenlight_span_for_masked_lm_objective = [] for i in range(len(target)): new_target_i = {} new_target_i["area"] = areas[i] new_target_i["iscrowd"] = 0 new_target_i["image_id"] = image_id new_target_i["category_id"] = target.extra_fields["labels"][i].item() new_target_i["id"] = None new_target_i['bbox'] = target.bbox[i].numpy().tolist() label_i = target.extra_fields["labels"][i].item() if label_i in label_to_positions: # NOTE: Only add those that actually appear in the final caption new_target_i["tokens_positive"] = [label_to_positions[label_i]] new_target.append(new_target_i) greenlight_span_for_masked_lm_objective.append(label_to_positions[label_i]) return new_target, pheso_caption, greenlight_span_for_masked_lm_objective
Convert object detection data into grounding data format, on the fly. ind_to_class: {0: "__background__", 1 : "person" ...}, contiguous id
9,929
import numpy as np import random import re import torch import pdb import logging def clean_name(name): name = re.sub(r"\(.*\)", "", name) name = re.sub(r"_", " ", name) name = re.sub(r" ", " ", name) return name def check_for_positive_overflow(target, ind_to_class, tokenizer, max_seq_length=256): # NOTE: Only call this function for OD data; DO NOT USE IT FOR GROUNDING DATA # NOTE: called only in coco_dt # Check if we have too many positive labels # generate a caption by appending the positive labels positive_label_set = set() for i in range(len(target)): label_i = target.extra_fields["labels"][i].item() positive_label_set.add(label_i) positive_label_list = list(positive_label_set) # random shuffule so we can sample different annotations at different epochs random.shuffle(positive_label_list) kept_lables = [] length = 0 for index, label in enumerate(positive_label_list): label_text = clean_name(ind_to_class[label]) + ". " # "dog. " tokenized = tokenizer.tokenize(label_text) length += len(tokenized) if length > max_seq_length: break else: kept_lables.append(label) ## filter boxes keep_box_index = [] for i in range(len(target)): label_i = target.extra_fields["labels"][i].item() if label_i in kept_lables: keep_box_index.append(i) keep_box_index = torch.LongTensor(keep_box_index) target = target[keep_box_index] ## filter boxes return target, length
null
9,930
import numpy as np import random import re import torch import pdb import logging def clean_name(name): name = re.sub(r"\(.*\)", "", name) name = re.sub(r"_", " ", name) name = re.sub(r" ", " ", name) return name def generate_control_options_given_probabilities( control_probabilities, full_positive, full_negative): # The function was originally designed to perform data augmentation by randomly dropping negative and positive classes. Later, we decided to only consider dropping negative classes. So the returned 'num_positives' by this function will be ignored. outer_prob = random.random() probability_one_negative = control_probabilities[0] probability_one_positive = control_probabilities[1] probability_full = control_probabilities[2] probability_drop_positive = control_probabilities[3] assert(probability_drop_positive == 0) if outer_prob < probability_one_negative: # a. probability_one_negative: only give one negative class to mimic evaluation (10%) num_negatives = 1 num_positives = 0 elif outer_prob < probability_one_positive + probability_one_negative: # b. probability_one_positive: only give one positive class to mimic evaluation (10%) num_negatives = 0 num_positives = 1 elif outer_prob < probability_full + probability_one_positive + probability_one_negative: # c. probability_full: add both all positive and all negatives (20%) num_negatives = full_negative num_positives = full_positive else: if random.random() < 1.0: # - probability_random_negative: probability of randomly sample X negatives (100%) num_negatives = np.random.choice(max(1, full_negative)) + 1 # mininum 1 else: num_negatives = full_negative # Full if random.random() < probability_drop_positive: # num_positives = np.random.choice(max(1, full_positive)) + 1 else: num_positives = full_positive # Full return num_negatives, num_positives The provided code snippet includes necessary dependencies for implementing the `convert_object_detection_to_grounding_optimized_for_od` function. Write a Python function `def convert_object_detection_to_grounding_optimized_for_od( target, image_id, ind_to_class, disable_shuffle, add_detection_prompt, add_detection_prompt_advanced, random_sample_negative, control_probabilities, restricted_negative_list=None, separation_tokens=" ", max_num_labels=-1, max_seq_length=256, tokenizer=None, positive_caption_length=0 )` to solve the following problem: ind_to_class: {0: "__background__", 1 : "person" ...} target: restricted_negative_list : for datasets with restricted negatives, sample only the negatives Convert object detection data into grounding data format, on the fly. Control options: 1. add_detection_prompt: add "object detection : " to the front of the prompt 2. num_negatives: randomly sampled negative classes 3. num_positives: how many positives to keep (-1 means do not cut any) Probabilities to generate the control options: a. probability_one_negative: only give one negative class to mimic evaluation b. probability_one_positive: only give one positive class to mimic evaluation c. probability_full: add both all positive and all negatives d. other: randomly sample some negatives and some positives The below control options are independent of each other: - probability_random_negative: probability of randomly sample X negatives - probability_random_positive: probability of randomly sample some positives Here is the function: def convert_object_detection_to_grounding_optimized_for_od( target, image_id, ind_to_class, disable_shuffle, add_detection_prompt, add_detection_prompt_advanced, random_sample_negative, control_probabilities, restricted_negative_list=None, separation_tokens=" ", max_num_labels=-1, max_seq_length=256, tokenizer=None, positive_caption_length=0 ): ''' ind_to_class: {0: "__background__", 1 : "person" ...} target: restricted_negative_list : for datasets with restricted negatives, sample only the negatives Convert object detection data into grounding data format, on the fly. Control options: 1. add_detection_prompt: add "object detection : " to the front of the prompt 2. num_negatives: randomly sampled negative classes 3. num_positives: how many positives to keep (-1 means do not cut any) Probabilities to generate the control options: a. probability_one_negative: only give one negative class to mimic evaluation b. probability_one_positive: only give one positive class to mimic evaluation c. probability_full: add both all positive and all negatives d. other: randomly sample some negatives and some positives The below control options are independent of each other: - probability_random_negative: probability of randomly sample X negatives - probability_random_positive: probability of randomly sample some positives ''' if restricted_negative_list is None: valid_negative_indexes = list(ind_to_class.keys()) else: valid_negative_indexes = restricted_negative_list def generate_senetence_given_labels( positive_label_list, negative_label_list, prompt_engineer_version="v2", disable_shuffle=False, positive_question_probability=0.6, negative_question_probability=0.8, full_question_probability=0.5): ''' v3: with simple prompt such as "there are", "are there?" v4: try to merge some are there / there are together, to avoid sequence being too long ''' label_to_positions = {} assert (prompt_engineer_version == "v2") num_negatives = len(negative_label_list) num_positives = len(positive_label_list) label_list = negative_label_list + positive_label_list if not disable_shuffle: random.shuffle(label_list) if add_detection_prompt: if add_detection_prompt_advanced and (num_negatives == 0 or num_positives == 0) and not disable_shuffle: pheso_caption = "object detection query : " else: pheso_caption = "object detection : " else: pheso_caption = "" for index, label in enumerate(label_list): start_index = len(pheso_caption) pheso_caption += clean_name(ind_to_class[label]) # NOTE: slight change... end_index = len(pheso_caption) # e.g.: pheso_caption = "cat dog", where cat is label 4, and dog is label 17 # label_to_positions: {4: (0, 3), 17: (4, 7)} label_to_positions[label] = [start_index, end_index] if index != len(label_list) - 1: pheso_caption += separation_tokens return label_to_positions, pheso_caption if disable_shuffle: label_list = list(sorted(ind_to_class.keys()))[1:] # do not include the background label_to_positions, pheso_caption = generate_senetence_given_labels( positive_label_list=label_list, negative_label_list=[], disable_shuffle=True) # print(label_to_positions, pheso_caption) else: positive_label_set = set() for i in range(len(target)): label_i = target.extra_fields["labels"][i].item() positive_label_set.add(label_i) full_positive = len(positive_label_set) if max_num_labels <= 0: full_negative = random_sample_negative else: full_negative = max(min(max_num_labels-full_positive, random_sample_negative), 0) if full_negative > len(valid_negative_indexes): full_negative = len(valid_negative_indexes) num_negatives, num_positives = generate_control_options_given_probabilities( control_probabilities=control_probabilities, full_positive=full_positive, full_negative=full_negative) # num_positives not used # Keep some negatives negative_label_list = set() if num_negatives != -1: if num_negatives > len(valid_negative_indexes): num_negatives = len(valid_negative_indexes) for i in np.random.choice(valid_negative_indexes, size=num_negatives, replace=False): # label_sets.add(i) if i not in positive_label_set: negative_label_list.add(i) # Keep all positives; ignoring num_positives positive_label_list = list(positive_label_set) random.shuffle(positive_label_list) negative_label_list = list(negative_label_list) # e.g.: [17, 1, 13] where each number is the class name random.shuffle(negative_label_list) # Do a pre-screen. If we cannot afford this many negatives, we will sample less negative_max_length = max_seq_length - positive_caption_length screened_negative_label_list = [] for negative_label in negative_label_list: label_text = clean_name(ind_to_class[negative_label]) + ". " # "dog. " tokenized = tokenizer.tokenize(label_text) negative_max_length -= len(tokenized) if negative_max_length > 0: screened_negative_label_list.append(negative_label) # keep this negative else: break negative_label_list = screened_negative_label_list label_to_positions, pheso_caption = generate_senetence_given_labels( positive_label_list=positive_label_list, negative_label_list=negative_label_list) new_target = [] ''' Convert into: {'area': 10506.0, 'iscrowd': 0, 'image_id': 571335, 'category_id': 1, 'id': 2999421, 'bbox': [221, 319, 103, 102], 'tokens_positive': [[0, 3]]} tokens_positive is the char position ''' areas = target.area() greenlight_span_for_masked_lm_objective = [] for i in range(len(target)): new_target_i = {} new_target_i["area"] = areas[i] new_target_i["iscrowd"] = 0 new_target_i["image_id"] = image_id new_target_i["category_id"] = target.extra_fields["labels"][i].item() new_target_i["id"] = None new_target_i['bbox'] = target.bbox[i].numpy().tolist() label_i = target.extra_fields["labels"][i].item() new_target_i["original_od_label"] = label_i if label_i in label_to_positions: # NOTE: Only add those that actually appear in the final caption new_target_i["tokens_positive"] = [label_to_positions[label_i]] new_target.append(new_target_i) greenlight_span_for_masked_lm_objective.append(label_to_positions[label_i]) return new_target, pheso_caption, greenlight_span_for_masked_lm_objective, label_to_positions
ind_to_class: {0: "__background__", 1 : "person" ...} target: restricted_negative_list : for datasets with restricted negatives, sample only the negatives Convert object detection data into grounding data format, on the fly. Control options: 1. add_detection_prompt: add "object detection : " to the front of the prompt 2. num_negatives: randomly sampled negative classes 3. num_positives: how many positives to keep (-1 means do not cut any) Probabilities to generate the control options: a. probability_one_negative: only give one negative class to mimic evaluation b. probability_one_positive: only give one positive class to mimic evaluation c. probability_full: add both all positive and all negatives d. other: randomly sample some negatives and some positives The below control options are independent of each other: - probability_random_negative: probability of randomly sample X negatives - probability_random_positive: probability of randomly sample some positives
9,931
import os import os.path import math from PIL import Image, ImageDraw import random import numpy as np import torch import torchvision import torch.utils.data as data from maskrcnn_benchmark.structures.bounding_box import BoxList from maskrcnn_benchmark.structures.segmentation_mask import SegmentationMask from maskrcnn_benchmark.structures.keypoint import PersonKeypoints from maskrcnn_benchmark.config import cfg import pdb def _count_visible_keypoints(anno): def _has_only_empty_bbox(anno): def has_valid_annotation(anno): # if it's empty, there is no annotation if len(anno) == 0: return False # if all boxes have close to zero area, there is no annotation if _has_only_empty_bbox(anno): return False # keypoints task have a slight different critera for considering # if an annotation is valid if "keypoints" not in anno[0]: return True # for keypoint detection tasks, only consider valid images those # containing at least min_keypoints_per_image if _count_visible_keypoints(anno) >= cfg.DATALOADER.MIN_KPS_PER_IMS: return True return False
null
9,932
import os import os.path import math from PIL import Image, ImageDraw import random import numpy as np import torch import torchvision import torch.utils.data as data from maskrcnn_benchmark.structures.bounding_box import BoxList from maskrcnn_benchmark.structures.segmentation_mask import SegmentationMask from maskrcnn_benchmark.structures.keypoint import PersonKeypoints from maskrcnn_benchmark.config import cfg import pdb def pil_loader(path, retry=5): # open path as file to avoid ResourceWarning (https://github.com/python-pillow/Pillow/issues/835) ri = 0 while ri < retry: try: with open(path, 'rb') as f: img = Image.open(f) return img.convert('RGB') except: ri += 1
null
9,933
import os import os.path import math from PIL import Image, ImageDraw import random import numpy as np import torch import torchvision import torch.utils.data as data from maskrcnn_benchmark.structures.bounding_box import BoxList from maskrcnn_benchmark.structures.segmentation_mask import SegmentationMask from maskrcnn_benchmark.structures.keypoint import PersonKeypoints from maskrcnn_benchmark.config import cfg import pdb def rgb2id(color): if isinstance(color, np.ndarray) and len(color.shape) == 3: if color.dtype == np.uint8: color = color.astype(np.int32) return color[:, :, 0] + 256 * color[:, :, 1] + 256 * 256 * color[:, :, 2] return int(color[0] + 256 * color[1] + 256 * 256 * color[2])
null
9,934
import os import os.path as op import json import base64 import yaml import errno import io import math from PIL import Image, ImageDraw from maskrcnn_benchmark.structures.bounding_box import BoxList from .box_label_loader import LabelLoader def load_linelist_file(linelist_file): if linelist_file is not None: line_list = [] with open(linelist_file, 'r') as fp: for i in fp: line_list.append(int(i.strip())) return line_list
null
9,935
import os import os.path as op import json import base64 import yaml import errno import io import math from PIL import Image, ImageDraw from maskrcnn_benchmark.structures.bounding_box import BoxList from .box_label_loader import LabelLoader def img_from_base64(imagestring): try: img = Image.open(io.BytesIO(base64.b64decode(imagestring))) return img.convert('RGB') except ValueError: return None
null
9,936
import os import os.path as op import json import base64 import yaml import errno import io import math from PIL import Image, ImageDraw from maskrcnn_benchmark.structures.bounding_box import BoxList from .box_label_loader import LabelLoader def find_file_path_in_yaml(fname, root): if fname is not None: if op.isfile(fname): return fname elif op.isfile(op.join(root, fname)): return op.join(root, fname) else: raise FileNotFoundError( errno.ENOENT, os.strerror(errno.ENOENT), op.join(root, fname) )
null
9,937
import os import os.path as op import json import base64 import yaml import errno import io import math from PIL import Image, ImageDraw from maskrcnn_benchmark.structures.bounding_box import BoxList from .box_label_loader import LabelLoader def create_lineidx(filein, idxout): idxout_tmp = idxout + '.tmp' with open(filein, 'r') as tsvin, open(idxout_tmp, 'w') as tsvout: fsize = os.fstat(tsvin.fileno()).st_size fpos = 0 while fpos != fsize: tsvout.write(str(fpos) + "\n") tsvin.readline() fpos = tsvin.tell() os.rename(idxout_tmp, idxout)
null
9,938
import os import os.path as op import json import base64 import yaml import errno import io import math from PIL import Image, ImageDraw from maskrcnn_benchmark.structures.bounding_box import BoxList from .box_label_loader import LabelLoader def read_to_character(fp, c): result = [] while True: s = fp.read(32) assert s != '' if c in s: result.append(s[: s.index(c)]) break else: result.append(s) return ''.join(result)
null
9,939
import os import os.path as op import json import base64 import yaml import errno import io import math from PIL import Image, ImageDraw from maskrcnn_benchmark.structures.bounding_box import BoxList from .box_label_loader import LabelLoader def load_list_file(fname): with open(fname, 'r') as fp: lines = fp.readlines() result = [line.strip() for line in lines] if len(result) > 0 and result[-1] == '': result = result[:-1] return result
null
9,940
import logging import os import os.path import math from PIL import Image, ImageDraw import random import numpy as np import torch import torchvision import torch.utils.data as data from pycocotools import mask as coco_mask from maskrcnn_benchmark.structures.bounding_box import BoxList from maskrcnn_benchmark.structures.segmentation_mask import SegmentationMask from maskrcnn_benchmark.data.datasets.coco import has_valid_annotation from .od_to_grounding import convert_od_to_grounding_simple, check_for_positive_overflow, sanity_check_target_after_processing, convert_object_detection_to_grounding_optimized_for_od import pdb import json def create_greenlight_map(tok_list, tokenized): # An example tok_list: # [(0, 5), (10, 13), (-1, -1, -1)] # The last one is a special indicator.. greenlight_map = torch.zeros(256, dtype=torch.float) for item in tok_list: if len(item) != 2: assert(len(item) == 3) # Make everything unmakable greenlight_map[:] = -1 break beg, end = item beg_pos = tokenized.char_to_token(beg) end_pos = tokenized.char_to_token(end - 1) if beg_pos is None: try: beg_pos = tokenized.char_to_token(beg + 1) if beg_pos is None: beg_pos = tokenized.char_to_token(beg + 2) except: beg_pos = None if end_pos is None: try: end_pos = tokenized.char_to_token(end - 2) if end_pos is None: end_pos = tokenized.char_to_token(end - 3) except: end_pos = None if beg_pos is None or end_pos is None: continue assert beg_pos is not None and end_pos is not None greenlight_map[beg_pos: end_pos + 1].fill_(1) return greenlight_map
null
9,941
import logging import os import os.path import math from PIL import Image, ImageDraw import random import numpy as np import torch import torchvision import torch.utils.data as data from pycocotools import mask as coco_mask from maskrcnn_benchmark.structures.bounding_box import BoxList from maskrcnn_benchmark.structures.segmentation_mask import SegmentationMask from maskrcnn_benchmark.data.datasets.coco import has_valid_annotation from .od_to_grounding import convert_od_to_grounding_simple, check_for_positive_overflow, sanity_check_target_after_processing, convert_object_detection_to_grounding_optimized_for_od import pdb import json The provided code snippet includes necessary dependencies for implementing the `create_positive_map_for_od_labels` function. Write a Python function `def create_positive_map_for_od_labels(tokenized, label_to_positions)` to solve the following problem: construct a map such that positive_map[i] = j, where j is the object detection label of the token i Here is the function: def create_positive_map_for_od_labels(tokenized, label_to_positions): """construct a map such that positive_map[i] = j, where j is the object detection label of the token i""" """ {3: [1: 5)} 256 : -1 3 3 3 3 -1 .. 8 8 .. the woman in the garden -1 -1 -1 -1 -1 """ positive_map = torch.ones(256, dtype=torch.float) * -1 # -1 means no match keys = list(label_to_positions.keys()) for j, key in enumerate(keys): tok_list = label_to_positions[key] # one label only mapps to one location beg, end = tok_list beg_pos = tokenized.char_to_token(beg) end_pos = tokenized.char_to_token(end - 1) if beg_pos is None: try: beg_pos = tokenized.char_to_token(beg + 1) if beg_pos is None: beg_pos = tokenized.char_to_token(beg + 2) except: beg_pos = None if end_pos is None: try: end_pos = tokenized.char_to_token(end - 2) if end_pos is None: end_pos = tokenized.char_to_token(end - 3) except: end_pos = None if beg_pos is None or end_pos is None: continue assert beg_pos is not None and end_pos is not None positive_map[beg_pos: end_pos + 1].fill_(key) return positive_map
construct a map such that positive_map[i] = j, where j is the object detection label of the token i
9,942
import logging import os import os.path import math from PIL import Image, ImageDraw import random import numpy as np import torch import torchvision import torch.utils.data as data from pycocotools import mask as coco_mask from maskrcnn_benchmark.structures.bounding_box import BoxList from maskrcnn_benchmark.structures.segmentation_mask import SegmentationMask from maskrcnn_benchmark.data.datasets.coco import has_valid_annotation from .od_to_grounding import convert_od_to_grounding_simple, check_for_positive_overflow, sanity_check_target_after_processing, convert_object_detection_to_grounding_optimized_for_od import pdb import json def convert_coco_poly_to_mask(segmentations, height, width): masks = [] for polygons in segmentations: rles = coco_mask.frPyObjects(polygons, height, width) mask = coco_mask.decode(rles) if len(mask.shape) < 3: mask = mask[..., None] mask = torch.as_tensor(mask, dtype=torch.uint8) mask = mask.any(dim=2) masks.append(mask) if masks: masks = torch.stack(masks, dim=0) else: masks = torch.zeros((0, height, width), dtype=torch.uint8) return masks
null
9,943
import logging import os import os.path import math from PIL import Image, ImageDraw import random import numpy as np import torch import torchvision import torch.utils.data as data from pycocotools import mask as coco_mask from maskrcnn_benchmark.structures.bounding_box import BoxList from maskrcnn_benchmark.structures.segmentation_mask import SegmentationMask from maskrcnn_benchmark.data.datasets.coco import has_valid_annotation from .od_to_grounding import convert_od_to_grounding_simple, check_for_positive_overflow, sanity_check_target_after_processing, convert_object_detection_to_grounding_optimized_for_od import pdb import json The provided code snippet includes necessary dependencies for implementing the `create_positive_map` function. Write a Python function `def create_positive_map(tokenized, tokens_positive)` to solve the following problem: construct a map such that positive_map[i,j] = True iff box i is associated to token j Here is the function: def create_positive_map(tokenized, tokens_positive): """construct a map such that positive_map[i,j] = True iff box i is associated to token j""" positive_map = torch.zeros((len(tokens_positive), 256), dtype=torch.float) for j, tok_list in enumerate(tokens_positive): for (beg, end) in tok_list: beg_pos = tokenized.char_to_token(beg) end_pos = tokenized.char_to_token(end - 1) if beg_pos is None: try: beg_pos = tokenized.char_to_token(beg + 1) if beg_pos is None: beg_pos = tokenized.char_to_token(beg + 2) except: beg_pos = None if end_pos is None: try: end_pos = tokenized.char_to_token(end - 2) if end_pos is None: end_pos = tokenized.char_to_token(end - 3) except: end_pos = None if beg_pos is None or end_pos is None: continue assert beg_pos is not None and end_pos is not None positive_map[j, beg_pos: end_pos + 1].fill_(1) return positive_map / (positive_map.sum(-1)[:, None] + 1e-6)
construct a map such that positive_map[i,j] = True iff box i is associated to token j
9,944
import logging import os import os.path import math from PIL import Image, ImageDraw import random import numpy as np import torch import torchvision import torch.utils.data as data from pycocotools import mask as coco_mask from maskrcnn_benchmark.structures.bounding_box import BoxList from maskrcnn_benchmark.structures.segmentation_mask import SegmentationMask from maskrcnn_benchmark.data.datasets.coco import has_valid_annotation from .od_to_grounding import convert_od_to_grounding_simple, check_for_positive_overflow, sanity_check_target_after_processing, convert_object_detection_to_grounding_optimized_for_od import pdb import json def pil_loader(path, retry=5): # open path as file to avoid ResourceWarning (https://github.com/python-pillow/Pillow/issues/835) ri = 0 while ri < retry: try: with open(path, 'rb') as f: img = Image.open(f) return img.convert('RGB') except: ri += 1
null
9,945
import os import os.path import json from PIL import Image import torch.utils.data as data def pil_loader(path): # open path as file to avoid ResourceWarning (https://github.com/python-pillow/Pillow/issues/835) with open(path, 'rb') as f: img = Image.open(f) return img.convert('RGB')
null
9,946
import json import os import time from collections import defaultdict import pycocotools.mask as mask_utils import torchvision from PIL import Image from .modulated_coco import ConvertCocoPolysToMask def _isArrayLike(obj): return hasattr(obj, "__iter__") and hasattr(obj, "__len__")
null
9,947
from __future__ import division import os from collections import defaultdict import numpy as np from maskrcnn_benchmark.structures.bounding_box import BoxList from maskrcnn_benchmark.structures.boxlist_ops import boxlist_iou def eval_detection_voc(pred_boxlists, gt_boxlists, iou_thresh=0.5, use_07_metric=False): """Evaluate on voc dataset. Args: pred_boxlists(list[BoxList]): pred boxlist, has labels and scores fields. gt_boxlists(list[BoxList]): ground truth boxlist, has labels field. iou_thresh: iou thresh use_07_metric: boolean Returns: dict represents the results """ assert len(gt_boxlists) == len( pred_boxlists ), "Length of gt and pred lists need to be same." prec, rec = calc_detection_voc_prec_rec( pred_boxlists=pred_boxlists, gt_boxlists=gt_boxlists, iou_thresh=iou_thresh ) ap = calc_detection_voc_ap(prec, rec, use_07_metric=use_07_metric) return {"ap": ap, "map": np.nanmean(ap)} def do_voc_evaluation(dataset, predictions, output_folder, logger): # TODO need to make the use_07_metric format available # for the user to choose pred_boxlists = [] gt_boxlists = [] for image_id, prediction in enumerate(predictions): img_info = dataset.get_img_info(image_id) if len(prediction) == 0: continue image_width = img_info["width"] image_height = img_info["height"] prediction = prediction.resize((image_width, image_height)) pred_boxlists.append(prediction) gt_boxlist = dataset.get_groundtruth(image_id) gt_boxlists.append(gt_boxlist) result = eval_detection_voc( pred_boxlists=pred_boxlists, gt_boxlists=gt_boxlists, iou_thresh=0.5, use_07_metric=True, ) result_str = "mAP: {:.4f}\n".format(result["map"]) for i, ap in enumerate(result["ap"]): if i == 0: # skip background continue result_str += "{:<16}: {:.4f}\n".format( dataset.map_class_id_to_class_name(i), ap ) logger.info(result_str) if output_folder: with open(os.path.join(output_folder, "result.txt"), "w") as fid: fid.write(result_str) return result
null
9,948
from maskrcnn_benchmark.structures.boxlist_ops import boxlist_iou from maskrcnn_benchmark.structures.bounding_box import BoxList import json import numpy as np import os.path as osp import os from prettytable import PrettyTable import xml.etree.ElementTree as ET from collections import defaultdict from pathlib import Path from typing import Any, Dict, List, Optional, Sequence, Tuple, Union import maskrcnn_benchmark.utils.mdetr_dist as dist The provided code snippet includes necessary dependencies for implementing the `get_sentence_data` function. Write a Python function `def get_sentence_data(filename) -> List[Dict[str, Any]]` to solve the following problem: Parses a sentence file from the Flickr30K Entities dataset input: filename - full file path to the sentence file to parse output: a list of dictionaries for each sentence with the following fields: sentence - the original sentence phrases - a list of dictionaries for each phrase with the following fields: phrase - the text of the annotated phrase first_word_index - the position of the first word of the phrase in the sentence phrase_id - an identifier for this phrase phrase_type - a list of the coarse categories this phrase belongs to Here is the function: def get_sentence_data(filename) -> List[Dict[str, Any]]: """ Parses a sentence file from the Flickr30K Entities dataset input: filename - full file path to the sentence file to parse output: a list of dictionaries for each sentence with the following fields: sentence - the original sentence phrases - a list of dictionaries for each phrase with the following fields: phrase - the text of the annotated phrase first_word_index - the position of the first word of the phrase in the sentence phrase_id - an identifier for this phrase phrase_type - a list of the coarse categories this phrase belongs to """ with open(filename, "r") as f: sentences = f.read().split("\n") annotations = [] for sentence in sentences: if not sentence: continue first_word = [] phrases = [] phrase_id = [] phrase_type = [] words = [] current_phrase = [] add_to_phrase = False for token in sentence.split(): if add_to_phrase: if token[-1] == "]": add_to_phrase = False token = token[:-1] current_phrase.append(token) phrases.append(" ".join(current_phrase)) current_phrase = [] else: current_phrase.append(token) words.append(token) else: if token[0] == "[": add_to_phrase = True first_word.append(len(words)) parts = token.split("/") phrase_id.append(parts[1][3:]) phrase_type.append(parts[2:]) else: words.append(token) sentence_data = {"sentence": " ".join(words), "phrases": []} for index, phrase, p_id, p_type in zip(first_word, phrases, phrase_id, phrase_type): sentence_data["phrases"].append( {"first_word_index": index, "phrase": phrase, "phrase_id": p_id, "phrase_type": p_type} ) annotations.append(sentence_data) return annotations
Parses a sentence file from the Flickr30K Entities dataset input: filename - full file path to the sentence file to parse output: a list of dictionaries for each sentence with the following fields: sentence - the original sentence phrases - a list of dictionaries for each phrase with the following fields: phrase - the text of the annotated phrase first_word_index - the position of the first word of the phrase in the sentence phrase_id - an identifier for this phrase phrase_type - a list of the coarse categories this phrase belongs to