Initial commit

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

scripts/backfill_dimensions.py

@@ -0,0 +1,145 @@
+#!/usr/bin/env python3
+"""
+Backfill missing dimensions (width/height) for files in file_inventory.
+
+Uses PIL for images and ffprobe for videos.
+"""
+
+import os
+import sys
+from pathlib import Path
+
+# Bootstrap PostgreSQL adapter before any sqlite3 imports
+sys.path.insert(0, '/opt/media-downloader')
+from modules.db_bootstrap import bootstrap_database
+bootstrap_database()
+import sqlite3
+
+# Database path (routed to PostgreSQL via pgadapter)
+DB_PATH = "/opt/media-downloader/database/media_downloader.db"
+
+
+def get_image_dimensions(file_path: str) -> tuple:
+    """Get dimensions for an image file using PIL."""
+    try:
+        from PIL import Image
+        with Image.open(file_path) as img:
+            return img.size  # (width, height)
+    except Exception as e:
+        print(f"  PIL error for {file_path}: {e}")
+        return None, None
+
+
+def get_video_dimensions(file_path: str) -> tuple:
+    """Get dimensions for a video file using ffprobe."""
+    try:
+        import subprocess
+        result = subprocess.run(
+            ['ffprobe', '-v', 'error', '-select_streams', 'v:0',
+             '-show_entries', 'stream=width,height', '-of', 'csv=p=0',
+             file_path],
+            capture_output=True, text=True, timeout=30
+        )
+        if result.returncode == 0 and result.stdout.strip():
+            parts = result.stdout.strip().split(',')
+            if len(parts) >= 2:
+                return int(parts[0]), int(parts[1])
+    except Exception as e:
+        print(f"  ffprobe error for {file_path}: {e}")
+    return None, None
+
+
+def main():
+    if not os.path.exists(DB_PATH):
+        print(f"Database not found: {DB_PATH}")
+        sys.exit(1)
+
+    conn = sqlite3.connect(DB_PATH)
+    conn.row_factory = sqlite3.Row
+    cursor = conn.cursor()
+
+    # Get count of files missing dimensions
+    cursor.execute('''
+        SELECT COUNT(*) FROM file_inventory
+        WHERE (width IS NULL OR height IS NULL)
+        AND location IN ('final', 'review')
+    ''')
+    total_missing = cursor.fetchone()[0]
+    print(f"Found {total_missing} files with missing dimensions")
+
+    if total_missing == 0:
+        print("No files need dimension backfill!")
+        conn.close()
+        return
+
+    # Process in batches
+    batch_size = 100
+    processed = 0
+    updated = 0
+    errors = 0
+
+    cursor.execute('''
+        SELECT id, file_path, content_type FROM file_inventory
+        WHERE (width IS NULL OR height IS NULL)
+        AND location IN ('final', 'review')
+    ''')
+
+    update_cursor = conn.cursor()
+    batch_updates = []
+
+    for row in cursor:
+        file_id = row['id']
+        file_path = row['file_path']
+        content_type = row['content_type']
+
+        if not os.path.exists(file_path):
+            processed += 1
+            continue
+
+        width, height = None, None
+
+        if content_type == 'image':
+            width, height = get_image_dimensions(file_path)
+        elif content_type == 'video':
+            width, height = get_video_dimensions(file_path)
+        else:
+            # Try to determine from extension
+            ext = Path(file_path).suffix.lower()
+            image_exts = {'.jpg', '.jpeg', '.png', '.gif', '.heic', '.heif', '.webp', '.bmp', '.tiff'}
+            if ext in image_exts:
+                width, height = get_image_dimensions(file_path)
+            else:
+                width, height = get_video_dimensions(file_path)
+
+        if width and height:
+            batch_updates.append((width, height, file_id))
+            updated += 1
+        else:
+            errors += 1
+
+        processed += 1
+
+        # Commit in batches
+        if len(batch_updates) >= batch_size:
+            update_cursor.executemany(
+                'UPDATE file_inventory SET width = ?, height = ? WHERE id = ?',
+                batch_updates
+            )
+            conn.commit()
+            print(f"  Processed {processed}/{total_missing} files, updated {updated}, errors {errors}")
+            batch_updates = []
+
+    # Final batch
+    if batch_updates:
+        update_cursor.executemany(
+            'UPDATE file_inventory SET width = ?, height = ? WHERE id = ?',
+            batch_updates
+        )
+        conn.commit()
+
+    print(f"\nComplete! Processed {processed} files, updated {updated} with dimensions, {errors} errors")
+    conn.close()
+
+
+if __name__ == '__main__':
+    main()