{
"event": "PostToolUse",
"tool_name": "Read",
"tool_response": {
"type": "text",
"file": {
"filePath": "\/var\/www\/scripts\/pipeline\/enrich.py",
"content": "#!\/usr\/bin\/env python3\n\"\"\"\nVision-Enrichment module for KI-System Pipeline.\nEnriches chunks with visual context from page-level vision analysis.\n\nUsage:\n python enrich.py <document_id> # Enrich all chunks for a document\n python enrich.py --all # Enrich all documents with vision data\n\"\"\"\n\nimport json\nimport re\nimport sys\n\nfrom db import db\n\n\ndef get_vision_context(document_id, page_number):\n \"\"\"\n Get vision analysis for a specific page.\n\n Returns dict with structured vision info or None.\n \"\"\"\n cursor = db.execute(\n \"\"\"SELECT vision_analysis\n FROM document_pages\n WHERE document_id = %s AND page_number = %s\"\"\",\n (document_id, page_number),\n )\n result = cursor.fetchone()\n cursor.close()\n\n if not result or not result.get(\"vision_analysis\"):\n return None\n\n try:\n vision_data = json.loads(result[\"vision_analysis\"])\n return vision_data\n except (json.JSONDecodeError, TypeError):\n return None\n\n\ndef extract_vision_summary(vision_data):\n \"\"\"\n Extract key information from vision analysis for chunk enrichment.\n\n Returns compact dict with:\n - detected_elements: list of visual elements found\n - page_title: extracted title if any\n - has_images: bool",
"numLines": 50,
"startLine": 1,
"totalLines": 298
}
}
}