import json,re
from pathlib import Path
from collections import Counter

base=Path(r"C:\Apache24\htdocs\.codex_tmp")
rows=json.loads((base/'ocr_merged.json').read_text(encoding='utf-8'))
trans=str.maketrans('٠١٢٣٤٥٦٧٨٩۰۱۲۳۴۵۶۷۸۹','01234567890123456789')
rot={}
for angle in (90,180,270):
    data=json.loads((base/f'ocr_rot{angle}_results.json').read_text(encoding='utf-8-sig'))
    for r in data:
        sid=int(r['file'].split('_')[1]); text=(r.get('text') or '').translate(trans)
        cs=[]
        for idx,line in enumerate(text.split('|')):
            digits=re.sub(r'\D','',line)
            if 6<=len(digits)<=9 and idx<=2: cs.append(digits)
        rot.setdefault(sid,[]).extend(cs[:1])
for r in rows:
    if r['chosen']: continue
    vals=rot.get(r['id'],[])
    counts=Counter(vals)
    if vals and (len(counts)==1 or counts.most_common(1)[0][1]>=2):
        r['chosen']=counts.most_common(1)[0][0]
        r['source']='rotation-consensus'
    r['rotation_candidates']=vals
(base/'ocr_final_candidates.json').write_text(json.dumps(rows,ensure_ascii=False,indent=2),encoding='utf-8')
print(json.dumps({'resolved':sum(bool(r['chosen']) for r in rows),'unresolved':sum(not r['chosen'] for r in rows),'rotation_conflicts':sum(not r['chosen'] and len(set(r.get('rotation_candidates',[])))>1 for r in rows)},ensure_ascii=False))