revert: shorts_carousel.py back to previous (9:16 ratio, strict regex)

shorts_carousel.py

@@ -12,90 +12,49 @@ BASE_24H = "https://www.24h.com.vn"
 
 
 def scrape_24h_news_shorts():
-    """Scrape https://www.24h.com.vn/video/video-tin-tuc-cvd769.html
-    Accept any article link that ends with .html and has content."""
+    """Scrape https://www.24h.com.vn/video/video-tin-tuc-cvd769.html"""
     try:
         url = f"{BASE_24H}/video/video-tin-tuc-cvd769.html"
         r = requests.get(url, headers=HEADERS, timeout=15)
         r.encoding = "utf-8"
         soup = BeautifulSoup(r.text, "lxml")
         articles, seen = [], set()
-        
-        # Look at all <article> tags first (more reliable structure)
-        for art in soup.find_all("article"):
-            a = art.find("a", href=True)
-            if not a:
-                continue
+        for a in soup.find_all("a", href=True):
             href = a.get("href", "")
-            if not href or href == "#" or "javascript:" in href:
+            if not re.search(r'-c\d+a\d+\.html$', href):
                 continue
             if not href.startswith("http"):
                 href = BASE_24H + href
-            # Skip category/listing pages
-            if href.endswith("cvd769.html") or "-cvd" in href:
-                continue
             if href in seen:
                 continue
             seen.add(href)
-            
-            # Get title from img alt, a title, or text
-            title = ""
-            img_tag = art.find("img")
-            if img_tag:
-                title = img_tag.get("alt", "")
+            title = a.get("title", "") or ""
+            img = a.find("img")
+            if img and not title:
+                title = img.get("alt", "")
             if not title:
-                title = a.get("title", "") or a.get_text(strip=True)
-            if not title or len(title) < 8:
+                title = a.get_text(strip=True)
+            if not title or len(title) < 10:
                 continue
-            
-            # Get image
             img_src = None
-            if img_tag:
-                img_src = img_tag.get("data-original") or img_tag.get("data-src") or img_tag.get("src") or ""
-                if not img_src or "base64" in img_src or len(img_src) < 20:
+            if img:
+                img_src = img.get("data-original") or img.get("data-src") or img.get("src") or ""
+                if "base64" in img_src or len(img_src) < 20:
                     img_src = None
-            
+            if not img_src:
+                parent = a.parent
+                if parent:
+                    parent_img = parent.find("img")
+                    if parent_img:
+                        img_src = parent_img.get("data-original") or parent_img.get("data-src") or parent_img.get("src") or ""
+                        if "base64" in img_src or len(img_src) < 20:
+                            img_src = None
             if not img_src:
                 continue
-            
             articles.append({
                 "title": title, "link": href, "img": img_src,
-                "source": "24h-shorts", "is_video": True
+                "source": "24h", "is_video": True
             })
-        
-        # Fallback: if <article> approach got nothing, try all <a> with images
-        if not articles:
-            for a in soup.find_all("a", href=True):
-                href = a.get("href", "")
-                if not href or href == "#" or "javascript:" in href:
-                    continue
-                if not href.endswith(".html"):
-                    continue
-                if "-cvd" in href:
-                    continue
-                if not href.startswith("http"):
-                    href = BASE_24H + href
-                if href in seen:
-                    continue
-                
-                img = a.find("img")
-                if not img:
-                    continue
-                
-                img_src = img.get("data-original") or img.get("data-src") or img.get("src") or ""
-                if not img_src or "base64" in img_src or len(img_src) < 20:
-                    continue
-                
-                title = img.get("alt", "") or a.get("title", "") or a.get_text(strip=True)
-                if not title or len(title) < 8:
-                    continue
-                
-                seen.add(href)
-                articles.append({
-                    "title": title, "link": href, "img": img_src,
-                    "source": "24h-shorts", "is_video": True
-                })
-        
         return articles[:20]
     except:
         return []
@@ -109,7 +68,6 @@ def _extract_24h_video_url(article_url):
         r.encoding = "utf-8"
         html = r.text
 
-        # Find all m3u8 URLs (normal + escaped)
         raw_m3u8 = re.findall(r'(https?://cdn\.24h\.com\.vn/[^\s"\'\\<>]+\.m3u8)', html)
         escaped_m3u8 = re.findall(r'(https?:\\/\\/cdn\.24h\.com\.vn\\/[^\s"\'<>]+\.m3u8)', html)
         escaped_m3u8 = [u.replace('\\/', '/') for u in escaped_m3u8]
@@ -124,7 +82,6 @@ def _extract_24h_video_url(article_url):
         primary_list = full_quality if full_quality else has_720p
         base_url = primary_list[0]
 
-        # Get poster
         soup = BeautifulSoup(html, "lxml")
         og = soup.find("meta", property="og:image")
         poster = og.get("content", "") if og else ""
@@ -132,14 +89,12 @@ def _extract_24h_video_url(article_url):
         if len(primary_list) > 1:
             return {"src": primary_list[0], "poster": poster, "vtype": "hls", "all_parts": primary_list}
 
-        # Probe for additional parts
         parts_found = [base_url]
 
         is_720p = '_720p.m3u8' in base_url
         clean_url = base_url.replace('_720p.m3u8', '.m3u8') if is_720p else base_url
         name_part = clean_url.replace('.m3u8', '')
 
-        # Pattern A: name ends with digit(s) → increment
         m = re.search(r'(\D)(\d{1,2})$', name_part)
         if m:
             current_num = int(m.group(2))
@@ -161,7 +116,6 @@ def _extract_24h_video_url(article_url):
                 else:
                     break
 
-        # Pattern B: CDN article ID increments + part number increments
         if len(parts_found) == 1:
             m2 = re.match(r'^(.+-)(\d{7,})(-.+?)(\d{1,2})((?:_720p)?\.m3u8)$', base_url)
             if m2:
@@ -188,20 +142,19 @@ def _extract_24h_video_url(article_url):
         return None
 
 
-# CSS: shorts slide uses 3:4 ratio thumbnail
 SHORTS_CSS = """
-/* Shorts 3:4 Slider */
-.vslide-shorts-item{flex:0 0 130px;scroll-snap-align:start;cursor:pointer;transition:transform .15s}
+/* Shorts 9:16 Slider */
+.vslide-shorts-item{flex:0 0 110px;scroll-snap-align:start;cursor:pointer;transition:transform .15s}
 .vslide-shorts-item:hover{transform:scale(1.04)}
-@media(min-width:768px){.vslide-shorts-item{flex:0 0 150px}}
-.vslide-shorts-thumb{position:relative;width:100%;aspect-ratio:3/4;border-radius:12px;overflow:hidden;background:#222}
+@media(min-width:768px){.vslide-shorts-item{flex:0 0 130px}}
+.vslide-shorts-thumb{position:relative;width:100%;aspect-ratio:9/16;border-radius:12px;overflow:hidden;background:#222}
 .vslide-shorts-thumb img{width:100%;height:100%;object-fit:cover}
 .vslide-shorts-title{color:#ccc;font-size:10.5px;margin:5px 0 0;line-height:1.3;display:-webkit-box;-webkit-line-clamp:2;-webkit-box-orient:vertical;overflow:hidden}
 """
 
 
 def render_shorts_carousel(shorts, esc_fn, safe_url_fn, make_id_fn, slug_fn):
-    """Render 3:4 shorts carousel HTML."""
+    """Render 9:16 vertical shorts carousel HTML."""
     vids = [v for v in shorts if v.get("img")]
     if not vids:
         return ""