#!/usr/bin/env python3
"""Cek apakah URL tertentu sudah pernah dipakai di jadwal Repliz/log Threads."""
import json, re, sys
from urllib.parse import urlparse

URL_RE = re.compile(r'https?://[^\s"\'\\]+')


def norm(u):
    u = u.strip().rstrip('").,;"\'')
    p = urlparse(u)
    return "%s%s" % (p.netloc.lower().replace("www.", ""), p.path.rstrip("/").lower())


docs = json.load(open("/tmp/repliz_all_20260928.json"))
used = set()
for d in docs:
    for m in URL_RE.findall(json.dumps(d, ensure_ascii=False)):
        n = norm(m)
        if n:
            used.add(n)
log = json.load(open("/home/mantapman/.hermes/profiles/atun/workspace/log_threads_harian.json"))
for date, accs in (log.get("threads_used_links") or {}).items():
    for acc, items in accs.items():
        for it in items:
            used.add(norm(it.get("url", "")))

for u in sys.argv[1:]:
    print("%s -> %s" % ("SUDAH DIPAKAI" if norm(u) in used else "FRESH", u))
