import json, re
from pathlib import Path

data=json.loads(Path(r"C:\Apache24\htdocs\.codex_tmp\ocr_results.json").read_text(encoding='utf-8-sig'))
trans=str.maketrans('٠١٢٣٤٥٦٧٨٩۰۱۲۳۴۵۶۷۸۹','01234567890123456789')
rows=[]
for r in data:
    text=(r.get('text') or '').translate(trans)
    compact=re.sub(r'[^0-9|]+','',text)
    candidates=[]
    for line in text.split('|'):
        digits=re.sub(r'\D','',line)
        if 6 <= len(digits) <= 9:
            candidates.append(digits)
    sid=int(r['file'].split('_')[1])
    rows.append({'id':sid,'file':r['file'],'text':text,'candidates':candidates,'chosen':candidates[0] if candidates else ''})
out=Path(r"C:\Apache24\htdocs\.codex_tmp\ocr_candidates.json")
out.write_text(json.dumps(rows,ensure_ascii=False,indent=2),encoding='utf-8')
print(json.dumps({'rows':len(rows),'with_candidate':sum(bool(r['chosen']) for r in rows),'without':sum(not r['chosen'] for r in rows),'multi':sum(len(r['candidates'])>1 for r in rows)},ensure_ascii=False))
print(json.dumps([r for r in rows if not r['chosen']][:20],ensure_ascii=False))