codex2

pricewatch/app/stores/amazon/selectors.yml

@@ -54,12 +54,12 @@ specs_table:
 # ASIN (parfois dans les métadonnées)
 asin:
   - "input[name='ASIN']"
-  - "th:contains('ASIN') + td"
+  - "th:-soup-contains('ASIN') + td"
 
 # Messages captcha / robot check
 captcha_indicators:
   - "form[action*='validateCaptcha']"
-  - "p.a-last:contains('Sorry')"
+  - "p.a-last:-soup-contains('Sorry')"
   - "img[alt*='captcha']"
 
 # Notes pour le parsing:

pricewatch/app/stores/amazon/store.py

@@ -4,7 +4,9 @@ Store Amazon - Parsing de produits Amazon.fr et Amazon.com.
 Supporte l'extraction de: titre, prix, ASIN, images, specs, etc.
 """
 
+import json
 import re
+from html import unescape
 from datetime import datetime
 from pathlib import Path
 from typing import Optional
@@ -21,6 +23,7 @@ from pricewatch.app.core.schema import (
     StockStatus,
 )
 from pricewatch.app.stores.base import BaseStore
+from pricewatch.app.stores.price_parser import parse_price_text
 
 logger = get_logger("stores.amazon")
 
@@ -131,6 +134,8 @@ class AmazonStore(BaseStore):
         images = self._extract_images(soup, debug_info)
         category = self._extract_category(soup, debug_info)
         specs = self._extract_specs(soup, debug_info)
+        description = self._extract_description(soup, debug_info)
+        msrp = self._extract_msrp(soup, debug_info)
         reference = self.extract_reference(url) or self._extract_asin_from_html(soup)
 
         # Déterminer le statut final (ne pas écraser FAILED)
@@ -150,8 +155,10 @@ class AmazonStore(BaseStore):
             stock_status=stock_status,
             reference=reference,
             category=category,
+            description=description,
             images=images,
             specs=specs,
+            msrp=msrp,
             debug=debug_info,
         )
 
@@ -195,6 +202,17 @@ class AmazonStore(BaseStore):
         debug.errors.append("Titre non trouvé")
         return None
 
+    def _extract_description(self, soup: BeautifulSoup, debug: DebugInfo) -> Optional[str]:
+        """Extrait la description (meta tags)."""
+        meta = soup.find("meta", property="og:description") or soup.find(
+            "meta", attrs={"name": "description"}
+        )
+        if meta:
+            description = meta.get("content", "").strip()
+            if description:
+                return description
+        return None
+
     def _extract_price(self, soup: BeautifulSoup, debug: DebugInfo) -> Optional[float]:
         """Extrait le prix."""
         selectors = self.get_selector("price", [])
@@ -205,14 +223,9 @@ class AmazonStore(BaseStore):
             elements = soup.select(selector)
             for element in elements:
                 text = element.get_text(strip=True)
-                # Extraire nombre (format: "299,99" ou "299.99")
-                match = re.search(r"(\d+)[.,](\d+)", text)
-                if match:
-                    price_str = f"{match.group(1)}.{match.group(2)}"
-                    try:
-                        return float(price_str)
-                    except ValueError:
-                        continue
+                price = parse_price_text(text)
+                if price is not None:
+                    return price
 
         # Fallback: chercher les spans séparés a-price-whole et a-price-fraction
         whole = soup.select_one("span.a-price-whole")
@@ -220,15 +233,24 @@ class AmazonStore(BaseStore):
         if whole and fraction:
             whole_text = whole.get_text(strip=True)
             fraction_text = fraction.get_text(strip=True)
-            try:
-                price_str = f"{whole_text}.{fraction_text}"
-                return float(price_str)
-            except ValueError:
-                pass
+            price = parse_price_text(f"{whole_text}.{fraction_text}")
+            if price is not None:
+                return price
 
         debug.errors.append("Prix non trouvé")
         return None
 
+    def _extract_msrp(self, soup: BeautifulSoup, debug: DebugInfo) -> Optional[float]:
+        """Extrait le prix conseille."""
+        strike = soup.select_one("span.priceBlockStrikePriceString") or soup.select_one(
+            "span.a-text-price span.a-offscreen"
+        )
+        if strike:
+            price = parse_price_text(strike.get_text(strip=True))
+            if price is not None:
+                return price
+        return None
+
     def _extract_currency(self, soup: BeautifulSoup, debug: DebugInfo) -> Optional[str]:
         """Extrait la devise."""
         selectors = self.get_selector("currency", [])
@@ -270,6 +292,7 @@ class AmazonStore(BaseStore):
     def _extract_images(self, soup: BeautifulSoup, debug: DebugInfo) -> list[str]:
         """Extrait les URLs d'images."""
         images = []
+        seen = set()
         selectors = self.get_selector("images", [])
         if isinstance(selectors, str):
             selectors = [selectors]
@@ -278,19 +301,57 @@ class AmazonStore(BaseStore):
             elements = soup.select(selector)
             for element in elements:
                 # Attribut src ou data-src
-                url = element.get("src") or element.get("data-src")
+                url = element.get("src") or element.get("data-src") or element.get("data-old-hires")
                 if url and url.startswith("http"):
-                    images.append(url)
+                    if self._is_product_image(url) and url not in seen:
+                        images.append(url)
+                        seen.add(url)
+                dynamic = element.get("data-a-dynamic-image")
+                if dynamic:
+                    urls = self._extract_dynamic_images(dynamic)
+                    for dyn_url in urls:
+                        if self._is_product_image(dyn_url) and dyn_url not in seen:
+                            images.append(dyn_url)
+                            seen.add(dyn_url)
 
         # Fallback: chercher tous les img tags si aucune image trouvée
         if not images:
             all_imgs = soup.find_all("img")
             for img in all_imgs:
                 url = img.get("src") or img.get("data-src")
-                if url and url.startswith("http"):
-                    images.append(url)
+                if url and url.startswith("http") and self._is_product_image(url):
+                    if url not in seen:
+                        images.append(url)
+                        seen.add(url)
 
-        return list(set(images))  # Dédupliquer
+        return images
+
+    def _extract_dynamic_images(self, raw: str) -> list[str]:
+        """Extrait les URLs du JSON data-a-dynamic-image."""
+        try:
+            data = json.loads(unescape(raw))
+        except (TypeError, json.JSONDecodeError):
+            return []
+
+        urls = []
+        if isinstance(data, dict):
+            candidates = []
+            for url, dims in data.items():
+                if not isinstance(url, str) or not url.startswith("http"):
+                    continue
+                size = dims[0] if isinstance(dims, list) and dims else 0
+                candidates.append((size, url))
+            candidates.sort(key=lambda item: item[0], reverse=True)
+            for _, url in candidates:
+                urls.append(url)
+        return urls
+
+    def _is_product_image(self, url: str) -> bool:
+        """Filtre basique pour eviter les logos et sprites."""
+        lowered = url.lower()
+        if "prime_logo" in lowered or "sprite" in lowered:
+            return False
+        return True
 
     def _extract_category(self, soup: BeautifulSoup, debug: DebugInfo) -> Optional[str]:
         """Extrait la catégorie depuis les breadcrumbs."""