fix shorts_carousel.py ONLY: relax URL filter + 3:4 thumbnail ratio

shorts_carousel.py

@@ -1,4 +1,4 @@
-"""Self-contained 24h Shorts Carousel + multi-part video extraction."""
+"""24h Shorts Carousel + video extraction."""
 import re
 import requests
 from bs4 import BeautifulSoup
@@ -12,48 +12,46 @@ BASE_24H = "https://www.24h.com.vn"
 
 
 def scrape_24h_news_shorts():
-    """Scrape https://www.24h.com.vn/video/video-tin-tuc-cvd769.html"""
+    """Scrape https://www.24h.com.vn/video/video-tin-tuc-cvd769.html
+    Lấy tất cả link bài viết có ảnh thumbnail."""
     try:
         url = f"{BASE_24H}/video/video-tin-tuc-cvd769.html"
-        r = requests.get(url, headers=HEADERS, timeout=15)
+        r = requests.get(url, headers=HEADERS, timeout=12)
         r.encoding = "utf-8"
         soup = BeautifulSoup(r.text, "lxml")
         articles, seen = [], set()
         for a in soup.find_all("a", href=True):
             href = a.get("href", "")
-            if not re.search(r'-c\d+a\d+\.html$', href):
+            # Bỏ qua link javascript, anchor, category
+            if not href.endswith(".html"):
+                continue
+            if "javascript:" in href or href == "#":
+                continue
+            if "-cvd" in href or "-c953" in href:
                 continue
             if not href.startswith("http"):
                 href = BASE_24H + href
             if href in seen:
                 continue
-            seen.add(href)
-            title = a.get("title", "") or ""
+            # Phải có ảnh
             img = a.find("img")
-            if img and not title:
-                title = img.get("alt", "")
-            if not title:
-                title = a.get_text(strip=True)
-            if not title or len(title) < 10:
+            if not img:
+                p = a.parent
+                if p:
+                    img = p.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
-            img_src = None
-            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:
+            # Lấy title
+            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", "is_video": True
+                "source": "24h-shorts", "is_video": True
             })
         return articles[:20]
     except:
@@ -61,80 +59,56 @@ def scrape_24h_news_shorts():
 
 
 def _extract_24h_video_url(article_url):
-    """Extract m3u8 video URL(s) from a 24h article.
-    Probes for multi-part videos using CDN naming patterns."""
+    """Extract m3u8 video URL(s) from a 24h article."""
     try:
         r = requests.get(article_url, headers=HEADERS, timeout=10)
         r.encoding = "utf-8"
         html = r.text
-
         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]
-
         all_m3u8 = list(dict.fromkeys(raw_m3u8 + escaped_m3u8))
         full_quality = [u for u in all_m3u8 if '_720p' not in u]
         has_720p = [u for u in all_m3u8 if '_720p' in u]
-
         if not full_quality and not has_720p:
             return None
-
         primary_list = full_quality if full_quality else has_720p
         base_url = primary_list[0]
-
         soup = BeautifulSoup(html, "lxml")
         og = soup.find("meta", property="og:image")
         poster = og.get("content", "") if og else ""
-
         if len(primary_list) > 1:
             return {"src": primary_list[0], "poster": poster, "vtype": "hls", "all_parts": primary_list}
-
+        # Probe multi-part
         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', '')
-
+        name_part = base_url.replace('_720p.m3u8', '').replace('.m3u8', '')
         m = re.search(r'(\D)(\d{1,2})$', name_part)
         if m:
-            current_num = int(m.group(2))
-            num_width = len(m.group(2))
+            cur = int(m.group(2))
+            w = len(m.group(2))
             base_name = name_part[:m.start(2)]
-            for i in range(current_num + 1, current_num + 5):
-                next_name = f"{base_name}{i:02d}" if num_width == 2 else f"{base_name}{i}"
-                for suffix in ['_720p.m3u8' if is_720p else '.m3u8', '_720p.m3u8', '.m3u8']:
-                    probe_url = next_name + suffix
-                    if probe_url in parts_found:
-                        continue
+            for i in range(cur+1, cur+5):
+                nn = f"{base_name}{i:02d}" if w==2 else f"{base_name}{i}"
+                for sfx in [('_720p.m3u8' if is_720p else '.m3u8'), '_720p.m3u8', '.m3u8']:
+                    pu = nn+sfx
+                    if pu in parts_found: continue
                     try:
-                        pr = requests.head(probe_url, headers=HEADERS, timeout=3, allow_redirects=True)
-                        if pr.status_code == 200:
-                            parts_found.append(probe_url)
-                            break
-                    except:
-                        continue
-                else:
-                    break
-
-        if len(parts_found) == 1:
+                        if requests.head(pu, headers=HEADERS, timeout=3, allow_redirects=True).status_code==200:
+                            parts_found.append(pu); break
+                    except: continue
+                else: break
+        if len(parts_found)==1:
             m2 = re.match(r'^(.+-)(\d{7,})(-.+?)(\d{1,2})((?:_720p)?\.m3u8)$', base_url)
             if m2:
-                prefix = m2.group(1)
-                cdn_id = int(m2.group(2))
-                mid_part = m2.group(3)
-                part_num = int(m2.group(4))
-                suffix = m2.group(5)
-                for i in range(1, 5):
-                    probe_url = f"{prefix}{cdn_id + i}{mid_part}{part_num + i}{suffix}"
+                pfx,cid,mid,pn,sfx = m2.group(1),int(m2.group(2)),m2.group(3),int(m2.group(4)),m2.group(5)
+                for i in range(1,5):
+                    pu = f"{pfx}{cid+i}{mid}{pn+i}{sfx}"
                     try:
-                        pr = requests.head(probe_url, headers=HEADERS, timeout=3, allow_redirects=True)
-                        if pr.status_code == 200:
-                            parts_found.append(probe_url)
-                        else:
-                            break
-                    except:
-                        break
-
+                        if requests.head(pu, headers=HEADERS, timeout=3, allow_redirects=True).status_code==200:
+                            parts_found.append(pu)
+                        else: break
+                    except: break
         if len(parts_found) > 1:
             return {"src": parts_found[0], "poster": poster, "vtype": "hls", "all_parts": parts_found}
         return {"src": base_url, "poster": poster, "vtype": "hls"}
@@ -143,18 +117,18 @@ def _extract_24h_video_url(article_url):
 
 
 SHORTS_CSS = """
-/* Shorts 9:16 Slider */
-.vslide-shorts-item{flex:0 0 110px;scroll-snap-align:start;cursor:pointer;transition:transform .15s}
+/* Shorts 3:4 Slider */
+.vslide-shorts-item{flex:0 0 120px;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 130px}}
-.vslide-shorts-thumb{position:relative;width:100%;aspect-ratio:9/16;border-radius:12px;overflow:hidden;background:#222}
+@media(min-width:768px){.vslide-shorts-item{flex:0 0 140px}}
+.vslide-shorts-thumb{position:relative;width:100%;aspect-ratio:3/4;border-radius:10px;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 9:16 vertical shorts carousel HTML."""
+    """Render 3:4 shorts carousel."""
     vids = [v for v in shorts if v.get("img")]
     if not vids:
         return ""
@@ -173,7 +147,7 @@ def render_shorts_carousel(shorts, esc_fn, safe_url_fn, make_id_fn, slug_fn):
             + '<p class="vslide-shorts-title">' + title + '</p></div>'
         )
     h = '<div class="vslide-wrap"><div class="vslide-header">'
-    h += '<span class="vslide-label">\U0001f4f1 Shorts Tin T\u1ee9c 24h</span>'
+    h += '<span class="vslide-label">\U0001f4f1 Shorts 24h</span>'
     h += "<div class=\"vslide-nav\"><button class=\"vslide-btn\" onclick=\"window.bdpSlideScroll(-1,'vslide-shorts')\">\u25c0</button>"
     h += "<button class=\"vslide-btn\" onclick=\"window.bdpSlideScroll(1,'vslide-shorts')\">\u25b6</button></div></div>"
     h += '<div class="vslide-track" id="vslide-shorts">' + ''.join(items) + '</div></div>'