Merge pull request 'fix(metadata): לא להתיישב 'completed' בכשל-חילוץ-Gemini חולף (#138)' (#261) from worktree-metadata-no-settle-on-fail into main

mcp-server/tests/test_metadata_extract_failure_status.py

@@ -0,0 +1,63 @@
+"""Regression test for #138 — metadata extraction must distinguish a transient
+failure (Gemini hiccup despite the row having text) from a permanent empty
+(no text to extract). Conflating them as 'no_metadata' let the drain settle the
+row to 'completed' and silently strand it with empty metadata.
+
+``extract_and_apply`` returns:
+  * ``extraction_failed`` when ``extract_metadata`` yields nothing BUT the row
+    has full_text → retryable.
+  * ``no_metadata`` when the row has no text → genuinely nothing to do.
+
+Runs fully OFFLINE — monkeypatches ``extract_metadata`` and ``db.get_case_law``.
+"""
+
+from __future__ import annotations
+
+import asyncio
+from uuid import uuid4
+
+import pytest
+
+from legal_mcp.services import db, precedent_metadata_extractor as pme
+
+
+def _run(coro):
+    loop = asyncio.new_event_loop()
+    try:
+        return loop.run_until_complete(coro)
+    finally:
+        loop.close()
+
+
+@pytest.fixture()
+def empty_extract(monkeypatch: pytest.MonkeyPatch):
+    async def _empty(_cid):
+        return {}
+    monkeypatch.setattr(pme, "extract_metadata", _empty)
+
+
+def test_empty_result_with_text_is_transient_failure(empty_extract, monkeypatch):
+    async def _rec(_cid):
+        return {"full_text": "פסק דין ארוך עם תוכן ממשי לחילוץ"}
+    monkeypatch.setattr(db, "get_case_law", _rec)
+
+    out = _run(pme.extract_and_apply(uuid4()))
+    assert out["status"] == "extraction_failed", out
+
+
+def test_empty_result_without_text_is_no_metadata(empty_extract, monkeypatch):
+    async def _rec(_cid):
+        return {"full_text": ""}
+    monkeypatch.setattr(db, "get_case_law", _rec)
+
+    out = _run(pme.extract_and_apply(uuid4()))
+    assert out["status"] == "no_metadata", out
+
+
+def test_missing_record_is_no_metadata(empty_extract, monkeypatch):
+    async def _none(_cid):
+        return None
+    monkeypatch.setattr(db, "get_case_law", _none)
+
+    out = _run(pme.extract_and_apply(uuid4()))
+    assert out["status"] == "no_metadata", out