wing-ops/backend/scripts/hns-import/merge-batch.py

24 lines
658 B
Python

"""배치 JSON을 ocr.json에 병합."""
import json
import sys
from pathlib import Path
SCRIPT_DIR = Path(__file__).parent.resolve()
OUT_DIR = SCRIPT_DIR / 'out'
OCR_PATH = OUT_DIR / 'ocr.json'
BATCH_PATH = OUT_DIR / 'batch.json'
with open(OCR_PATH, encoding='utf-8') as f:
ocr = json.load(f)
with open(BATCH_PATH, encoding='utf-8') as f:
batch = json.load(f)
added = [k for k in batch if k not in ocr]
updated = [k for k in batch if k in ocr]
ocr.update(batch)
with open(OCR_PATH, 'w', encoding='utf-8') as f:
json.dump(ocr, f, ensure_ascii=False, indent=2)
print(f'merged: +{len(added)} added, ~{len(updated)} updated, total {len(ocr)}')