#!/usr/bin/env python3
"""Enrichissement des boutiques : logo, image de couverture, description.

Pour chaque boutique du registre : télécharge la page d'accueil, extrait
  - logo : apple-touch-icon > icon (plus grande taille) > og:logo
  - couverture : og:image
  - description : meta description > og:description
  - nom d'affichage : og:site_name (si plus propre que le nom actuel)
Écrit dans la table stores (colonnes logo_url, cover_url, description_meta).
Cache disque data/enrich_cache/<domain>.json ; Scrapfly en secours pour les 403.
"""
import concurrent.futures as cf
import json
import os
import re
import sys
from urllib.parse import urljoin

import requests

ROOT = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
sys.path.insert(0, ROOT)
CACHE = os.path.join(ROOT, "data", "enrich_cache")
os.makedirs(CACHE, exist_ok=True)

for line in open(os.path.join(ROOT, ".env")).read().splitlines():
    if "=" in line and not line.startswith("#"):
        k, _, v = line.partition("=")
        os.environ.setdefault(k.strip(), v.strip())

from fabrika import db as fdb  # noqa: E402
from fabrika.connectors.scrapfly import scrapfly_get, available  # noqa: E402

HDRS = {"User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/126.0.0.0 Safari/537.36",
        "Accept-Language": "fr-CA,fr;q=0.9,en;q=0.8"}

LINK_RE = re.compile(r'<link[^>]+>', re.I)
META_RE = re.compile(r'<meta[^>]+>', re.I)


def attr(tag, name):
    m = re.search(name + r'\s*=\s*["\']([^"\']+)["\']', tag, re.I)
    return m.group(1) if m else None


def extract(base_url, html):
    head = html[:120000]
    logo, cover, desc, site_name = None, None, None, None
    icons = []
    for tag in LINK_RE.findall(head):
        rel = (attr(tag, "rel") or "").lower()
        href = attr(tag, "href")
        if not href:
            continue
        if "apple-touch-icon" in rel:
            icons.append((200 if not attr(tag, "sizes") else int((attr(tag, "sizes") or "180x").split("x")[0] or 180), href))
        elif rel in ("icon", "shortcut icon"):
            sizes = attr(tag, "sizes") or "32x32"
            try:
                s = int(sizes.split("x")[0])
            except ValueError:
                s = 32
            if not href.endswith(".ico"):
                icons.append((s, href))
    if icons:
        icons.sort(reverse=True)
        logo = urljoin(base_url, icons[0][1])
    for tag in META_RE.findall(head):
        prop = (attr(tag, "property") or attr(tag, "name") or "").lower()
        content = attr(tag, "content")
        if not content:
            continue
        if prop == "og:image" and not cover:
            cover = urljoin(base_url, content)
        elif prop in ("description", "og:description") and not desc:
            desc = content.strip()[:400]
        elif prop == "og:site_name" and not site_name:
            site_name = content.strip()[:80]
        elif prop == "og:logo" and not logo:
            logo = urljoin(base_url, content)
    return {"logo_url": logo, "cover_url": cover, "description_meta": desc, "site_name": site_name}


def enrich_domain(store):
    dom = store["id"]
    cpath = os.path.join(CACHE, dom + ".json")
    if os.path.exists(cpath):
        return json.load(open(cpath))
    url = store.get("url") or f"https://{dom}"
    html = ""
    try:
        r = requests.get(url, headers=HDRS, timeout=20, allow_redirects=True)
        if r.status_code == 200:
            html, url = r.text, r.url
    except Exception:
        pass
    if not html and available():
        try:
            status, content = scrapfly_get(url)
            if status == 200:
                html = content
        except Exception:
            pass
    rec = {"domain": dom}
    if html:
        rec.update(extract(url, html))
        # fallback logo : favicon.ico si rien trouvé
        if not rec.get("logo_url"):
            rec["logo_url"] = urljoin(url, "/favicon.ico")
    json.dump(rec, open(cpath, "w"))
    return rec


def main(only_live=True):
    con = fdb.connect()
    # colonnes d'enrichissement
    for col in ("logo_url", "cover_url", "description_meta"):
        try:
            con.execute(f"ALTER TABLE stores ADD COLUMN {col} TEXT")
        except Exception:
            pass
    con.commit()
    rows = [dict(r) for r in con.execute(
        "SELECT id, url, product_count FROM stores ORDER BY product_count DESC")]
    con.close()
    if only_live:
        pass  # on enrichit tout le registre, les vivantes d'abord (tri ci-dessus)
    print(f"enrichissement de {len(rows)} boutiques…")
    done = 0
    results = []
    with cf.ThreadPoolExecutor(16) as ex:
        for rec in ex.map(enrich_domain, rows):
            results.append(rec)
            done += 1
            if done % 200 == 0:
                print(f"  {done}/{len(rows)}", flush=True)
    con = fdb.connect()
    n = 0
    for rec in results:
        if rec.get("logo_url") or rec.get("description_meta"):
            con.execute("UPDATE stores SET logo_url=?, cover_url=?, description_meta=? WHERE id=?",
                        (rec.get("logo_url"), rec.get("cover_url"),
                         rec.get("description_meta"), rec["domain"]))
            n += 1
    con.commit(); con.close()
    print(f"boutiques enrichies: {n}/{len(results)}")


if __name__ == "__main__":
    main()