Merge remote-tracking branch 'origin/v2' into v2

app/chain/search.py

@@ -374,8 +374,13 @@ class SearchChain(ChainBase):
             logger.warn(f'{title} 未搜索到资源')
             return []
         # 组装上下文
-        contexts = [Context(meta_info=MetaInfo(title=torrent.title, subtitle=torrent.description),
+        contexts = [
-                            torrent_info=torrent) for torrent in torrents]
+            Context(
+                meta_info=MetaInfo(title=torrent.title, subtitle=torrent.description),
+                torrent_info=torrent,
+                resource_source="search",
+            ) for torrent in torrents
+        ]
         # 保存到本地文件
         if cache_local:
             self.save_cache(contexts, self.__result_temp_file)
@@ -446,8 +451,13 @@ class SearchChain(ChainBase):
             logger.warn(f'{title} 未搜索到资源')
             return []
         # 组装上下文
-        contexts = [Context(meta_info=MetaInfo(title=torrent.title, subtitle=torrent.description),
+        contexts = [
-                            torrent_info=torrent) for torrent in torrents]
+            Context(
+                meta_info=MetaInfo(title=torrent.title, subtitle=torrent.description),
+                torrent_info=torrent,
+                resource_source="search",
+            ) for torrent in torrents
+        ]
         # 保存到本地文件
         if cache_local:
             await self.async_save_cache(contexts, self.__result_temp_file)
@@ -470,8 +480,11 @@ class SearchChain(ChainBase):
         async for event in self.__async_search_all_sites_stream(keyword=title, sites=sites, page=page):
             result = event.pop("items", []) or []
             batch_contexts = [
-                Context(meta_info=MetaInfo(title=torrent.title, subtitle=torrent.description),
+                Context(
-                        torrent_info=torrent)
+                    meta_info=MetaInfo(title=torrent.title, subtitle=torrent.description),
+                    torrent_info=torrent,
+                    resource_source="search",
+                )
                 for torrent in result
             ]
             if batch_contexts:
@@ -718,7 +731,7 @@ class SearchChain(ChainBase):
                         and mediainfo.imdb_id \
                         and torrent.imdbid == mediainfo.imdb_id:
                     logger.info(f'{mediainfo.title} 通过IMDBID匹配到资源：{torrent.site_name} - {torrent.title}')
-                    _match_torrents.append((torrent, torrent_meta))
+                    _match_torrents.append((torrent, torrent_meta, "imdbid"))
                     continue
 
                 # 比对种子
@@ -726,7 +739,7 @@ class SearchChain(ChainBase):
                                                torrent_meta=torrent_meta,
                                                torrent=torrent):
                     # 匹配成功
-                    _match_torrents.append((torrent, torrent_meta))
+                    _match_torrents.append((torrent, torrent_meta, "title"))
                     continue
             # 匹配完成
             logger.info(f"匹配完成，共匹配到 {len(_match_torrents)} 个资源")
@@ -736,9 +749,17 @@ class SearchChain(ChainBase):
             # 去掉mediainfo中多余的数据
             mediainfo.clear()
             # 组装上下文
-            contexts = [Context(torrent_info=t[0],
+            contexts = [
-                                media_info=mediainfo,
+                Context(
-                                meta_info=t[1]) for t in _match_torrents]
+                    torrent_info=t[0],
+                    media_info=mediainfo,
+                    meta_info=t[1],
+                    resource_source="search",
+                    match_source=t[2],
+                    candidate_recognized=False,
+                    media_info_is_target=True,
+                ) for t in _match_torrents
+            ]
         finally:
             torrents.clear()
             del torrents
@@ -989,9 +1010,13 @@ class SearchChain(ChainBase):
                 result = event.pop("items", []) or []
                 torrents.extend(result)
                 batch_contexts = [
-                    Context(meta_info=MetaInfo(title=torrent.title, subtitle=torrent.description),
+                    Context(
-                            media_info=mediainfo,
+                        meta_info=MetaInfo(title=torrent.title, subtitle=torrent.description),
-                            torrent_info=torrent)
+                        media_info=mediainfo,
+                        torrent_info=torrent,
+                        resource_source="search",
+                        media_info_is_target=True,
+                    )
                     for torrent in result
                 ]
                 candidate_contexts.extend(batch_contexts)

app/chain/subscribe.py

@@ -833,6 +833,11 @@ class SubscribeChain(ChainBase):
                             re_mediainfo.clear()
                             # 更新种子缓存
                             context.media_info = re_mediainfo
+                            context.match_source = self.__get_media_id_match_source(re_mediainfo)
+                            context.candidate_recognized = bool(
+                                re_mediainfo.tmdb_id or re_mediainfo.douban_id
+                            )
+                            context.media_info_is_target = False
                             # 重置失败次数
                             context.media_recognize_fail_count = 0
                             logger.debug(f'种子 {context.torrent_info.title} 重新识别成功')
@@ -941,6 +946,11 @@ class SubscribeChain(ChainBase):
                                         torrent_mediainfo.clear()
                                         # 更新种子缓存
                                         _context.media_info = torrent_mediainfo
+                                        _context.match_source = self.__get_media_id_match_source(torrent_mediainfo)
+                                        _context.candidate_recognized = bool(
+                                            torrent_mediainfo.tmdb_id or torrent_mediainfo.douban_id
+                                        )
+                                        _context.media_info_is_target = False
 
                             # 如果仍然没有识别到媒体信息，尝试标题匹配
                             if not torrent_mediainfo or (
@@ -956,6 +966,9 @@ class SubscribeChain(ChainBase):
                                     torrent_mediainfo = mediainfo
                                     # 更新种子缓存
                                     _context.media_info = mediainfo
+                                    _context.match_source = "title"
+                                    _context.candidate_recognized = False
+                                    _context.media_info_is_target = True
                                 else:
                                     continue
 
@@ -971,6 +984,18 @@ class SubscribeChain(ChainBase):
                                     continue
                                 logger.info(
                                     f'{mediainfo.title_year} 通过媒体ID匹配到可选资源：{torrent_info.site_name} - {torrent_info.title}')
+                                match_source = getattr(_context, "match_source", "unknown")
+                                if match_source == "title":
+                                    # 标题兜底使用的是订阅目标 media_info，不能标记为候选自身识别结果。
+                                    _context.candidate_recognized = False
+                                    _context.media_info_is_target = True
+                                elif match_source == "unknown":
+                                    _context.match_source = self.__get_media_id_match_source(torrent_mediainfo)
+                                    _context.candidate_recognized = True
+                                    _context.media_info_is_target = False
+                                else:
+                                    _context.candidate_recognized = True
+                                    _context.media_info_is_target = False
                             else:
                                 continue
 
@@ -2537,6 +2562,17 @@ class SubscribeChain(ChainBase):
 
         return min_ep <= start_ep and max_ep >= end_ep
 
+    @staticmethod
+    def __get_media_id_match_source(mediainfo: Optional[MediaInfo]) -> str:
+        """
+        返回候选自身识别命中的明确媒体 ID 类型。
+        """
+        if mediainfo and mediainfo.tmdb_id:
+            return "tmdbid"
+        if mediainfo and mediainfo.douban_id:
+            return "doubanid"
+        return "unknown"
+
     @staticmethod
     def get_states_for_search(state: str) -> str:
         """

app/chain/torrents.py

@@ -60,8 +60,8 @@ class TorrentsChain(ChainBase):
         else:
             torrents_cache = self.load_cache(self._rss_file) or {}
 
-        # 兼容性处理：为旧版本的Context对象添加失败次数字段
+        # 兼容性处理：为旧版本的Context对象补齐新增候选识别字段
-        self._ensure_context_compatibility(torrents_cache)
+        self._ensure_context_compatibility(torrents_cache, stype=stype)
 
         return torrents_cache
 
@@ -80,8 +80,8 @@ class TorrentsChain(ChainBase):
         else:
             torrents_cache = await self.async_load_cache(self._rss_file) or {}
 
-        # 兼容性处理：为旧版本的Context对象添加失败次数字段
+        # 兼容性处理：为旧版本的Context对象补齐新增候选识别字段
-        self._ensure_context_compatibility(torrents_cache)
+        self._ensure_context_compatibility(torrents_cache, stype=stype)
 
         return torrents_cache
 
@@ -285,8 +285,18 @@ class TorrentsChain(ChainBase):
                             mediainfo = MediaInfo()
                         # 清理多余数据，减少内存占用
                         mediainfo.clear()
+                        candidate_recognized = bool(mediainfo and (mediainfo.tmdb_id or mediainfo.douban_id))
+                        match_source = self._get_media_id_match_source(mediainfo)
                         # 上下文
-                        context = Context(meta_info=meta, media_info=mediainfo, torrent_info=torrent)
+                        context = Context(
+                            meta_info=meta,
+                            media_info=mediainfo,
+                            torrent_info=torrent,
+                            resource_source="spider" if stype == "spider" else "rss",
+                            match_source=match_source if candidate_recognized else "unknown",
+                            candidate_recognized=candidate_recognized,
+                            media_info_is_target=False,
+                        )
                         # 如果未识别到媒体信息，设置初始失败次数为1
                         if not mediainfo or (not mediainfo.tmdb_id and not mediainfo.douban_id):
                             context.media_recognize_fail_count = 1
@@ -317,19 +327,44 @@ class TorrentsChain(ChainBase):
         return torrents_cache
 
     @staticmethod
-    def _ensure_context_compatibility(torrents_cache: Dict[str, List[Context]]):
+    def _ensure_context_compatibility(torrents_cache: Dict[str, List[Context]], stype: Optional[str] = None):
         """
         确保Context对象的兼容性，为旧版本添加缺失的字段
         """
         for domain, contexts in torrents_cache.items():
             for context in contexts:
-                # 如果Context对象没有media_recognize_fail_count字段，添加默认值
+                context_fields = vars(context)
-                if not hasattr(context, 'media_recognize_fail_count'):
+                # 旧 pickle 实例会读到 dataclass 类默认值，必须检查实例字段，避免跳过兼容回填。
+                if "media_recognize_fail_count" not in context_fields:
                     context.media_recognize_fail_count = 0
                     # 如果媒体信息未识别，设置初始失败次数
                     if (not context.media_info or
                             (not context.media_info.tmdb_id and not context.media_info.douban_id)):
                         context.media_recognize_fail_count = 1
+                if "resource_source" not in context_fields:
+                    context.resource_source = "spider" if stype == "spider" else "rss"
+                if "candidate_recognized" not in context_fields:
+                    context.candidate_recognized = bool(
+                        context.media_info and (context.media_info.tmdb_id or context.media_info.douban_id)
+                    )
+                if "match_source" not in context_fields:
+                    context.match_source = (
+                        TorrentsChain._get_media_id_match_source(context.media_info)
+                        if context.candidate_recognized else "unknown"
+                    )
+                if "media_info_is_target" not in context_fields:
+                    context.media_info_is_target = False
+
+    @staticmethod
+    def _get_media_id_match_source(mediainfo: Optional[MediaInfo]) -> str:
+        """
+        返回候选自身识别命中的明确媒体 ID 类型。
+        """
+        if mediainfo and mediainfo.tmdb_id:
+            return "tmdbid"
+        if mediainfo and mediainfo.douban_id:
+            return "doubanid"
+        return "unknown"
 
     def __renew_rss_url(self, domain: str, site: dict):
         """

app/core/context.py

@@ -832,6 +832,14 @@ class Context:
     torrent_info: TorrentInfo = None
     # 媒体识别失败次数
     media_recognize_fail_count: int = 0
+    # 候选资源来源：rss、spider、search、unknown。
+    resource_source: str = "unknown"
+    # 候选匹配来源：tmdbid、doubanid、imdbid、title、plugin、unknown。
+    match_source: str = "unknown"
+    # 候选自身是否已经识别出有效媒体 ID。
+    candidate_recognized: bool = False
+    # 当前 media_info 是否为目标媒体回填，而不是候选自身识别结果。
+    media_info_is_target: bool = False
 
     def to_dict(self):
         """
@@ -841,5 +849,9 @@ class Context:
             "meta_info": self.meta_info.to_dict() if self.meta_info else None,
             "torrent_info": self.torrent_info.to_dict() if self.torrent_info else None,
             "media_info": self.media_info.to_dict() if self.media_info else None,
-            "media_recognize_fail_count": self.media_recognize_fail_count
+            "media_recognize_fail_count": self.media_recognize_fail_count,
+            "resource_source": self.resource_source,
+            "match_source": self.match_source,
+            "candidate_recognized": self.candidate_recognized,
+            "media_info_is_target": self.media_info_is_target,
         }

app/schemas/context.py

@@ -248,6 +248,14 @@ class Context(BaseModel):
     media_info: Optional[Union[MediaInfo, Any]] = None
     # 种子信息
     torrent_info: Optional[TorrentInfo] = None
+    # 候选资源来源：rss、spider、search、unknown
+    resource_source: Optional[str] = "unknown"
+    # 候选匹配来源：tmdbid、doubanid、imdbid、title、plugin、unknown
+    match_source: Optional[str] = "unknown"
+    # 候选自身是否已经识别出有效媒体 ID
+    candidate_recognized: Optional[bool] = False
+    # 当前 media_info 是否为目标媒体回填
+    media_info_is_target: Optional[bool] = False
 
 
 class MediaSeason(BaseModel):