import os import re import sys import tempfile from collections import deque from copy import deepcopy from functools import cached_property, reduce from pathlib import Path from shutil import copy from tomlkit.toml_document import TOMLDocument from typing import Any from typing import ClassVar from typing import Dict from typing import Optional from typing import Tuple from typing import Type from typing import Union import tomlkit from pydantic import ValidationError from pydantic_settings import BaseSettings from pydantic_settings import PydanticBaseSettingsSource from pydantic_settings import PyprojectTomlConfigSettingsSource from pydantic_settings import SettingsConfigDict from anaconda_cli_base.exceptions import ( AnacondaConfigTomlSyntaxError, AnacondaConfigValidationError, ) if sys.version_info >= (3, 11): import tomllib else: import tomli as tomllib def anaconda_secrets_dir() -> Optional[Path]: path = Path( os.path.expandvars( os.path.expanduser(os.getenv("ANACONDA_SECRETS_DIR", "/run/secrets")) ) ) return path if path.is_dir() else None def anaconda_config_path() -> Path: return Path( os.path.expandvars( os.path.expanduser( os.getenv("ANACONDA_CONFIG_TOML", "~/.anaconda/config.toml") ) ) ) class AnacondaConfigTomlSettingsSource(PyprojectTomlConfigSettingsSource): _cache: ClassVar[Dict[Path, Dict[str, Any]]] = {} def _read_file(self, file_path: Path) -> Dict[str, Any]: try: result = self._cache.get(file_path) if result is None: result = super()._read_file(file_path) self._cache[file_path] = result return result except tomllib.TOMLDecodeError as e: arg = f"{anaconda_config_path()}: {e.args[0]}" raise AnacondaConfigTomlSyntaxError(arg) class AnacondaBaseSettings(BaseSettings): def __init_subclass__( cls, plugin_name: Optional[Union[str, tuple]] = None, **kwargs: Any ) -> None: base_env_prefix: str = "ANACONDA_" pyproject_toml_table_header: Tuple[str, ...] if plugin_name is None: pyproject_toml_table_header = () env_prefix = base_env_prefix elif isinstance(plugin_name, tuple): if not all(isinstance(entry, str) for entry in plugin_name): raise ValueError( f"plugin_name={plugin_name} error: All values must be strings." ) pyproject_toml_table_header = ("plugin", *plugin_name) env_prefix = base_env_prefix + "_".join(plugin_name).upper() + "_" elif isinstance(plugin_name, str): pyproject_toml_table_header = ("plugin", plugin_name) env_prefix = base_env_prefix + f"{plugin_name.upper()}_" else: raise ValueError( f"plugin_name={plugin_name} is not supported. It must be either a str or tuple." ) cls.model_config = SettingsConfigDict( env_file=".env", pyproject_toml_table_header=pyproject_toml_table_header, env_prefix=env_prefix, env_nested_delimiter="__", extra="ignore", ignored_types=(cached_property,), secrets_dir=anaconda_secrets_dir(), validate_assignment=True, ) return super().__init_subclass__(**kwargs) def __init__(self, **kwargs: Any) -> None: try: super().__init__(**kwargs) except ValidationError as e: errors = [] for error in e.errors(): input_value = error["input"] msg = error["msg"] env_prefix = self.model_config.get("env_prefix", "") delimiter = self.model_config.get("env_nested_delimiter", "") or "" env_var = env_prefix + delimiter.join( str(loc).upper() for loc in error["loc"] ) kwarg = error["loc"][0] if kwarg in kwargs: value = kwargs[str(kwarg)] msg = f"- Error in init kwarg {e.title}({error['loc'][0]}={value})\n {msg}" elif env_var in os.environ: msg = f"- Error in environment variable {env_var}={input_value}\n {msg}" else: table_header = ".".join( self.model_config.get("pyproject_toml_table_header", []) ) key = ".".join(str(loc) for loc in error["loc"]) msg = f"- Error in {anaconda_config_path()} in [{table_header}] for {key} = {input_value}\n {msg}" errors.append(msg) message = "\n" + "\n".join(errors) raise AnacondaConfigValidationError(message) @classmethod def settings_customise_sources( cls, settings_cls: Type[BaseSettings], init_settings: PydanticBaseSettingsSource, env_settings: PydanticBaseSettingsSource, dotenv_settings: PydanticBaseSettingsSource, file_secret_settings: PydanticBaseSettingsSource, ) -> Tuple[PydanticBaseSettingsSource, ...]: return ( init_settings, env_settings, file_secret_settings, dotenv_settings, AnacondaConfigTomlSettingsSource(settings_cls, anaconda_config_path()), ) def write_config( self, preserve_existing_keys: bool = True, dry_run: bool = False, ) -> None: """ Write the current configuration to the Anaconda config.toml file. This method writes the configuration instance to the config.toml file, preserving existing comments and formatting. Only non-default and non-None values are written. If a value is set to its default, the corresponding entry is removed from the config file. The write operation is atomic - the config file is only updated if the entire write succeeds, preventing corruption from interrupted writes. Args: preserve_existing_keys: If True (default) updates to existing keys in the config.toml file, will not remove the key if set to the default value. If False fields set to default value are removed from the file dry_run: If True, displays a diff of proposed changes without writing to the file. If False (default), writes changes to config.toml. Raises: ValidationError: If any attribute has been manually set to an invalid value that fails pydantic validation. OSError: If backup creation fails, config file cannot be read, or config directory cannot be created due to permissions or I/O errors. ValueError: If the existing config.toml contains invalid TOML syntax. Behavior: - Creates ~/.anaconda/config.toml if it doesn't exist - Creates timestamped backup (e.g., config.backup.20231218_143022.toml) - Keeps last 5 backups, automatically deletes older ones - Preserves comments and formatting in existing config - Only writes non-default, non-None values - Removes keys when values are set to their defaults - Validates all values before writing - Uses atomic write to prevent file corruption """ values = self.model_dump( exclude_unset=False, exclude_defaults=True, exclude_none=True, exclude_computed_fields=True, ) # save a timestamped backup of the config.toml config_toml = anaconda_config_path() if config_toml.exists(): from datetime import datetime timestamp = datetime.now().strftime("%Y%m%d_%H%M%S_%f") backup_path = config_toml.with_name(f"config.backup.{timestamp}.toml") try: copy(config_toml, backup_path) except (OSError, IOError) as e: raise OSError( f"Failed to create backup of {config_toml} at {backup_path}: {e}" ) from e # Clean up old backups, keeping only the last 5 try: backups = sorted( config_toml.parent.glob("config.backup.*.toml"), key=lambda p: p.stat().st_mtime, reverse=True, ) # Keep the 5 most recent backups, delete the rest for old_backup in backups[5:]: old_backup.unlink() except (OSError, IOError): # If cleanup fails, continue anyway - backup was already created pass try: with open(config_toml, "rt") as f: config = tomlkit.load(f) except (OSError, IOError) as e: raise OSError(f"Failed to read {config_toml}: {e}") from e except Exception as e: raise ValueError( f"Failed to parse {config_toml} as TOML. " f"The file may be corrupted or contain invalid TOML syntax: {e}" ) from e else: try: config_toml.parent.mkdir(parents=True, exist_ok=True) except (OSError, IOError) as e: raise OSError( f"Failed to create directory {config_toml.parent}: {e}" ) from e config = tomlkit.TOMLDocument() to_update = deepcopy(config) table_header = self.model_config.get("pyproject_toml_table_header", tuple()) if table_header: def nestitem(a: tomlkit.TOMLDocument, b: Any) -> Any: if b not in a: a.add(b, tomlkit.table()) return a[b] parent = reduce(nestitem, table_header, to_update) else: parent = to_update def deepmerge( orig: tomlkit.TOMLDocument, new: Dict[str, Any], full_model: Dict[str, Any], preserve_existing_keys: bool = True, ) -> None: stack = deque[Tuple[TOMLDocument, Dict[str, Any], Dict[str, Any]]]( [(orig, new, full_model)] ) while stack: current_original, current_update, current_full = stack.popleft() removed_keys = ( current_original.keys() - current_update.keys() - {"plugin"} ) for k in removed_keys: # If a key was already present in toml # ensure that it remains set even if the # new value is the default for the class value = current_full.get(k, None) if (value is not None) and preserve_existing_keys: current_original[k] = value else: del current_original[k] for k, v in current_update.items(): if isinstance(v, dict): if k not in current_original: current_original.add(k, tomlkit.table()) to_append = (current_original.get(k), v, current_full.get(k)) stack.append(to_append) # type: ignore else: current_original[k] = v full_dump = self.model_dump() deepmerge( parent, values, full_dump, preserve_existing_keys=preserve_existing_keys ) if dry_run: import difflib import datetime as dt from rich.syntax import Syntax from anaconda_cli_base.console import console original = config.as_string() updated = to_update.as_string() dt_format = "%m-%d-%y %H:%M" config_toml = anaconda_config_path() if config_toml.exists(): modified = dt.datetime.fromtimestamp( config_toml.stat().st_mtime ).strftime(dt_format) else: modified = "" now = dt.datetime.now().strftime(dt_format) diffs = difflib.unified_diff( original.splitlines(False), updated.splitlines(False), fromfile=str(config_toml), fromfiledate=modified, tofile=str(config_toml), tofiledate=now, lineterm="", ) diff = "\n".join(diffs) if not diff: console.print( f"[bold green]No change to {anaconda_config_path()}[/bold green]" ) return syntax = Syntax(code=diff, lexer="diff", line_numbers=False, word_wrap=True) console.print(syntax) return # Use atomic write to prevent corruption if write fails # Write to temp file in same directory, then atomically rename tmp_fd, tmp_path = tempfile.mkstemp( dir=config_toml.parent, prefix=".config_", suffix=".toml.tmp", text=True, ) try: config_dump = tomlkit.dumps(to_update) config_dump = re.sub(r"\n+$", "\n", config_dump, flags=re.DOTALL) with os.fdopen(tmp_fd, "wt") as f: f.write(config_dump) # Atomic rename - if this fails, original file is untouched os.replace(tmp_path, config_toml) # ensure that any existing cache of the config.toml file # is cleared. AnacondaConfigTomlSettingsSource._cache.clear() except Exception: # Clean up temp file if write or rename failed try: os.unlink(tmp_path) except OSError: pass raise