Enhance HTMLExtractionOptions with new annotateNumberClasses feature

- Added `annotateNumberClasses` option to the HTMLExtractionOptions interface for appending CSS class names to numeric elements in markdown output.
- Updated README to include documentation for the new feature.
- Improved HTML to Markdown conversion logic to support number class annotation, preserving semantic meaning.
- Added unit tests to ensure correct functionality across various scenarios.

Author: azhong-git

a.html

@@ -0,0 +1,55 @@
+function scrape(document) {
+  const products = [];
+  const productLinks = document.querySelectorAll('div[data-testid="list-view"] a.w-100.h-100.z-1');
+
+  productLinks.forEach(link => {
+    const productTile = link.closest('div[data-testid="list-view"] > div > div');
+    if (!productTile) return;
+
+    const name = link.textContent.trim();
+    const productUrl = link.href;
+
+    const imageUrl = productTile.querySelector('img[data-testid="productTileImage"]')?.src;
+
+    let price;
+    const priceElement = productTile.querySelector('div.flex.flex-wrap.justify-start > div.mr1.mr2-xl.b');
+    if (priceElement) {
+      const priceText = priceElement.textContent.replace(/[^0-9.]/g, '');
+      price = parseFloat(priceText);
+    }
+
+    let originalPrice;
+    const originalPriceElement = productTile.querySelector('div.flex.flex-wrap.justify-start > div.gray.mr1.f7:nth-of-type(2)');
+    if (originalPriceElement) {
+      const originalPriceText = originalPriceElement.textContent.replace(/[^0-9.]/g, '');
+      originalPrice = parseFloat(originalPriceText);
+    }
+
+    const brand = productTile.querySelector('div.mb1.mt2.b')?.textContent.trim();
+
+    let rating;
+    let reviewCount;
+    const ratingReviewElement = productTile.querySelector('div.flex.items-center.mt2');
+    if (ratingReviewElement) {
+      const ratingReviewText = ratingReviewElement.textContent;
+      const match = ratingReviewText.match(/(\d+\.?\d*)\s+out of 5 stars\.\s+(\d+)\s+reviews/);
+      if (match) {
+        rating = parseFloat(match[1]);
+        reviewCount = parseInt(match[2], 10);
+      }
+    }
+
+    products.push({
+      name,
+      brand: brand || undefined,
+      price: price || undefined,
+      originalPrice: originalPrice || undefined,
+      rating: rating || undefined,
+      reviewCount: reviewCount || undefined,
+      productUrl,
+      imageUrl: imageUrl || undefined
+    });
+  });
+
+  return { products };
+}

a.out

@@ -0,0 +1,3 @@
+{
+  "products": []
+}

aa.html

@@ -0,0 +1,3 @@
+{
+  "products": []
+}

scrape-debug-1775576198343/annotated-markdown.md

@@ -0,0 +1,4 @@
+{
+  "isValid": false,
+  "issues": "The 'products' array is empty, but the annotated content clearly shows multiple product listings with details such as name, price, brand, rating, review count, product URL, and image URL. The scraping code needs to be fixed to extract all these product details and populate the 'products' array accordingly."
+}