# -*- coding: utf-8 -*-
"""原创离线合成样例；不连接模型、不读取账户。输出位于脚本旁。
已有输入保留，修改输入后可以重跑。失败边界由 --negative 单独演示。
"""
from pathlib import Path
import sys, json, csv, math, io, hashlib, html, base64, re
sys.stdout.reconfigure(encoding='utf-8')
sys.stderr.reconfigure(encoding='utf-8')
BASE = Path(__file__).resolve().parent
NEG = '--negative' in sys.argv
def write(name, value):
    p=BASE/name; p.parent.mkdir(parents=True,exist_ok=True)
    if isinstance(value,bytes): p.write_bytes(value)
    else: p.write_text(value,encoding='utf-8',newline='')
def seed(name,value):
    if not (BASE/name).exists(): write(name,value)
def jwrite(name,value): write(name,json.dumps(value,ensure_ascii=False,indent=2))
def jseed(name,value):
    if not (BASE/name).exists(): jwrite(name,value)
def load(name): return json.loads((BASE/name).read_text(encoding='utf-8'))
def table(name,heads,rows):
    f=io.StringIO(newline='');w=csv.writer(f);w.writerow(heads);w.writerows(rows);write(name,f.getvalue())
def page(name,title,body):
    write(name,'<!doctype html><html lang="zh-CN"><meta charset="utf-8"><meta name="viewport" content="width=device-width,initial-scale=1"><title>'+html.escape(title)+'</title><style>body{max-width:1000px;margin:2rem auto;padding:0 1rem;font:17px/1.7 system-ui}table{border-collapse:collapse}td,th{border:1px solid #aaa;padding:.5rem}pre{white-space:pre-wrap;overflow-wrap:anywhere}img,svg{max-width:100%;height:auto}a{color:#125caa}</style><h1>'+html.escape(title)+'</h1>'+body+'</html>')
def emit(data): print(json.dumps(data,ensure_ascii=False))

seed('ocr-lines.txt','The inter-\nnational team uses state-of-the-art tools.\nModel ZX-204 costs 10-20 units.\nA soft hy\u00adphen stays in one word.\nWe re-\nenter later.\n')
lines=(BASE/'ocr-lines.txt').read_text(encoding='utf-8').splitlines();approved={'international'};out=[];candidates=[];i=0
if NEG: lines=['item ZX-','204 starts next paragraph']
while i<len(lines):
    line=lines[i].replace('\u00ad','')
    m=re.search(r'([A-Za-z]+)-$',line)
    if m and i+1<len(lines):
        nxt=re.match(r'([a-z]+)',lines[i+1]);word=(m.group(1)+nxt.group(1)).lower() if nxt else '';decision='merge' if word in approved else 'manual';candidates.append([i+1,i+2,word,decision])
        if decision=='merge': out.append(line[:-1]+lines[i+1]);i+=2;continue
    out.append(line);i+=1
write('repaired.txt','\n'.join(out)+'\n');table('repair-candidates.csv',['line_a','line_b','word','decision'],candidates);emit({'text':'\n'.join(out),'candidates':candidates})