From b47af1655da1eae4bd1392288363c22e5b44685b Mon Sep 17 00:00:00 2001
From: Michael SCHAL <michaelschal@gmail.com>
Date: Sat, 29 Nov 2025 17:46:20 +0100
Subject: [PATCH] feat: add debug script to display search site configurations
 and database entries

---
 analyze_missing_sites.py  | 93 +++++++++++++++++++++++++++++++++++++++
 app/core/search_config.py | 25 +++++++++++
 debug_sites.py            | 30 +++++++++++++
 3 files changed, 148 insertions(+)
 create mode 100644 analyze_missing_sites.py
 create mode 100644 debug_sites.py

diff --git a/analyze_missing_sites.py b/analyze_missing_sites.py
new file mode 100644
index 0000000..fa7eca5
--- /dev/null
+++ b/analyze_missing_sites.py
@@ -0,0 +1,93 @@
+"""
+Script pour analyser les sélecteurs CSS des sites manquants
+"""
+import asyncio
+from playwright.async_api import async_playwright
+
+async def analyze_site(url: str, site_name: str):
+    print(f"\n=== Analyse de {site_name} ===")
+    print(f"URL: {url}")
+    
+    async with async_playwright() as p:
+        browser = await p.chromium.launch(headless=False)
+        page = await browser.new_page()
+        
+        try:
+            await page.goto(url, wait_until="networkidle", timeout=30000)
+            await page.wait_for_timeout(3000)
+            
+            # Try cookie banners
+            cookie_selectors = [
+                "button:has-text('Accepter')",
+                "button:has-text('Tout accepter')",
+                "#didomi-notice-agree-button",
+                ".didomi-continue-without-agreeing"
+            ]
+            for selector in cookie_selectors:
+                try:
+                    await page.click(selector, timeout=2000)
+                    print(f"✓ Cookie banner fermé: {selector}")
+                    break
+                except:
+                    pass
+            
+            await page.wait_for_timeout(2000)
+            
+            # Save HTML
+            html = await page.content()
+            filename = f"{site_name.lower().replace(' ', '_')}_search.html"
+            with open(filename, "w", encoding="utf-8") as f:
+                f.write(html)
+            print(f"✓ HTML sauvegardé: {filename}")
+            
+            # Take screenshot
+            screenshot_path = f"{site_name.lower().replace(' ', '_')}_search.png"
+            await page.screenshot(path=screenshot_path, full_page=True)
+            print(f"✓ Screenshot: {screenshot_path}")
+            
+            # Test common product link selectors
+            test_selectors = [
+                "a.product-link",
+                "a.product-name",
+                "a[href*='/produit']",
+                "a[href*='/product']",
+                "a[href*='/p/']",
+                ".product-title a",
+                ".product-item a",
+                "article a",
+                "a.product",
+                "a[itemprop='url']",
+            ]
+            
+            print("\n--- Test de sélecteurs ---")
+            for selector in test_selectors:
+                try:
+                    elements = await page.query_selector_all(selector)
+                    if elements:
+                        print(f"✓ {selector}: {len(elements)} éléments trouvés")
+                        # Get first few hrefs
+                        for i, elem in enumerate(elements[:3]):
+                            href = await elem.get_attribute("href")
+                            text = await elem.inner_text()
+                            print(f"  [{i+1}] {text[:50]} -> {href}")
+                except Exception as e:
+                    pass
+            
+        except Exception as e:
+            print(f"❌ Erreur: {e}")
+        finally:
+            await browser.close()
+
+async def main():
+    sites = [
+        ("https://bmstores.fr/module/ambjolisearch/jolisearch?s=chaise", "BM"),
+        ("https://www.centrakor.com/recherche?controller=search&s=chaise", "Centrakor"),
+        ("https://www.lincroyable.fr/recherche?query=chaise", "L'Incroyable"),
+    ]
+    
+    for url, name in sites:
+        await analyze_site(url, name)
+        await asyncio.sleep(2)
+
+if __name__ == "__main__":
+    asyncio.run(main())
diff --git a/app/core/search_config.py b/app/core/search_config.py
index 7fc6450..3be93a0 100644
--- a/app/core/search_config.py
+++ b/app/core/search_config.py
@@ -119,6 +119,31 @@ SITE_CONFIGS = {
         "category": "Culture & Tech",
         "requires_proxy": False,
     },
+    # === NOUVEAUX SITES ===
+    "bmstores.fr": {
+        "name": "B&M",
+        "search_url": "https://bmstores.fr/module/ambjolisearch/jolisearch?s={query}",
+        "product_selector": "a.thumbnail.product-thumbnail",
+        "wait_selector": ".products, .product-miniature",
+        "category": "Discount",
+        "requires_proxy": False,
+    },
+    "centrakor.com": {
+        "name": "Centrakor",
+        "search_url": "https://www.centrakor.com/search/{query}",
+        "product_selector": "a.link.link--block",
+        "wait_selector": ".product-list, .product-item",
+        "category": "Discount",
+        "requires_proxy": False,
+    },
+    "lincroyable.fr": {
+        "name": "L'Incroyable",
+        "search_url": "https://www.lincroyable.fr/recherche-query={query}/",
+        "product_selector": "a[href*='.html']",
+        "wait_selector": ".product-item, .product-list",
+        "category": "Discount",
+        "requires_proxy": False,
+    },
 }
 
 # === COOKIE BANNERS ===
diff --git a/debug_sites.py b/debug_sites.py
new file mode 100644
index 0000000..4b45bcc
--- /dev/null
+++ b/debug_sites.py
@@ -0,0 +1,30 @@
+import sys
+import os
+
+# Add app to path
+sys.path.append(os.getcwd())
+
+from app.core.database import SessionLocal
+from app.models import SearchSite
+from app.core.search_config import SITE_CONFIGS
+
+def debug_sites():
+    db = SessionLocal()
+    try:
+        print("=== SITE_CONFIGS Keys ===")
+        for key in SITE_CONFIGS.keys():
+            print(f"- {key}")
+            
+        print("\n=== DB SearchSites ===")
+        sites = db.query(SearchSite).all()
+        for site in sites:
+            print(f"ID: {site.id} | Name: {site.name} | Domain: {site.domain} | Active: {site.is_active}")
+            print(f"   URL: {site.search_url}")
+            print(f"   Selector: {site.product_link_selector}")
+            print("-" * 20)
+            
+    finally:
+        db.close()
+
+if __name__ == "__main__":
+    debug_sites()