feat: expand metadata sources and media server sync (#6129)

2026-08-21 16:23:34 +08:00 · 2026-07-21 11:34:18 +08:00
parent 1d708870c9
commit e3f80af74f
43 changed files with 2581 additions and 152 deletions
--- a/app/core/config.py
+++ b/app/core/config.py
@@ -38,6 +38,8 @@ class SystemConfModel(BaseModel):
    douban: int = 0
    # Bangumi请求缓存数量
    bangumi: int = 0
+    # AniList请求缓存数量
+    anilist: int = 0
    # Fanart请求缓存数量
    fanart: int = 0
    # 元数据缓存过期时间（秒）
@@ -197,11 +199,11 @@ class ConfigModel(BaseModel):
    DOH_RESOLVERS: str = "1.0.0.1,1.1.1.1,9.9.9.9,149.112.112.112"

    # ==================== 媒体元数据配置 ====================
-    # 媒体搜索来源 themoviedb/douban/bangumi，多个用,分隔
+    # 媒体搜索来源 themoviedb/douban/bangumi/anilist，多个用,分隔
    SEARCH_SOURCE: str = "themoviedb"
-    # 媒体识别来源 themoviedb/douban
+    # 媒体识别来源 themoviedb/douban/bangumi/anilist
    RECOGNIZE_SOURCE: str = "themoviedb"
-    # 刮削来源 themoviedb/douban
+    # 刮削来源 themoviedb/douban/bangumi/anilist
    SCRAP_SOURCE: str = "themoviedb"
    # 电视剧动漫的分类genre_ids
    ANIME_GENREIDS: List[int] = Field(default=[16])
--- a/app/core/context.py
+++ b/app/core/context.py
@@ -10,6 +10,7 @@ from app.schemas.types import MediaType
 from app.utils.string import StringUtils

 BANGUMI_MOVIE_PLATFORMS = frozenset({"movie", "电影", "剧场版"})
+ANILIST_MOVIE_FORMATS = frozenset({"MOVIE"})


@dataclass
@@ -251,8 +252,10 @@ class MediaInfo:

    # 内部标记：是否命中本地识别缓存，不参与序列化
    recognize_cache_hit = False
-    # 来源：themoviedb、douban、bangumi
+    # 来源：themoviedb、douban、bangumi、anilist
    source: str = None
+    # 请求级刮削来源；为空时使用系统设置
+    scrape_source: str = None
    # 类型 电影、电视剧
    type: MediaType = None
    # 媒体标题
@@ -279,6 +282,10 @@ class MediaInfo:
    douban_id: str = None
    # Bangumi ID
    bangumi_id: int = None
+    # AniList ID
+    anilist_id: int = None
+    # AniDB ID（AniList外部映射）
+    anidb_id: int = None
    # 合集ID
    collection_id: int = None
    # 媒体原语种
@@ -315,6 +322,8 @@ class MediaInfo:
    douban_info: dict = field(default_factory=dict)
    # Bangumi INFO
    bangumi_info: dict = field(default_factory=dict)
+    # AniList INFO
+    anilist_info: dict = field(default_factory=dict)
    # 导演
    directors: List[dict] = field(default_factory=list)
    # 演员
@@ -380,6 +389,8 @@ class MediaInfo:
            self.set_douban_info(self.douban_info)
        if self.bangumi_info:
            self.set_bangumi_info(self.bangumi_info)
+        if self.anilist_info:
+            self.set_anilist_info(self.anilist_info)

    def __setattr__(self, name: str, value: Any):
        self.__dict__[name] = value
@@ -750,7 +761,7 @@ class MediaInfo:
        self.source = "bangumi"
        # 本体
        self.bangumi_info = info
-        # 豆瓣ID
+        # Bangumi ID
        self.bangumi_id = info.get("id")
        # 类型
        if not self.type:
@@ -804,13 +815,166 @@ class MediaInfo:
        if self.type == MediaType.TV and not self.seasons:
            meta = MetaInfo(self.title)
            season = meta.begin_season if meta.begin_season is not None else 1
-            episodes_count = info.get("total_episodes")
+            episodes_count = info.get("total_episodes") or info.get("eps")
            if episodes_count:
                self.seasons[season] = list(range(1, episodes_count + 1))
+                self.number_of_episodes = episodes_count
+                self.number_of_seasons = 1
+        # 风格
+        if not self.genres:
+            self.genres = [
+                {"id": tag.get("name"), "name": tag.get("name")}
+                for tag in info.get("tags") or []
+                if tag.get("name")
+            ]
+        # 制作公司与导演
+        if info.get("infobox"):
+            companies = []
+            directors = []
+            for item in info.get("infobox"):
+                values = item.get("value")
+                if not isinstance(values, list):
+                    values = [values]
+                normalized_values = [
+                    value.get("v") if isinstance(value, dict) else value
+                    for value in values
+                    if value
+                ]
+                if item.get("key") in {"动画制作", "制作"}:
+                    companies.extend({"name": value} for value in normalized_values)
+                elif item.get("key") == "导演":
+                    directors.extend({"name": value} for value in normalized_values)
+            if companies and not self.production_companies:
+                self.production_companies = companies
+            if directors and not self.directors:
+                self.directors = directors
        # 演员
        if not self.actors:
            self.actors = info.get("actors") or []

+    @staticmethod
+    def get_anilist_media_type(info: dict) -> MediaType:
+        """
+        根据 AniList 发布格式获取标准媒体类型。
+
+        :param info: AniList 媒体信息
+        :return: 标准媒体类型
+        """
+        return (
+            MediaType.MOVIE
+            if str(info.get("format") or "").upper() in ANILIST_MOVIE_FORMATS
+            else MediaType.TV
+        )
+
+    @staticmethod
+    def _anilist_date(date_info: dict) -> Optional[str]:
+        """
+        将 AniList 模糊日期转换为标准日期文本。
+
+        :param date_info: AniList FuzzyDate 字段
+        :return: YYYY、YYYY-MM 或 YYYY-MM-DD 日期文本
+        """
+        if not date_info or not date_info.get("year"):
+            return None
+        values = [str(date_info.get("year"))]
+        if date_info.get("month"):
+            values.append(str(date_info.get("month")).zfill(2))
+        if date_info.get("day"):
+            values.append(str(date_info.get("day")).zfill(2))
+        return "-".join(values)
+
+    def set_anilist_info(self, info: dict) -> None:
+        """
+        初始化 AniList 媒体信息。
+
+        :param info: AniList 媒体详情
+        """
+        if not info:
+            return
+        self.source = "anilist"
+        self.anilist_info = info
+        self.anilist_id = info.get("id")
+        self.type = self.type or self.get_anilist_media_type(info)
+
+        titles = info.get("title") or {}
+        self.title = self.title or titles.get("english") or titles.get("romaji") or titles.get("native")
+        self.en_title = self.en_title or titles.get("english")
+        self.original_title = self.original_title or titles.get("native") or titles.get("romaji")
+        self.names = list(
+            dict.fromkeys(
+                value
+                for value in [
+                    titles.get("english"),
+                    titles.get("romaji"),
+                    titles.get("native"),
+                    *(info.get("synonyms") or []),
+                ]
+                if value and value != self.title
+            )
+        )
+
+        self.release_date = self.release_date or self._anilist_date(info.get("startDate") or {})
+        self.first_air_date = self.first_air_date or self.release_date
+        self.last_air_date = self.last_air_date or self._anilist_date(info.get("endDate") or {})
+        self.year = self.year or (
+            str(info.get("startDate", {}).get("year"))
+            if info.get("startDate", {}).get("year")
+            else str(info.get("seasonYear")) if info.get("seasonYear") else None
+        )
+
+        cover = info.get("coverImage") or {}
+        self.poster_path = self.poster_path or cover.get("extraLarge") or cover.get("large")
+        self.backdrop_path = self.backdrop_path or info.get("bannerImage")
+        self.overview = self.overview or re.sub(
+            r"<[^>]+>",
+            "",
+            str(info.get("description") or "").replace("<br>", "\n").replace("<br />", "\n"),
+        ).strip()
+        self.vote_average = self.vote_average or (
+            round(float(info.get("averageScore")) / 10, 1)
+            if info.get("averageScore") is not None
+            else 0
+        )
+        self.popularity = self.popularity or info.get("popularity")
+        self.runtime = self.runtime or info.get("duration")
+        self.adult = self.adult or bool(info.get("isAdult"))
+        self.status = self.status or info.get("status")
+        self.original_language = self.original_language or (
+            "ja" if info.get("countryOfOrigin") == "JP" else None
+        )
+        self.origin_country = self.origin_country or (
+            [info.get("countryOfOrigin")] if info.get("countryOfOrigin") else []
+        )
+        self.production_companies = self.production_companies or [
+            {"name": studio.get("name")}
+            for studio in info.get("studios", {}).get("nodes") or []
+            if studio.get("name")
+        ]
+        self.genres = self.genres or [
+            {"id": genre, "name": genre} for genre in info.get("genres") or []
+        ]
+        self.actors = self.actors or info.get("actors") or []
+        self.directors = self.directors or info.get("directors") or []
+
+        if self.season is None:
+            self.season = MetaInfo(self.title).begin_season if self.title else None
+        episodes_count = info.get("episodes")
+        if self.type == MediaType.TV and episodes_count:
+            season = self.season if self.season is not None else 1
+            self.seasons[season] = list(range(1, episodes_count + 1))
+            self.number_of_episodes = episodes_count
+            self.number_of_seasons = 1
+            if self.year:
+                self.season_years[season] = self.year
+
+        for external_link in info.get("externalLinks") or []:
+            if str(external_link.get("site") or "").casefold() != "anidb":
+                continue
+            match = re.search(r"\d+", external_link.get("url") or "")
+            if match:
+                self.anidb_id = int(match.group())
+                break
+
    @property
    def title_year(self):
        if self.title:
@@ -831,6 +995,8 @@ class MediaInfo:
            return "https://movie.douban.com/subject/%s" % self.douban_id
        elif self.bangumi_id:
            return "http://bgm.tv/subject/%s" % self.bangumi_id
+        elif self.anilist_id:
+            return "https://anilist.co/anime/%s" % self.anilist_id
        return ""

    @property
@@ -895,6 +1061,16 @@ class MediaInfo:
        dicts["tmdb_info"] = None
        dicts["douban_info"] = None
        dicts["bangumi_info"] = None
+        dicts["anilist_info"] = None
+        dicts["mediaid_prefix"] = self.source
+        source_ids = {
+            "themoviedb": self.tmdb_id,
+            "douban": self.douban_id,
+            "bangumi": self.bangumi_id,
+            "anilist": self.anilist_id,
+        }
+        media_id = source_ids.get(self.source)
+        dicts["media_id"] = str(media_id) if media_id is not None else None
        return dicts

    def clear(self):
@@ -904,6 +1080,7 @@ class MediaInfo:
        self.tmdb_info = {}
        self.douban_info = {}
        self.bangumi_info = {}
+        self.anilist_info = {}
        self.seasons = {}
        self.genres = []
        self.season_info = []
--- a/app/core/meta/infopath.py
+++ b/app/core/meta/infopath.py
@@ -23,7 +23,7 @@ def should_use_parent_title_for_file_stem(
    """
    if not file_meta.isfile or not stem or not parent_dir_name:
        return False
-    if file_meta.tmdbid or file_meta.doubanid:
+    if file_meta.tmdbid or file_meta.doubanid or file_meta.media_id:
        return False
    if not PARENT_LATIN_TITLE_RE.search(parent_dir_name):
        return False
--- a/app/core/meta/metabase.py
+++ b/app/core/meta/metabase.py
@@ -97,6 +97,8 @@ class MetaBase(object):
    # 附加信息
    tmdbid: int = None
    doubanid: str = None
+    media_source: Optional[str] = None
+    media_id: Optional[str] = None
    episode_group: Optional[str] = None
    # 帧率信息（纯数值）
    fps: Optional[int] = None
@@ -683,6 +685,11 @@ class MetaBase(object):
        # doubanid
        if not self.doubanid and meta.doubanid:
            self.doubanid = meta.doubanid
+        # 通用媒体来源与ID
+        if not self.media_source and meta.media_source:
+            self.media_source = meta.media_source
+        if not self.media_id and meta.media_id:
+            self.media_id = meta.media_id
        # 剧集组
        if not self.episode_group and meta.episode_group:
            self.episode_group = meta.episode_group
--- a/app/core/metainfo.py
+++ b/app/core/metainfo.py
@@ -29,6 +29,8 @@ _ANIME_SQUARE_BRACKET_RE = re.compile(r'\[[+0-9XVPI-]+]\s*\[', re.IGNORECASE)
 _BRACED_METAINFO_RE = re.compile(r'(?<={\[)[\W\w]+(?=]})')
 _BRACED_TMDBID_RE = re.compile(r'(?<=tmdbid=)\d+')
 _BRACED_DOUBANID_RE = re.compile(r'(?<=doubanid=)\d+')
+_BRACED_BANGUMIID_RE = re.compile(r'(?<=bangumiid=)\d+')
+_BRACED_ANILISTID_RE = re.compile(r'(?<=anilistid=)\d+')
 _BRACED_TYPE_RE = re.compile(r'(?<=type=)\w+')
 _BRACED_EPISODE_GROUP_RE = re.compile(r'(?:^|;)g=([0-9a-fA-F]+)(?=;|$)')
 _BRACED_BEGIN_SEASON_RE = re.compile(r'(?<=s=)\d+')
@@ -41,6 +43,24 @@ _EMBY_TMDB_RE_LIST = (
    re.compile(r'\{tmdbid[=\-](\d+)\}'),
    re.compile(r'\{tmdb[=\-](\d+)\}'),
 )
+_EXTENDED_MEDIA_ID_RE_LIST = {
+    "bangumi": (
+        re.compile(r'\[bangumiid[=\-](\d+)\]'),
+        re.compile(r'\[bangumi[=\-](\d+)\]'),
+        re.compile(r'\{bangumiid[=\-](\d+)\}'),
+        re.compile(r'\{bangumi[=\-](\d+)\}'),
+    ),
+    "anilist": (
+        re.compile(r'\[anilistid[=\-](\d+)\]'),
+        re.compile(r'\[anilist[=\-](\d+)\]'),
+        re.compile(r'\{anilistid[=\-](\d+)\}'),
+        re.compile(r'\{anilist[=\-](\d+)\}'),
+    ),
+}
+_EXTENDED_MEDIA_ID_TAG_RE = re.compile(
+    r'(?:bangumi(?:id)?|anilist(?:id)?)[=\-]\d+',
+    re.IGNORECASE,
+)
 _RUST_PARSE_OPTIONS_CACHE_KEY = "_cache_key"


@@ -51,6 +71,10 @@ def _empty_metainfo() -> dict:
    return {
        'tmdbid': None,
        'doubanid': None,
+        'bangumiid': None,
+        'anilistid': None,
+        'media_source': None,
+        'media_id': None,
        'type': None,
        'episode_group': None,
        'begin_season': None,
@@ -115,6 +139,14 @@ def _find_metainfo_python(title: str) -> Tuple[str, dict]:
            doubanid = _BRACED_DOUBANID_RE.search(result)
            if doubanid and doubanid.group(0).isdigit():
                metainfo['doubanid'] = doubanid.group(0)
+            # 查找Bangumi ID信息
+            bangumiid = _BRACED_BANGUMIID_RE.search(result)
+            if bangumiid and bangumiid.group(0).isdigit():
+                metainfo['bangumiid'] = bangumiid.group(0)
+            # 查找AniList ID信息
+            anilistid = _BRACED_ANILISTID_RE.search(result)
+            if anilistid and anilistid.group(0).isdigit():
+                metainfo['anilistid'] = anilistid.group(0)
            # 查找媒体类型
            mtype = _BRACED_TYPE_RE.search(result)
            if mtype:
@@ -142,7 +174,18 @@ def _find_metainfo_python(title: str) -> Tuple[str, dict]:
            if end_episode and end_episode.group(0).isdigit():
                metainfo['end_episode'] = int(end_episode.group(0))
            # 去除title中该部分
-            if tmdbid or mtype or episode_group or begin_season or end_season or begin_episode or end_episode:
+            if (
+                tmdbid
+                or doubanid
+                or bangumiid
+                or anilistid
+                or mtype
+                or episode_group
+                or begin_season
+                or end_season
+                or begin_episode
+                or end_episode
+            ):
                title = title.replace(f"{{[{result}]}}", '')

    # 支持Emby格式的ID标签；第一个 [tmdbid] 历史上始终优先处理，用于覆盖前面 {[...]} 中的旧标签。
@@ -159,6 +202,31 @@ def _find_metainfo_python(title: str) -> Tuple[str, dict]:
                title = tmdb_re.sub('', title).strip()
                break

+    for source, patterns in _EXTENDED_MEDIA_ID_RE_LIST.items():
+        key = f"{source}id"
+        if metainfo.get(key):
+            continue
+        for media_id_re in patterns:
+            media_id_match = media_id_re.search(title)
+            if not media_id_match:
+                continue
+            metainfo[key] = media_id_match.group(1)
+            title = media_id_re.sub('', title).strip()
+            break
+
+    if metainfo.get('tmdbid'):
+        metainfo['media_source'] = 'themoviedb'
+        metainfo['media_id'] = metainfo['tmdbid']
+    elif metainfo.get('doubanid'):
+        metainfo['media_source'] = 'douban'
+        metainfo['media_id'] = metainfo['doubanid']
+    elif metainfo.get('bangumiid'):
+        metainfo['media_source'] = 'bangumi'
+        metainfo['media_id'] = metainfo['bangumiid']
+    elif metainfo.get('anilistid'):
+        metainfo['media_source'] = 'anilist'
+        metainfo['media_id'] = metainfo['anilistid']
+
    # 计算季集总数
    _apply_range_total(metainfo, 'begin_season', 'end_season', 'total_season')
    _apply_range_total(metainfo, 'begin_episode', 'end_episode', 'total_episode')
@@ -202,6 +270,10 @@ def _build_meta_info(
            logger.warn("tmdbid 必须是数字")
    if metainfo.get('doubanid'):
        meta.doubanid = metainfo['doubanid']
+    if metainfo.get('media_source'):
+        meta.media_source = metainfo['media_source']
+    if metainfo.get('media_id'):
+        meta.media_id = str(metainfo['media_id'])
    if metainfo.get('type'):
        meta.type = MediaType(metainfo['type']) if isinstance(metainfo['type'], str) else metainfo['type']
    if metainfo.get('episode_group'):
@@ -319,6 +391,8 @@ def _meta_from_rust(parsed: dict) -> Optional[MetaBase]:
        "apply_words": parsed.get("apply_words") or [],
        "tmdbid": parsed.get("tmdbid"),
        "doubanid": parsed.get("doubanid"),
+        "media_source": parsed.get("media_source"),
+        "media_id": parsed.get("media_id"),
        "episode_group": parsed.get("episode_group"),
        "fps": parsed.get("fps"),
    }
@@ -327,6 +401,24 @@ def _meta_from_rust(parsed: dict) -> Optional[MetaBase]:
    return meta


+def _requires_python_metainfo(
+    title: str,
+    custom_words: Optional[List[str]] = None,
+) -> bool:
+    """
+    判断标题或临时识别词是否包含当前Rust扩展尚未支持的数据源ID标签。
+
+    :param title: 原始标题
+    :param custom_words: 临时识别词
+    :return: 是否必须使用Python解析器
+    """
+    candidates = [title or "", *(custom_words or [])]
+    contains_extended_id = any(
+        _EXTENDED_MEDIA_ID_TAG_RE.search(candidate) for candidate in candidates
+    )
+    return contains_extended_id and not rust_accel.supports_extended_media_ids()
+
+
 def MetaInfo(title: str, subtitle: Optional[str] = None, custom_words: List[str] = None) -> MetaBase:
    """
    根据标题和副标题识别元数据
@@ -335,9 +427,11 @@ def MetaInfo(title: str, subtitle: Optional[str] = None, custom_words: List[str]
    :param custom_words: 自定义识别词列表
    :return: MetaAnime、MetaVideo
    """
-    rust_meta = _meta_from_rust(
-        rust_accel.parse_metainfo(title, subtitle, _rust_parse_options(custom_words))
-    )
+    rust_meta = None
+    if not _requires_python_metainfo(title, custom_words):
+        rust_meta = _meta_from_rust(
+            rust_accel.parse_metainfo(title, subtitle, _rust_parse_options(custom_words))
+        )
    if rust_meta:
        return rust_meta
    meta = _build_meta_info(title=title, subtitle=subtitle, custom_words=custom_words)
@@ -355,9 +449,14 @@ def MetaInfoPath(path: Path, custom_words: List[str] = None) -> MetaBase:
    :param path: 路径
    :param custom_words: 自定义识别词列表
    """
-    rust_meta = _meta_from_rust(
-        rust_accel.parse_metainfo_path(str(path), _rust_parse_options(custom_words))
+    path_context = " ".join(
+        [path.name, path.parent.name, path.parent.parent.name]
    )
+    rust_meta = None
+    if not _requires_python_metainfo(path_context, custom_words):
+        rust_meta = _meta_from_rust(
+            rust_accel.parse_metainfo_path(str(path), _rust_parse_options(custom_words))
+        )
    if rust_meta:
        return rust_meta
    # 文件元数据，不包含后缀
@@ -400,7 +499,9 @@ def find_metainfo(title: str) -> Tuple[str, dict]:
    """
    从标题中提取媒体信息
    """
-    rust_result = rust_accel.find_metainfo(title)
+    rust_result = None
+    if not _requires_python_metainfo(title):
+        rust_result = rust_accel.find_metainfo(title)
    if rust_result:
        return rust_result["title"], rust_result["metainfo"]
    return _find_metainfo_python(title)