Commit
[ka6] fix connecteur donjjewellery.com: refonte WordPress/Woo → Next.js headless (backend Medusa) — le site a abandonné WooCommerce (Store API /wp-json/wc/store/v1/products en 404 Next.js « __next_error__ », wp-content ne survit que pour les médias sur media.donjjewellery.com; backend Medusa confirmé sur admin.donjjewellery.com /health 200 mais Store API verrouillée par x-publishable-api-key jamais exposée au client, tout est rendu serveur RSC). Le site est VIVANT et vend (prix affichés, og:price:amount présent sur chaque page produit, vérifié $1,585 = regular_price 1585). Fix en 2 volets: (1) fiche stores.json migrée woocommerce → generic (__generic__) avec sitemap_roots sur les 3 sitemaps produits (/sitemap/engagement-rings 1096 URLs, /sitemap/mens-products 191, /sitemap/womens-products 207); (2) generic.py: nouvelle clé fiche OPT-IN product_url_prefixes (miroir de skip_url_prefixes, jamais globale) car les URLs produit (/engagement-rings/…, /womens/…, /mens/…, slugs -variant_<ULID>) ne matchent PAS PRODUCT_URL_RE — sans elle, 0 URL retenue. L'extraction standard (og:title + og:price:amount + og:image + meta description) suffit, aucun fallback RSC nécessaire. Sync réel 2026-10-19: donjjewellery.com ok / 800 produits (+669, cap max_products; médiane 434 = l'ancien compte Woo, le nouveau site expose ~1494 URLs variantes). Non-régression generic nominal: atelierora.ca ok / 15 (0 erreur). pm2 restart fabri-ka-sync fait, web local 200. Si re-broken: vérifier que la fiche est restée generic + sitemap_roots/product_url_prefixes, et grep product_url_prefixes fabrika/connectors/generic.py (gare aux reverts concurrents).
modified
data/stores.json
+12 −2
@@ -42295,8 +42295,18 @@ | |||
| 42295 | "id": "donjjewellery.com", | 42295 | "id": "donjjewellery.com", |
| 42296 | "name": "Donj Jewellery", | 42296 | "name": "Donj Jewellery", |
| 42297 | "url": "https://donjjewellery.com", | 42297 | "url": "https://donjjewellery.com", |
| 42298 | − "platform": "woocommerce", | 42298 | + "platform": "generic", |
| 42299 | − "catalog_endpoint": "/wp-json/wc/store/v1/products", | 42299 | + "catalog_endpoint": "__generic__", |
| 42300 | + "sitemap_roots": [ | ||
| 42301 | + "/sitemap/engagement-rings", | ||
| 42302 | + "/sitemap/mens-products", | ||
| 42303 | + "/sitemap/womens-products" | ||
| 42304 | + ], | ||
| 42305 | + "product_url_prefixes": [ | ||
| 42306 | + "/engagement-rings/", | ||
| 42307 | + "/womens/", | ||
| 42308 | + "/mens/" | ||
| 42309 | + ], | ||
| 42300 | "city": "", | 42310 | "city": "", |
| 42301 | "region": "Laval", | 42311 | "region": "Laval", |
| 42302 | "postal_prefix": "H7V", | 42312 | "postal_prefix": "H7V", |
modified
fabrika/connectors/generic.py
+9 −1
@@ -203,6 +203,12 @@ class GenericConnector(BaseConnector): | |||
| 203 | extra = [u if u.startswith("http") else f"{self.base}{u}" | 203 | extra = [u if u.startswith("http") else f"{self.base}{u}" |
| 204 | for u in (self.store.get("sitemap_roots") or [])] | 204 | for u in (self.store.get("sitemap_roots") or [])] |
| 205 | roots = extra + roots | 205 | roots = extra + roots |
| 206 | + # URLs produit sous des préfixes propres au site, invisibles pour | ||
| 207 | + # PRODUCT_URL_RE (ex. donjjewellery.com, refonte Woo → Medusa headless | ||
| 208 | + # 2026-10 : produits sous /engagement-rings/…, /womens/…, /mens/…). | ||
| 209 | + # Inclusion OPT-IN par fiche registre (product_url_prefixes), jamais | ||
| 210 | + # globale — miroir de skip_url_prefixes. | ||
| 211 | + incl = tuple(self.store.get("product_url_prefixes") or []) | ||
| 206 | queue, depth_left = list(roots), 3 | 212 | queue, depth_left = list(roots), 3 |
| 207 | fetched_roots = 0 | 213 | fetched_roots = 0 |
| 208 | while queue and fetched_roots < 60: | 214 | while queue and fetched_roots < 60: |
@@ -231,7 +237,9 @@ class GenericConnector(BaseConnector): | |||
| 231 | for l in locs: | 237 | for l in locs: |
| 232 | if l.endswith(".xml"): | 238 | if l.endswith(".xml"): |
| 233 | continue | 239 | continue |
| 234 | − if PRODUCT_URL_RE.search(l) and not NON_PRODUCT_RE.search(l): | 240 | + if (PRODUCT_URL_RE.search(l) |
| 241 | + or (incl and urlparse(l).path.startswith(incl))) \ | ||
| 242 | + and not NON_PRODUCT_RE.search(l): | ||
| 235 | out.append(l) | 243 | out.append(l) |
| 236 | if len(out) >= self.max_products * 2: | 244 | if len(out) >= self.max_products * 2: |
| 237 | break | 245 | break |
| 238 | 246 | ||