fix(classification): preserve non-TMDB legacy categories

app/application/classification/execution.py

@@ -5,7 +5,7 @@ from __future__ import annotations
 from collections.abc import Callable, Mapping
 from copy import deepcopy
 from enum import Enum
-from typing import Protocol, TypeAlias
+from typing import Final, Protocol, TypeAlias, cast
 
 from app.application.classification.legacy import (
     build_legacy_tmdb_extension_facts,
@@ -16,6 +16,7 @@ from app.domain.classification.facts import build_classification_facts
 from app.domain.context import MediaInfo, MusicAlbumInfo, MusicArtistInfo, MusicInfo
 from app.schemas.category import (
     CategoryConfig,
+    ClassificationEvaluation,
     ClassificationFacts,
     ClassificationFactValue,
     ClassificationPolicy,
@@ -34,6 +35,9 @@ ClassificationExtensionFactsProvider: TypeAlias = Callable[
 ]
 """按当前插件注册表校验并提供来源扩展分类事实的端口。"""
 
+_LEGACY_TMDB_SOURCE: Final[str] = "themoviedb"
+_LEGACY_RULE_PREFIX: Final[str] = "legacy."
+
 
 class ClassificationRuntimePort(Protocol):
     """分类执行只需要的活动策略与 legacy 快照端口。"""
@@ -278,6 +282,9 @@ class ClassificationExecutionService:
     ) -> ClassificationSubject:
         """应用纯求值结果和人工覆盖，并更新兼容目录分类。"""
         evaluation = ClassificationEvaluator.evaluate(policy, facts)
+        legacy_evaluation = _evaluate_legacy_tmdb_compatibility(policy, facts)
+        if legacy_evaluation is not None and _uses_fallback(evaluation.result):
+            evaluation = legacy_evaluation
         result = evaluation.result.model_copy(deep=True)
         if effective_override:
             result.effective = effective_override.model_copy(deep=True)
@@ -346,6 +353,94 @@ class ClassificationExecutionService:
             media.set_library_category(_category_path_snapshot(effective))
 
 
+def _evaluate_legacy_tmdb_compatibility(
+    policy: ClassificationPolicy,
+    facts: ClassificationFacts,
+) -> ClassificationEvaluation | None:
+    """让旧 TMDB 分类规则消费非 TMDB 来源已经拥有的标准事实。"""
+    if facts.identity.media_source == _LEGACY_TMDB_SOURCE:
+        return None
+    legacy_rules = [
+        rule for rule in policy.rules if rule.id.startswith(_LEGACY_RULE_PREFIX)
+    ]
+    if not legacy_rules:
+        return None
+    compatibility_policy = policy.model_copy(
+        deep=True,
+        update={"rules": legacy_rules},
+    )
+    compatibility_facts = _legacy_tmdb_compatibility_facts(policy, facts)
+    return ClassificationEvaluator.evaluate(
+        compatibility_policy,
+        compatibility_facts,
+    )
+
+
+def _legacy_tmdb_compatibility_facts(
+    policy: ClassificationPolicy,
+    facts: ClassificationFacts,
+) -> ClassificationFacts:
+    """把跨来源标准字段投影到旧规则的 TMDB 扩展命名空间。"""
+    extensions = {
+        str(source): {str(key): value for key, value in values.items()}
+        for source, values in facts.extensions.items()
+    }
+    legacy_info = _legacy_tmdb_info_from_standard_facts(facts)
+    for source, values in build_legacy_tmdb_extension_facts(
+        policy,
+        legacy_info,
+    ).items():
+        target = extensions.setdefault(source, {})
+        for field, value in values.items():
+            target.setdefault(field, value)
+    identity = facts.identity.model_copy(
+        update={"media_source": _LEGACY_TMDB_SOURCE}
+    )
+    return cast(
+        ClassificationFacts,
+        facts.model_copy(
+            deep=True,
+            update={"identity": identity, "extensions": extensions},
+        ),
+    )
+
+
+def _legacy_tmdb_info_from_standard_facts(
+    facts: ClassificationFacts,
+) -> dict[str, object]:
+    """构造旧规则投影所需的有限 TMDB 字段，不伪造缺失事实。"""
+    media = facts.media
+    info: dict[str, object] = {}
+    if media.language:
+        info["original_language"] = media.language
+    if media.countries:
+        countries = [str(country) for country in media.countries if country]
+        if countries:
+            info["origin_country"] = countries
+            info["production_countries"] = [
+                {"iso_3166_1": country} for country in countries
+            ]
+    if media.year is not None:
+        info["release_date"] = str(media.year)
+    for field in (
+        "adult",
+        "runtime",
+        "content_rating",
+        "companies",
+        "networks",
+    ):
+        value = getattr(media, field, None)
+        if value not in (None, "", []):
+            info[field] = value
+    return info
+
+
+def _uses_fallback(result: ClassificationResult) -> bool:
+    """判断主来源求值是否没有命中具体分类规则。"""
+    selection = result.effective or result.recommended
+    return selection is None or selection.source in {None, "fallback", "source_fallback"}
+
+
 def _classification_extensions(
     policy: ClassificationPolicy,
     media: ClassificationSubject,

app/db/adapters/classification.py

@@ -34,7 +34,7 @@ DirectoryConfigurationNormalizer = Callable[
 
 
 def discard_removed_source_fallbacks(value: Any) -> Any:
-    """读取持久化策略时丢弃已删除的来源级默认分类字段，不再恢复其行为。"""
+    """读取旧持久化策略时把 TMDB 兜底平移为媒体类型全局兜底。"""
     if not isinstance(value, Mapping):
         return value
 
@@ -47,9 +47,27 @@ def discard_removed_source_fallbacks(value: Any) -> Any:
     if isinstance(history, list):
         policies.extend(item for item in history if isinstance(item, dict))
     for policy in policies:
-        policy.pop("source_fallbacks", None)
+        source_fallbacks = policy.pop("source_fallbacks", None)
+        if not isinstance(source_fallbacks, Mapping):
+            continue
+        fallbacks = policy.get("fallbacks")
+        tmdb_fallbacks = source_fallbacks.get("themoviedb")
+        if not isinstance(fallbacks, dict) or not isinstance(tmdb_fallbacks, Mapping):
+            continue
+        for media_type, category_id in tmdb_fallbacks.items():
+            if fallbacks.get(media_type) != _COMMON_FALLBACK_IDS.get(media_type):
+                continue
+            if category_id:
+                fallbacks[media_type] = category_id
     return state
 
+
+_COMMON_FALLBACK_IDS = {
+    "电影": "movie.uncategorized",
+    "电视剧": "tv.uncategorized",
+    "音乐": "music.uncategorized",
+}
+
 _CONFIGURATION_LOCK_KEYS = (
     SystemConfigKey.MediaClassificationPolicy.value,
     SystemConfigKey.Directories.value,

tests/test_media_classification_configuration_adapter.py

@@ -25,6 +25,7 @@ from app.application.directory import normalize_directory_configurations_for_pol
 from app.db.adapters.classification import (
     SystemConfigClassificationPolicyStore,
     SystemConfigDirectoryConfigurationStore,
+    discard_removed_source_fallbacks,
 )
 from app.db.models.systemconfig import SystemConfig
 from app.db.oper.systemconfig import SystemConfigOper
@@ -75,6 +76,37 @@ def _store(db: Any) -> tuple[SystemConfigClassificationPolicyStore, SystemConfig
     )
 
 
+def test_old_source_fallback_is_promoted_to_global_fallback() -> None:
+    """旧策略读取时应保留迁移生成的分类兜底，并移除废弃字段。"""
+    state = {
+        "active": {
+            "fallbacks": {
+                "电影": "movie.uncategorized",
+                "电视剧": "tv.uncategorized",
+                "音乐": "music.uncategorized",
+            },
+            "source_fallbacks": {
+                "themoviedb": {"电视剧": "legacy.tv.uncategorized"},
+            },
+        },
+        "history": [
+            {
+                "fallbacks": {"电视剧": "tv.uncategorized"},
+                "source_fallbacks": {
+                    "themoviedb": {"电视剧": "legacy.tv.previous"},
+                },
+            }
+        ],
+    }
+
+    normalized = discard_removed_source_fallbacks(state)
+
+    assert normalized["active"]["fallbacks"]["电视剧"] == "legacy.tv.uncategorized"
+    assert normalized["history"][0]["fallbacks"]["电视剧"] == "legacy.tv.previous"
+    assert "source_fallbacks" not in normalized["active"]
+    assert "source_fallbacks" not in normalized["history"][0]
+
+
 def test_adapter_initializes_and_round_trips_json_datetime(db: Any) -> None:
     """首次创建原子写入 JSON，并在提交后发布系统配置快照。"""
     store, oper = _store(db)

tests/test_media_classification_execution.py

@@ -181,6 +181,39 @@ def test_execution_classifies_copy_and_preserves_source_identity() -> None:
     assert source.classification.policy_revision == 1
 
 
+def test_legacy_tmdb_rules_use_non_tmdb_standard_facts() -> None:
+    """旧 TMDB 分类规则应使用豆瓣等来源已有的标准国家事实。"""
+    migration = migrate_legacy_category_config(
+        {
+            "movie": {},
+            "tv": {
+                "国产剧": {"origin_country": "CN,TW,HK"},
+                "未分类": None,
+            },
+        }
+    )
+    source = MediaInfo(
+        media_source=MediaSource.Douban,
+        media_id="35593344",
+        type=MediaType.TV,
+        title="测试剧",
+        production_countries=[{"name": "中国大陆"}],
+    )
+
+    finalized = ClassificationExecutionService(
+        _Runtime(migration.policy)
+    ).finalize(source)
+
+    assert finalized.media_source == MediaSource.Douban
+    assert finalized.library_category == "国产剧"
+    assert finalized.classification is not None
+    assert finalized.classification.effective.category_id == next(
+        category.id
+        for category in migration.policy.categories
+        if category.name == "国产剧"
+    )
+
+
 def test_execution_builds_complete_facts_without_mutating_media() -> None:
     """影响分析事实入口应复用插件字段构造，并保持原媒体对象不变。"""
     source = MediaInfo(