add scanner metadata cache reuse and control flags

tests/test_cli_scan.py

@@ -0,0 +1,89 @@
+"""Tests for CLI scan command options."""
+
+from pathlib import Path
+from unittest.mock import patch
+
+from click.testing import CliRunner
+
+from vlm.cli import main
+
+
+def test_scan_no_metadata_passes_flag_to_scanner(tmp_path):
+    """Test `scan --no-metadata` disables ffprobe metadata extraction."""
+    library_root = tmp_path / "library"
+    library_root.mkdir(parents=True)
+
+    config_file = tmp_path / "config.yaml"
+    config_file.write_text(
+        f"""
+library_root: {library_root}
+video_extensions:
+  - .mp4
+categories:
+  movie: [movie, movies]
+  series: [series, tv, shows]
+  anime: [anime]
+""".strip()
+    )
+
+    runner = CliRunner()
+    with patch("vlm.scanner.scan_library", return_value=[]) as mock_scan, patch(
+        "vlm.scanner.save_inventory_csv"
+    ) as mock_save:
+        result = runner.invoke(
+            main,
+            ["--config", str(config_file), "scan", "--no-metadata", "--output", str(tmp_path / "inv.csv")]
+        )
+
+    assert result.exit_code == 0
+    assert mock_scan.call_count == 1
+    _, kwargs = mock_scan.call_args
+    assert kwargs["include_video_metadata"] is False
+    assert "progress_callback" in kwargs
+    assert mock_save.call_count == 1
+
+
+def test_scan_force_refresh_metadata_disables_cache(tmp_path):
+    """Test `scan --force-refresh-metadata` skips cache loading."""
+    library_root = tmp_path / "library"
+    library_root.mkdir(parents=True)
+
+    config_file = tmp_path / "config.yaml"
+    config_file.write_text(
+        f"""
+library_root: {library_root}
+video_extensions:
+  - .mp4
+categories:
+  movie: [movie, movies]
+  series: [series, tv, shows]
+  anime: [anime]
+""".strip()
+    )
+
+    output_file = tmp_path / "inventory.csv"
+    output_file.write_text(
+        "# Generated: 2026-02-09T00:00:00\n"
+        f"# Library Root: {library_root}\n"
+        "path,filename,size_bytes,modified_timestamp,category,resolution,codec,duration_seconds,bitrate_kbps\n"
+    )
+
+    runner = CliRunner()
+    with patch("vlm.scanner.load_inventory_csv") as mock_load_cache, patch(
+        "vlm.scanner.scan_library", return_value=[]
+    ) as mock_scan, patch("vlm.scanner.save_inventory_csv") as mock_save:
+        result = runner.invoke(
+            main,
+            [
+                "--config", str(config_file), "scan",
+                "--output", str(output_file),
+                "--force-refresh-metadata"
+            ]
+        )
+
+    assert result.exit_code == 0
+    mock_load_cache.assert_not_called()
+    _, kwargs = mock_scan.call_args
+    assert kwargs["include_video_metadata"] is True
+    assert kwargs["metadata_cache"] is None
+    assert mock_save.call_count == 1

tests/test_scanner.py

@@ -3,6 +3,7 @@
 import os
 import tempfile
 import time
+import csv
 from datetime import datetime, timezone
 from pathlib import Path
 from unittest.mock import patch, MagicMock
@@ -17,6 +18,8 @@ from vlm.scanner import (
     categorize_file,
     scan_library,
     extract_metadata,
+    save_inventory_csv,
+    load_inventory_csv,
 )
 
 
@@ -698,6 +701,63 @@ class TestExtractMetadata:
             assert vf.duration_seconds is None
             assert vf.bitrate_kbps is None
 
+    def test_scan_library_with_metadata_disabled(self, tmp_path):
+        """Test that metadata extraction can be disabled for faster scans."""
+        movie_dir = tmp_path / "movie"
+        movie_dir.mkdir()
+
+        video_file = movie_dir / "test.mp4"
+        video_file.touch()
+
+        config = Config(library_root=tmp_path)
+
+        with patch('subprocess.run') as mock_run:
+            result = scan_library(tmp_path, config, include_video_metadata=False)
+
+        assert len(result) == 1
+        assert result[0].filename == "test.mp4"
+        assert result[0].resolution is None
+        assert result[0].codec is None
+        assert result[0].duration_seconds is None
+        assert result[0].bitrate_kbps is None
+        mock_run.assert_not_called()
+
+    def test_scan_library_reuses_cached_metadata_when_unchanged(self, tmp_path):
+        """Test unchanged files reuse metadata from previous inventory cache."""
+        movie_dir = tmp_path / "movie"
+        movie_dir.mkdir()
+        video_file = movie_dir / "test.mp4"
+        video_file.write_text("test")
+
+        original = VideoFile(
+            path=video_file,
+            filename=video_file.name,
+            size_bytes=video_file.stat().st_size,
+            modified_timestamp=datetime.fromtimestamp(video_file.stat().st_mtime, tz=timezone.utc),
+            category="movie",
+            resolution="1920x1080",
+            codec="h264",
+            duration_seconds=120.5,
+            bitrate_kbps=5000,
+        )
+
+        cache_file = tmp_path / "cached_inventory.csv"
+        save_inventory_csv([original], cache_file, tmp_path)
+        cached_entries = load_inventory_csv(cache_file)
+        metadata_cache = {str(vf.path): vf for vf in cached_entries}
+
+        config = Config(library_root=tmp_path)
+        with patch("subprocess.run") as mock_run:
+            result = scan_library(tmp_path, config, metadata_cache=metadata_cache)
+
+        assert len(result) == 1
+        scanned = result[0]
+        assert scanned.resolution == "1920x1080"
+        assert scanned.codec == "h264"
+        assert scanned.duration_seconds == 120.5
+        assert scanned.bitrate_kbps == 5000
+        mock_run.assert_not_called()
+
 
 
 class TestInventoryReports: