fix: restore identity_id after face_dedup, rebuild package v20260512

- Re-ran identity_bind.py to restore identity_id on face_detections
- Dedup cleanup had removed rows with identity_id, kept NULL rows
- 70691 face_detections now have identity_id, 428 identities
- Full package rebuild: 169MB sqlite, 1358MB tar.gz
- identities.json: 428 identities + 5483 bindings + 5483 trace maps
- TMDB matching complete: Audrey Hepburn 843 traces, Cary Grant 482

scripts/package_file.sh

@@ -0,0 +1,119 @@
+#!/bin/bash
+# Package File Content — single video's complete data
+set -euo pipefail
+UUID="${1:?Usage: $0 <file_uuid> [version]}"
+VERSION="${2:-v1.0.0}"
+PROJECT="/Users/accusys/momentry_core_0.1"
+OUTPUT="$PROJECT/release/files/$UUID/$VERSION"
+OUTPUT_DEV="/Users/accusys/momentry/output_dev"
+PG_BIN="/Users/accusys/pgsql/18.3/bin"
+T0=$(date +%s)
+
+mkdir -p "$OUTPUT/processors"
+
+echo "=== File Package ${UUID} ${VERSION} ==="
+
+# 1. metadata
+echo "[1/8] metadata.json..."
+$PG_BIN/psql -U accusys -d momentry -t -A -c "
+SELECT json_build_object(
+  'file_uuid', file_uuid,
+  'file_name', file_name,
+  'file_path', file_path,
+  'file_type', file_type,
+  'duration', duration,
+  'width', width,
+  'height', height,
+  'fps', fps,
+  'status', status,
+  'total_frames', total_frames,
+  'registration_time', registration_time::text
+) FROM dev.videos WHERE file_uuid='$UUID';
+" 2>/dev/null | python3 -c "import json,sys;d=json.load(sys.stdin);json.dump(d,open('$OUTPUT/metadata.json','w'),indent=2)" 2>/dev/null || echo "  WARN: no metadata"
+echo "  $(ls -lh "$OUTPUT/metadata.json" | awk '{print $5}')"
+
+# 2. Processor outputs
+echo "[2/8] Processor outputs..."
+for type in asr asrx asr-1 yolo face pose ocr cut scene; do
+  src="$OUTPUT_DEV/${UUID}.${type}.json"
+  if [ -f "$src" ]; then
+    cp "$src" "$OUTPUT/processors/"
+    echo "  ${type}.json"
+  fi
+done
+
+# 3. Identities (related to this file)
+echo "[3/8] Identities..."
+$PG_BIN/psql -U accusys -d momentry -c "
+COPY (
+  SELECT DISTINCT i.uuid, i.name, i.identity_type, i.source, i.status, i.metadata
+  FROM dev.identities i
+  JOIN dev.identity_bindings ib ON ib.identity_id = i.id
+  WHERE ib.file_uuid = '$UUID'
+) TO '$OUTPUT/identities.csv' WITH CSV HEADER;
+" 2>/dev/null && echo "  $(wc -l < "$OUTPUT/identities.csv") rows"
+
+# 4. Face detections
+echo "[4/8] Face detections..."
+$PG_BIN/psql -U accusys -d momentry -c "
+COPY (
+  SELECT id, file_uuid, frame_number, timestamp_secs, face_id, x, y, width, height, confidence, trace_id, identity_id
+  FROM dev.face_detections WHERE file_uuid = '$UUID'
+  ORDER BY frame_number
+) TO '$OUTPUT/face_detections.csv' WITH CSV HEADER;
+" 2>/dev/null && echo "  $(wc -l < "$OUTPUT/face_detections.csv") rows"
+
+# 5. Chunks
+echo "[5/8] Chunks..."
+$PG_BIN/psql -U accusys -d momentry -c "
+COPY (
+  SELECT chunk_id, chunk_type, start_frame, end_frame, start_time, end_time, fps, text_content
+  FROM dev.chunk WHERE file_uuid = '$UUID'
+  ORDER BY id
+) TO '$OUTPUT/chunks.csv' WITH CSV HEADER;
+" 2>/dev/null && echo "  $(wc -l < "$OUTPUT/chunks.csv") rows"
+
+# 6. Vectors
+echo "[6/8] chunk_vectors..."
+$PG_BIN/psql -U accusys -d momentry -c "
+COPY (
+  SELECT cv.chunk_id, cv.embedding::text
+  FROM dev.chunk_vectors cv
+  JOIN dev.chunk c ON c.file_uuid=cv.uuid AND c.chunk_id=cv.chunk_id
+  WHERE cv.uuid = '$UUID'
+) TO '$OUTPUT/chunk_vectors.csv' WITH CSV HEADER;
+" 2>/dev/null && echo "  $(wc -l < "$OUTPUT/chunk_vectors.csv") rows"
+
+# 7. TKG
+echo "[7/8] TKG..."
+$PG_BIN/psql -U accusys -d momentry -c "
+COPY (SELECT * FROM dev.tkg_nodes WHERE file_uuid='$UUID') TO '$OUTPUT/tkg_nodes.csv' WITH CSV HEADER;
+" 2>/dev/null
+$PG_BIN/psql -U accusys -d momentry -c "
+COPY (SELECT * FROM dev.tkg_edges WHERE file_uuid='$UUID') TO '$OUTPUT/tkg_edges.csv' WITH CSV HEADER;
+" 2>/dev/null
+echo "  nodes+edges exported"
+
+# 8. RELEASE_INFO
+echo "[8/8] RELEASE_INFO..."
+SENTENCE=$($PG_BIN/psql -U accusys -d momentry -t -A -c "SELECT count(*) FROM dev.chunk WHERE file_uuid='$UUID' AND chunk_type='sentence';" 2>/dev/null)
+VECTORS=$($PG_BIN/psql -U accusys -d momentry -t -A -c "SELECT count(*) FROM dev.chunk_vectors cv JOIN dev.chunk c ON c.file_uuid=cv.uuid AND c.chunk_id=cv.chunk_id WHERE cv.uuid='$UUID';" 2>/dev/null)
+cat > "$OUTPUT/RELEASE_INFO.txt" << EOF
+Release: ${VERSION}
+Type: file
+UUID: ${UUID}
+Date: $(date +%Y-%m-%d)
+
+Chunks: sentence=${SENTENCE}
+Vectors: ${VECTORS}
+Processors: $(ls "$OUTPUT/processors/" 2>/dev/null | wc -l | tr -d ' ')
+EOF
+
+# Symlink latest
+ln -sfn "$OUTPUT" "$PROJECT/release/files/$UUID/latest"
+
+ELAPSED=$(($(date +%s) - T0))
+echo ""
+echo "=== File Package done (${ELAPSED}s) ==="
+echo "  $OUTPUT"
+du -sh "$OUTPUT"