feat: isolate probe failures in reports

tests/test_analysis.py

@@ -1,7 +1,7 @@
 from pathlib import Path
 
 from video_metadata_analyzer.analysis import build_summary
-from video_metadata_analyzer.models import VideoMetadata
+from video_metadata_analyzer.models import ProbeFailure, VideoMetadata
 
 
 def test_build_summary_computes_totals_and_histograms() -> None:
@@ -81,3 +81,38 @@ def test_build_summary_computes_totals_and_histograms() -> None:
         ("1280x720", 1),
         ("3840x2160", 1),
     ]
+
+
+def test_build_summary_tracks_successful_and_failed_files() -> None:
+    records = [
+        VideoMetadata(
+            path=Path("a.mp4"),
+            size_bytes=1024,
+            duration_seconds=10,
+            format_name="mov,mp4,m4a,3gp,3g2,mj2",
+            bit_rate=1_000_000,
+            video_codec="h264",
+            audio_codec="aac",
+            width=1920,
+            height=1080,
+            frame_rate=30.0,
+            video_stream_count=1,
+            audio_stream_count=1,
+            subtitle_stream_count=0,
+        )
+    ]
+    failures = [
+        ProbeFailure(
+            path=Path("broken.mp4"),
+            error_type="CalledProcessError",
+            error_message="moov atom not found",
+        )
+    ]
+
+    summary = build_summary(records, failures=failures)
+
+    assert summary.total_files == 2
+    assert summary.successful_files == 1
+    assert summary.failed_files == 1
+    assert summary.total_size_bytes == 1024
+    assert summary.total_duration_seconds == 10

tests/test_cli.py

@@ -1,6 +1,7 @@
 from __future__ import annotations
 
 import json
+import subprocess
 from pathlib import Path
 
 from video_metadata_analyzer.cli import main
@@ -50,3 +51,66 @@ def test_main_scans_input_directory_and_writes_reports(tmp_path: Path) -> None:
     assert (output_dir / "report.json").exists()
     assert (output_dir / "files.csv").exists()
     assert (output_dir / "report.html").exists()
+
+
+def test_main_continues_when_single_file_probe_fails(tmp_path: Path) -> None:
+    input_dir = tmp_path / "videos"
+    input_dir.mkdir()
+    good_sample = input_dir / "good.mp4"
+    bad_sample = input_dir / "bad.mp4"
+    good_sample.write_bytes(b"x" * 2048)
+    bad_sample.write_bytes(b"y" * 1024)
+    output_dir = tmp_path / "output"
+
+    def fake_runner(command: list[str]) -> str:
+        if command[:2] == ["ffprobe", "-version"]:
+            return "ffprobe version 8.0.1\nbuilt with ...\n"
+        if command[:2] == ["ffmpeg", "-version"]:
+            return "ffmpeg version 8.0.1\nbuilt with ...\n"
+        if command[0] == "ffprobe" and command[-1] == str(good_sample):
+            return json.dumps(
+                {
+                    "format": {
+                        "filename": str(good_sample),
+                        "format_name": "mov,mp4,m4a,3gp,3g2,mj2",
+                        "duration": "8.000000",
+                        "bit_rate": "1024000",
+                        "size": "2048",
+                    },
+                    "streams": [
+                        {
+                            "codec_type": "video",
+                            "codec_name": "h264",
+                            "width": 1280,
+                            "height": 720,
+                            "avg_frame_rate": "25/1",
+                        }
+                    ],
+                }
+            )
+        if command[0] == "ffprobe" and command[-1] == str(bad_sample):
+            raise subprocess.CalledProcessError(
+                returncode=1,
+                cmd=command,
+                stderr="moov atom not found",
+            )
+        raise AssertionError(f"unexpected command: {command}")
+
+    exit_code = main(
+        [str(input_dir), "--output-dir", str(output_dir)],
+        command_runner=fake_runner,
+    )
+
+    assert exit_code == 0
+    payload = json.loads((output_dir / "report.json").read_text(encoding="utf-8"))
+    assert payload["summary"]["total_files"] == 2
+    assert payload["summary"]["successful_files"] == 1
+    assert payload["summary"]["failed_files"] == 1
+    assert payload["files"][0]["path"] == good_sample.as_posix()
+    assert payload["failures"] == [
+        {
+            "path": bad_sample.as_posix(),
+            "error_type": "CalledProcessError",
+            "error_message": "moov atom not found",
+        }
+    ]

tests/test_report.py

@@ -4,7 +4,7 @@ import json
 from pathlib import Path
 
 from video_metadata_analyzer.analysis import build_summary
-from video_metadata_analyzer.models import VideoMetadata
+from video_metadata_analyzer.models import ProbeFailure, VideoMetadata
 from video_metadata_analyzer.reporting import write_report_bundle
 
 
@@ -25,12 +25,18 @@ def test_write_report_bundle_creates_json_csv_and_html(tmp_path: Path) -> None:
         audio_stream_count=1,
         subtitle_stream_count=0,
     )
-    summary = build_summary([record])
+    failure = ProbeFailure(
+        path=Path("videos/broken.mp4"),
+        error_type="CalledProcessError",
+        error_message="moov atom not found",
+    )
+    summary = build_summary([record], failures=[failure])
 
     paths = write_report_bundle(
         output_dir=output_dir,
         scanned_inputs=[Path("videos")],
         records=[record],
+        failures=[failure],
         summary=summary,
         tool_versions={"ffprobe": "ffprobe version x", "ffmpeg": "ffmpeg version y"},
     )
@@ -41,15 +47,23 @@ def test_write_report_bundle_creates_json_csv_and_html(tmp_path: Path) -> None:
     assert paths["html"].exists()
 
     payload = json.loads(paths["json"].read_text(encoding="utf-8"))
-    assert payload["summary"]["total_files"] == 1
+    assert payload["summary"]["total_files"] == 2
+    assert payload["summary"]["successful_files"] == 1
+    assert payload["summary"]["failed_files"] == 1
     assert payload["tool_versions"]["ffmpeg"] == "ffmpeg version y"
     assert payload["files"][0]["path"] == "videos/sample.mp4"
+    assert payload["failures"][0]["path"] == "videos/broken.mp4"
+    assert payload["failures"][0]["error_message"] == "moov atom not found"
 
     csv_text = paths["csv"].read_text(encoding="utf-8")
-    assert "path,size_bytes,duration_seconds" in csv_text
+    assert "status,path,size_bytes,duration_seconds" in csv_text
     assert "videos/sample.mp4" in csv_text
+    assert "videos/broken.mp4" in csv_text
+    assert "CalledProcessError" in csv_text
 
     html_text = paths["html"].read_text(encoding="utf-8")
     assert "视频元数据分析报告" in html_text
     assert "时长分布" in html_text
+    assert "错误清单" in html_text
+    assert "moov atom not found" in html_text
     assert "<svg" in html_text