# ==============================================================================
# Author: Simon-Pierre Boucher <contact@spboucher.ai>
# File:   creaka/connectors/tiktok.py
# Desc:   Connecteur ENRICHISSEMENT TikTok — profil public par créateur :
#         abonnés, badge vérifié, bio, lien de bio (découverte croisée).
#         Mode d'accès : Scrapfly (plateforme hostile au scraping, API
#         officielle limitée aux comptes connectés) — palier 4 (§9), §15.
# ==============================================================================
"""Enrichissement TikTok, créateur par créateur.

La page publique tiktok.com/@handle embarque un JSON d'hydratation
(__UNIVERSAL_DATA_FOR_REHYDRATION__) contenant stats.followerCount, la bio
(`signature`), le badge vérifié et le lien de bio (`bioLink`). On le lit via
Scrapfly. Lien de bio → page link-in-bio ou compte croisé (signal cross_link).
"""
from __future__ import annotations

from datetime import datetime, timedelta, timezone

import json
import re

from ..identity import account, merge_accounts
from ..normalize import parse_count, platform_from_url
from ..schema import Creator, now_iso
from .base import BaseConnector
from .linkinbio import is_supported as is_linkinbio

PROFILE_URL = "https://www.tiktok.com/@{h}"
_STATE_RE = re.compile(
    r'<script id="__UNIVERSAL_DATA_FOR_REHYDRATION__"[^>]*>(.*?)</script>', re.S)


def parse_profile(html: str) -> dict | None:
    """HTML de profil TikTok → dict userInfo {user:…, stats:…} (None si absent)."""
    m = _STATE_RE.search(html)
    if not m:
        return None
    try:
        data = json.loads(m.group(1))
        return (data.get("__DEFAULT_SCOPE__", {})
                    .get("webapp.user-detail", {})
                    .get("userInfo")) or None
    except Exception:
        return None



def _fresh(last_checked: str, hours: int = 72) -> bool:
    """Compte déjà rafraîchi récemment (acteur Apify) — épargner Scrapfly."""
    if not last_checked:
        return False
    try:
        seen = datetime.fromisoformat(last_checked.replace("Z", "+00:00"))
    except ValueError:
        return False
    return datetime.now(timezone.utc) - seen < timedelta(hours=hours)

class TikTokConnector(BaseConnector):
    source_id = "tiktok-profil"
    kind = "enrichment"
    request_delay = 1.5
    max_profiles = 700

    def enrich(self, creators: list[Creator]) -> list[Creator]:
        enriched: list[Creator] = []
        fetched = 0
        for cr in creators:
            tk = next((a for a in cr.platforms if a.platform == "tiktok"), None)
            if tk is None or fetched >= self.max_profiles:
                continue
            if _fresh(tk.last_checked):
                continue
            try:
                result = self.scrapfly(PROFILE_URL.format(h=tk.handle),
                                       render_js=False)
            except Exception:
                continue
            fetched += 1
            if result.get("status_code") != 200:
                continue
            info = parse_profile(result.get("content") or "")
            if not info:
                continue
            user = info.get("user") or {}
            stats = info.get("stats") or {}
            if user.get("privateAccount"):
                continue  # §15 : uniquement du public
            tk.followers = parse_count(stats.get("followerCount")) or tk.followers
            tk.verified = bool(user.get("verified"))
            tk.last_checked = now_iso()
            extra = {
                "following": parse_count(stats.get("followingCount")),
                "likes": parse_count(stats.get("heartCount")),
                "videos": parse_count(stats.get("videoCount")),
            }
            tk.metrics.update({k: v for k, v in extra.items() if v is not None})
            if not cr.avatar_url and user.get("avatarLarger"):
                cr.avatar_url = user["avatarLarger"]
            if not cr.bio and user.get("signature"):
                cr.bio = user["signature"]
            url = ((user.get("bioLink") or {}).get("link") or "").strip()
            if url and not url.startswith("http"):
                url = "https://" + url
            if url:
                if is_linkinbio(url) and not cr.link_in_bio_url:
                    cr.link_in_bio_url = url
                else:
                    hit = platform_from_url(url)
                    if hit and hit[0] != "tiktok":
                        cr.platforms = merge_accounts(cr.platforms, [
                            account(hit[0], hit[1], "cross_link",
                                    url=url).finalize()])
            enriched.append(cr)
        return enriched