From c23cc8353a7e9498635790009d5d932c2e373d23 Mon Sep 17 00:00:00 2001
From: clucraft <clucraft@users.noreply.github.com>
Date: Wed, 21 Jan 2026 21:17:14 -0500
Subject: [PATCH] Remove B&H Photo scraper (Cloudflare protection too strong)

B&H Photo Video uses aggressive Cloudflare protection that blocks
headless browsers even with stealth plugins. Removing the site-specific
scraper for now. The Puppeteer fallback remains in place for other
sites with less aggressive protection.

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
---
 backend/src/services/scraper.ts | 128 --------------------------------
 1 file changed, 128 deletions(-)

diff --git a/backend/src/services/scraper.ts b/backend/src/services/scraper.ts
index d5c8a33..5cefef6 100644
--- a/backend/src/services/scraper.ts
+++ b/backend/src/services/scraper.ts
@@ -628,134 +628,6 @@ const siteScrapers: SiteScraper[] = [
     },
   },
 
-  // B&H Photo Video
-  {
-    match: (url) => /bhphotovideo\.com/i.test(url),
-    scrape: ($) => {
-      let price: ParsedPrice | null = null;
-      let name: string | null = null;
-      let imageUrl: string | null = null;
-      let stockStatus: StockStatus = 'unknown';
-
-      // Debug: Check page title and body length
-      const pageTitle = $('title').text();
-      const bodyLength = $('body').html()?.length || 0;
-      console.log(`[B&H] Page title: "${pageTitle}", body length: ${bodyLength}`);
-
-      // Try to get data from JSON-LD first
-      try {
-        const scripts = $('script[type="application/ld+json"]');
-        console.log(`[B&H] Found ${scripts.length} JSON-LD scripts`);
-        scripts.each((_i, script) => {
-          const content = $(script).html();
-          if (!content) return;
-          try {
-            const data = JSON.parse(content);
-            console.log(`[B&H] JSON-LD type: ${data['@type']}`);
-            if (data['@type'] === 'Product' || data.offers) {
-              if (data.name && !name) {
-                name = data.name;
-                console.log(`[B&H] Found name: ${name}`);
-              }
-              if (data.image && !imageUrl) {
-                imageUrl = Array.isArray(data.image) ? data.image[0] : data.image;
-              }
-              if (data.offers && !price) {
-                const offer = Array.isArray(data.offers) ? data.offers[0] : data.offers;
-                console.log(`[B&H] Offer data: ${JSON.stringify(offer).slice(0, 200)}`);
-                if (offer.price) {
-                  price = {
-                    price: parseFloat(String(offer.price)),
-                    currency: offer.priceCurrency || 'USD',
-                  };
-                  console.log(`[B&H] Found price from JSON-LD: ${price.price}`);
-                }
-                // Check availability from JSON-LD
-                if (offer.availability) {
-                  const avail = offer.availability.toLowerCase();
-                  if (avail.includes('instock')) {
-                    stockStatus = 'in_stock';
-                  } else if (avail.includes('outofstock')) {
-                    stockStatus = 'out_of_stock';
-                  }
-                }
-              }
-            }
-          } catch (_e) {
-            // JSON-LD parse error, continue
-          }
-        });
-      } catch (_e) {
-        // JSON-LD extraction error, continue
-      }
-
-      // Fallback to HTML selectors
-      if (!price) {
-        console.log(`[B&H] No price from JSON-LD, trying HTML selectors`);
-        const priceSelectors = [
-          '[data-selenium="pricingPrice"]',
-          '[data-selenium="uppedDecimalPriceFirst"]',
-          '.price_1DPoToKrLP1U',
-          '[class*="price_"] span',
-          '.priceInfo span[class*="price"]',
-        ];
-
-        for (const selector of priceSelectors) {
-          const el = $(selector).first();
-          console.log(`[B&H] Selector "${selector}": found ${el.length} elements`);
-          if (el.length) {
-            const text = el.text().trim();
-            console.log(`[B&H] Element text: "${text.slice(0, 100)}"`);
-            price = parsePrice(text);
-            if (price) {
-              console.log(`[B&H] Parsed price: ${price.price}`);
-              break;
-            }
-          }
-        }
-      }
-
-      // Try combining dollars and cents if still no price
-      if (!price) {
-        const priceContainer = $('[data-selenium="pricingPrice"]').first();
-        if (priceContainer.length) {
-          const fullText = priceContainer.text().replace(/\s+/g, '');
-          price = parsePrice(fullText);
-        }
-      }
-
-      if (!name) {
-        name = $('h1[data-selenium="productTitle"]').text().trim() ||
-               $('h1[class*="title_"]').text().trim() ||
-               $('[data-selenium="productTitle"]').text().trim() ||
-               null;
-      }
-
-      if (!imageUrl) {
-        imageUrl = $('[data-selenium="mainImage"] img').attr('src') ||
-                   $('img[data-selenium="mainImage"]').attr('src') ||
-                   $('meta[property="og:image"]').attr('content') ||
-                   null;
-      }
-
-      // Stock status from HTML
-      if (stockStatus === 'unknown') {
-        const addToCartBtn = $('[data-selenium="addToCartButton"]').length > 0;
-        const notifyBtn = $('[data-selenium="notifyAvailabilityButton"]').length > 0;
-        const outOfStockText = $('body').text().toLowerCase();
-
-        if (addToCartBtn) {
-          stockStatus = 'in_stock';
-        } else if (notifyBtn || outOfStockText.includes('notify when available') ||
-                   outOfStockText.includes('temporarily unavailable')) {
-          stockStatus = 'out_of_stock';
-        }
-      }
-
-      console.log(`[B&H] Final result - name: ${name?.slice(0, 50)}, price: ${price?.price}, stock: ${stockStatus}`);
-      return { name, price, imageUrl, stockStatus };
-    },
-  },
 ];
 
 // Generic selectors as fallback