"""E8 · cotejo con la fuente (§7.2): para los casos de la muestra con ficha de la PLACSP descargada, ¿aparecen en la ficha el
objeto (sus primeras palabras) y el importe sin IVA de nuestra fila? Sin red. Uso: c5_cotejo_fuente.py <mercado>.
Salida: priv/cotejo_fuente_<mercado>.csv (lo mira una persona; el script solo busca)."""
import csv, html, re, sys, unicodedata, os
m = sys.argv[1]
def nz(t):
    t = unicodedata.normalize('NFKD', (t or '').lower())
    t = ''.join(c for c in t if not unicodedata.combining(c))
    return re.sub(r'\s+', ' ', re.sub(r'[^a-z0-9,.€ ]', ' ', t)).strip()
def formatos(x):
    v = float(x); s = []
    for d in (2,):
        a = f'{v:,.{d}f}'                       # 1,234.56
        s += [a.replace(',', 'X').replace('.', ',').replace('X', '.'), a.replace(',', ''), a.replace(',', '').replace('.', ',')]
    if v == int(v):
        s += [f'{int(v):,}'.replace(',', '.'), str(int(v))]
    return set(s)
out = []
for r in csv.DictReader(open(f'/p/priv/cotejo_{m}.csv')):
    f = f"/ref/cotejo/{m}/{m}_{int(r['orden']):02d}.html"
    if not os.path.exists(f):
        continue
    raw = open(f, encoding='utf-8', errors='replace').read()
    txt = html.unescape(re.sub(r'<[^>]+>', ' ', raw))
    t = nz(txt)
    obj = nz(r['objeto'].split(' — ')[0])[:70]
    pal = [w for w in obj.split() if len(w) > 3][:8]
    obj_ok = obj[:40] in t or (len(pal) >= 3 and sum(w in t for w in pal) >= len(pal) - 1)
    imp_ok = any(s in txt for s in formatos(r['imp_sin_iva']))
    ctx = ' | '.join(re.sub(r'\s+', ' ', txt[mm.start():mm.start() + 160]) for mm in re.finditer(r'Importe', txt))[:600]
    out.append(dict(orden=r['orden'], objeto_en_ficha=obj_ok, importe_en_ficha=imp_ok, imp=r['imp_sin_iva'], contexto=ctx))
with open(f'/p/priv/cotejo_fuente_{m}.csv', 'w', newline='') as fh:
    w = csv.DictWriter(fh, fieldnames=['orden', 'objeto_en_ficha', 'importe_en_ficha', 'imp', 'contexto']); w.writeheader(); w.writerows(out)
print(m, 'fichas', len(out), 'objeto', sum(o['objeto_en_ficha'] for o in out), 'importe', sum(o['importe_en_ficha'] for o in out))
for o in out:
    if not (o['objeto_en_ficha'] and o['importe_en_ficha']): print(' ', o['orden'], o['objeto_en_ficha'], o['importe_en_ficha'], o['imp'], o['contexto'][:300])