Add job tracking with PostgreSQL, image storage, and review workflow

- Add PostgreSQL service to docker-compose with health check and postgres_data volume
- Mount ./ocr_images as bind volume for persistent image storage
- Add backend/database.py with schema init and get_db() context manager
- Add 5 new API endpoints: POST /api/jobs, GET /api/jobs (search), GET /api/jobs/{id},
  GET /api/jobs/{id}/image, PUT /api/jobs/{id}/review
- Jobs are saved with author/book/chapter/page metadata, auto UUID, and submitted_at timestamp
- Jobs start as 'unreviewed'; review captures edited text, reviewer name, and reviewed_at
- Add MetadataForm.jsx (author/book/chapter/page inputs) to the New Job panel
- Add JobsPanel.jsx with search/filter, paginated list, and detail pane with review form
- Add "Commit Job" button to ResultPanel (plain_ocr mode only) with success/error feedback
- Add "New Job" / "Browse Jobs" navigation to the app header

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

.env.example

@@ -17,6 +17,15 @@ CORS_ORIGINS=http://localhost:3000
 # Upload Configuration
 MAX_UPLOAD_SIZE_MB=100
 
+# PostgreSQL Configuration
+POSTGRES_USER=ocr_user
+POSTGRES_PASSWORD=ocr_password
+POSTGRES_DB=ocr_db
+DATABASE_URL=postgresql://ocr_user:ocr_password@postgres:5432/ocr_db
+
+# OCR Image Storage (host path mounted into container)
+OCR_IMAGES_DIR=/data/ocr_images
+
 # Processing Configuration
 BASE_SIZE=1024
 IMAGE_SIZE=640