refactor review-plan safety and validation

2026-04-07 11:00:47 +08:00
parent fb128c70d6
commit 0a6bddcc7e
19 changed files with 1246 additions and 628 deletions
@@ -614,6 +614,114 @@ def test_generate_plan_with_analysis_by_reputation_missing_scores_uses_fallback_
    assert "评分依据不足" in plan.human_summary


+
+
+def test_generate_plan_with_analysis_by_quality_missing_quality_data_marks_manual_review(config):
+    """Missing quality entries should not silently keep the first duplicate."""
+    config.duplicate_keep = "by_quality"
+    p_low = Path("/mnt/nas/videos/movie/Test.2020.720p.WEB-DL.mkv")
+    p_high = Path("/mnt/nas/videos/movie/Test.2020.1080p.BluRay.mkv")
+    vf_low = VideoFile(
+        path=p_low,
+        filename="Test.2020.720p.WEB-DL.mkv",
+        size_bytes=1000000,
+        modified_timestamp=datetime.now(timezone.utc),
+        category="movie",
+    )
+    vf_high = VideoFile(
+        path=p_high,
+        filename="Test.2020.1080p.BluRay.mkv",
+        size_bytes=2000000,
+        modified_timestamp=datetime.now(timezone.utc),
+        category="movie",
+    )
+    identity = MovieIdentity(
+        title="Test",
+        year=2020,
+        confidence=0.9,
+        needs_review=False,
+        original_filename="Test.2020.mkv",
+    )
+
+    plan = generate_plan(
+        [(vf_low, identity), (vf_high, identity)],
+        config,
+        analysis_data={
+            "metadata": {"source_identities": "identities.json"},
+            "completeness": [],
+            "duplicates": [
+                {
+                    "identity": {"type": "movie", "title": "Test", "year": 2020},
+                    "files": [str(p_low), str(p_high)],
+                    "quality_comparison": [
+                        {"path": str(p_low), "resolution": "1280x720", "size_bytes": 1000000},
+                    ],
+                }
+            ],
+        },
+    )
+
+    assert all(op.operation_type == "no-op" for op in plan.operations)
+    assert "Duplicate group needs manual review" in plan.operations[0].reason
+    issues = plan.metadata["validation_snapshot"]["duplicate_resolution_issues"]
+    assert len(issues) == 1
+    assert "missing quality comparison entries" in issues[0]["reason"]
+    assert "转人工复核" in plan.human_summary
+
+
+def test_generate_plan_normalizes_duplicate_paths_before_matching(config):
+    """Duplicate matching should survive harmless path-format differences."""
+    config.duplicate_keep = "by_quality"
+    p_720 = Path("/mnt/nas/videos/movie/Test.2020.720p.WEB-DL.mkv")
+    p_1080 = Path("/mnt/nas/videos/movie/Test.2020.1080p.BluRay.mkv")
+    vf_720 = VideoFile(
+        path=p_720,
+        filename="Test.2020.720p.WEB-DL.mkv",
+        size_bytes=1000000,
+        modified_timestamp=datetime.now(timezone.utc),
+        category="movie",
+    )
+    vf_1080 = VideoFile(
+        path=p_1080,
+        filename="Test.2020.1080p.BluRay.mkv",
+        size_bytes=2000000,
+        modified_timestamp=datetime.now(timezone.utc),
+        category="movie",
+    )
+    identity = MovieIdentity(
+        title="Test",
+        year=2020,
+        confidence=0.9,
+        needs_review=False,
+        original_filename="Test.2020.mkv",
+    )
+
+    normalized_low = str(p_720.parent / "." / p_720.name)
+    normalized_high = str(p_1080.parent / "." / p_1080.name)
+    plan = generate_plan(
+        [(vf_720, identity), (vf_1080, identity)],
+        config,
+        analysis_data={
+            "metadata": {"source_identities": "identities.json"},
+            "completeness": [],
+            "duplicates": [
+                {
+                    "identity": {"type": "movie", "title": "Test", "year": 2020},
+                    "files": [normalized_low, normalized_high],
+                    "quality_comparison": [
+                        {"path": normalized_low, "resolution": "1280x720", "size_bytes": 1000000},
+                        {"path": normalized_high, "resolution": "1920x1080", "size_bytes": 2000000},
+                    ],
+                }
+            ],
+        },
+    )
+
+    assert plan.summary["quarantine"] == 1
+    assert plan.operations[0].operation_type == "quarantine"
+    assert plan.operations[1].operation_type == "move"
+
+
 def test_generate_plan_human_summary_marks_disc_files_as_high_risk(config):
    """Disc/part files should be listed with a conservative risk note in summary."""
    config.duplicate_keep = "by_reputation"