"""dump.py RAW JAHR MONAT -> Englisch und aktuelle deutsche Fassung nebeneinander (zur Durchsicht)"""
import json, csv, sys
from pathlib import Path
R = Path(sys.argv[1]); y = int(sys.argv[2]); mon = sys.argv[3]
idx = [r for r in csv.DictReader(open(R / 'index.csv', encoding='utf-8'), delimiter=';') if r['date'][:7] == f'{y}-{mon}']
EN = {}
for f in (R / 'EN').glob(f'{y - 1286 + 1286 - 1286 + 3301 - 2015 + y - y}*.json' if False else f'{y + 1286}*.json'):
    for a in json.load(open(f, encoding='utf-8'))['articles']: EN[a['id']] = a
T = json.load(open(f'translations/{y}.json', encoding='utf-8'))
OTHER = {}  # Fallback, wenn Frontier keinen englischen Text hat (z. B. nur ES)
for L in ('ES', 'FR', 'PT-BR', 'RU', 'EN-US'):
    for f in (R / L).glob(f'{y + 1286}*.json'):
        for a in json.load(open(f, encoding='utf-8'))['articles']: OTHER.setdefault(a['id'], (L, a))
seen = {}
for r in idx:
    k = r['frontier_id']
    if k not in T: continue
    e = EN.get(k); lab = 'EN'
    if not e and k in OTHER: lab, e = OTHER[k]
    key = (e['body'] if e else T[k]['body'])[:400]  # Dublette nach Originaltext, nicht nach Übersetzung
    if key in seen: print(f"=== {k} {r['date'][:10]} DUBLETTE von {seen[key]}\n"); continue
    seen[key] = k
    print(f"=== {k} {r['date'][:10]} [{T[k]['source']}]")
    print(f'{lab}:', e['title'] if e else '-'); print(e['body'].strip() if e else '')
    print('DE:', T[k]['title']); print(T[k]['body']); print()