fix(search): preserve cached resource contexts

app/schemas/search.py

@@ -5,14 +5,14 @@ from typing import Literal, Union
 from pydantic import BaseModel, Field
 
 from app.schemas.common import JsonData
-from app.schemas.context import SubtitleInfo, TorrentInfo
+from app.schemas.context import Context, SubtitleInfo
 
 
 class SearchLastContextData(BaseModel):
     """上一次搜索的请求参数与结果。"""
 
     params: dict[str, JsonData] = Field(default_factory=dict)
-    results: list[Union[TorrentInfo, SubtitleInfo]] = Field(default_factory=list)
+    results: list[Union[Context, SubtitleInfo]] = Field(default_factory=list)
 
 
 class SearchRecommendStatusData(BaseModel):

tests/test_search_response_models.py

@@ -0,0 +1,48 @@
+"""搜索接口响应模型测试。"""
+
+from app.schemas.context import Context, MetaInfo, SubtitleInfo, TorrentInfo
+from app.schemas.response import Response
+from app.schemas.search import SearchLastContextData
+
+
+def test_last_search_context_preserves_nested_torrent_context() -> None:
+    """最近搜索响应必须保留资源上下文的嵌套元数据和种子信息。"""
+    context = Context(
+        meta_info=MetaInfo(title="Example", name="Example"),
+        torrent_info=TorrentInfo(title="Example torrent", site_name="Site A"),
+    )
+
+    response = Response[SearchLastContextData](
+        success=True,
+        data={
+            "params": {"keyword": "Example", "result_type": "torrent"},
+            "results": [context.model_dump(mode="json")],
+        },
+    ).model_dump(mode="json")
+
+    result = response["data"]["results"][0]
+    assert result["meta_info"]["name"] == "Example"
+    assert result["torrent_info"]["title"] == "Example torrent"
+    assert result["torrent_info"]["site_name"] == "Site A"
+
+
+def test_last_search_context_keeps_subtitle_result_shape() -> None:
+    """修正资源上下文类型后仍必须完整保留字幕搜索结果。"""
+    subtitle = SubtitleInfo(
+        title="Example subtitle",
+        site_name="Subtitle Site",
+        enclosure="https://example.test/subtitle.srt",
+    )
+
+    response = Response[SearchLastContextData](
+        success=True,
+        data={
+            "params": {"keyword": "Example", "result_type": "subtitle"},
+            "results": [subtitle.model_dump(mode="json")],
+        },
+    ).model_dump(mode="json")
+
+    result = response["data"]["results"][0]
+    assert result["title"] == "Example subtitle"
+    assert result["site_name"] == "Subtitle Site"
+    assert result["enclosure"] == "https://example.test/subtitle.srt"