diff --git a/app/application/formatting.py b/app/application/formatting.py index a738abca8..0f86557cf 100644 --- a/app/application/formatting.py +++ b/app/application/formatting.py @@ -3,7 +3,7 @@ from collections import defaultdict from dataclasses import dataclass from functools import lru_cache from pathlib import Path -from typing import Dict, Iterable, List, Match, Optional, Tuple, Union, cast +from typing import Dict, Iterable, List, Match, Optional, Pattern, Tuple, Union, cast import anitopy @@ -12,8 +12,11 @@ from app.domain.meta.metabase import MetaBase from app.domain.meta.words import calculate_episode_offset from app.domain.metainfo import MetaInfoPath from app.runtime.log import logger +from app.schemas.file import FileItem from app.schemas.transfer import EpisodeFormatRule -from app.schemas.workflow import FileItem + +# 推荐结果保留既有字典输出，字段值只包含文本、计数、开关、原因列表或空值。 +EpisodeFormatRecommendation = Dict[str, Union[str, int, bool, List[str], None]] @dataclass(frozen=True) @@ -28,7 +31,8 @@ class _TemplateParseResult: def _compile_template_pattern( template: str, ep_group_name: Optional[str] = None, -): +) -> Pattern[str]: + """编译定位模板，并缓存转义花括号与命名分组的解析结果。""" parts: List[str] = ["^"] cursor = 0 while cursor < len(template): @@ -68,6 +72,7 @@ def _match_template( text: str, ep_group_name: Optional[str] = None, ) -> Optional[_TemplateParseResult]: + """匹配模板并保留命名分组在原文件名中的位置。""" pattern = _compile_template_pattern(template, ep_group_name) result = pattern.match(text) if not result: @@ -85,11 +90,11 @@ def _match_template( class FormatParser(object): """解析剧集定位模板并从文件名提取集数区间。""" - _key = "" + _key: Optional[str] = "" _split_chars = r"\.|\s+|\(|\)|\[|]|-|\+|【|】|/|～|;|&|\||#|_|「|」|~" def __init__(self, eformat: str, details: Optional[str] = None, part: Optional[str] = None, - offset: Optional[str] = None, key: Optional[str] = "ep"): + offset: Optional[str] = None, key: Optional[str] = "ep") -> None: """ :params eformat: 格式化字符串 :params details: 格式化详情 @@ -98,8 +103,8 @@ class FormatParser(object): :prams key: EP关键字 """ self._format = eformat - self._start_ep = None - self._end_ep = None + self._start_ep: Optional[Union[int, str]] = None + self._end_ep: Optional[Union[int, str]] = None if not offset: self.__offset = "EP" elif "EP" in offset: @@ -110,7 +115,7 @@ class FormatParser(object): else: self.__offset = f"EP+{offset}" self._key = key - self._part = None + self._part: Optional[str] = None self._compiled_pattern = ( _compile_template_pattern(self._format, self._key) if self._format @@ -131,27 +136,27 @@ class FormatParser(object): self._start_ep = self._end_ep = int(tmp[0]) @property - def format(self): + def format(self) -> str: """返回原始定位模板。""" return self._format @property - def start_ep(self): + def start_ep(self) -> Optional[Union[int, str]]: """返回规则限定的起始集数。""" return self._start_ep @property - def end_ep(self): + def end_ep(self) -> Optional[Union[int, str]]: """返回规则限定的结束集数。""" return self._end_ep @property - def part(self): + def part(self) -> Optional[str]: """返回规则指定的分段信息。""" return self._part @property - def offset(self): + def offset(self) -> str: """返回集数偏移表达式。""" return self.__offset @@ -164,9 +169,11 @@ class FormatParser(object): return False if self._start_ep is None: return True - if self._start_ep <= s <= self._end_ep: - return True - return False + if isinstance(self._start_ep, str): + # 连字符形式保存的是原始区间文本，比较时使用数字边界，偏移仍由拆集负责。 + start_ep, end_ep = (int(value) for value in self._start_ep.split("-")) + return start_ep <= s <= end_ep + return isinstance(self._end_ep, int) and self._start_ep <= s <= self._end_ep def split_episode(self, file_name: str, file_meta: MetaBase) -> Tuple[Optional[int], Optional[int], Optional[str]]: """ @@ -209,7 +216,7 @@ class FormatParser(object): if not self._format: return None, None ret = self._compiled_pattern.match(file) if self._compiled_pattern else None - if not ret or self._key not in ret.groupdict(): + if not ret or self._key is None or self._key not in ret.groupdict(): return None, None episodes = ret.group(self._key) if not re.compile( @@ -288,7 +295,7 @@ class EpisodeFormatRuleHelper: self, rules: List[EpisodeFormatRule], sample_files: List[FileItem], - ) -> Tuple[bool, str, Optional[dict]]: + ) -> Tuple[bool, str, Optional[EpisodeFormatRecommendation]]: """ 推荐集数定位模板 """ @@ -304,7 +311,7 @@ class EpisodeFormatRuleHelper: continue sample_file, match_result = matched_samples[0] - episode_format = self._build_template(sample_file.name, match_result) + episode_format = self._build_template(sample_file.name or "", match_result) if not episode_format: continue if not self._validate_template(episode_format, matched_samples): @@ -338,7 +345,7 @@ class EpisodeFormatRuleHelper: def _auto_recommend( self, sample_files: List[FileItem], - ) -> Tuple[bool, str, Optional[dict]]: + ) -> Tuple[bool, str, Optional[EpisodeFormatRecommendation]]: """ 自动生成集数定位模板：anitopy 反向定位 + 多文件对比 """ @@ -655,7 +662,7 @@ class EpisodeFormatRuleHelper: _AutoRecommendSample( file_name=file_name, ep_span=ep_span, - expected_episode=normalized_episode, + expected_episode=normalized_episode or "", source_kind=self._get_file_kind(item), native_episode=native_episode, native_verified=native_verified, @@ -668,7 +675,7 @@ class EpisodeFormatRuleHelper: def _locate_episode( cls, file_name: str, - episode_value: str, + episode_value: Optional[str], ) -> Optional[Tuple[int, int]]: """ 三级策略反向定位 episode_number 在文件名中的位置 @@ -702,7 +709,7 @@ class EpisodeFormatRuleHelper: return None @staticmethod - def _normalize_episode_value(episode_value) -> str: + def _normalize_episode_value(episode_value: object) -> str: """将单集或范围集数归一为不含 EP 前缀的文本。""" if isinstance(episode_value, list): parts = [str(part) for part in episode_value]