Fix: rotating User-Agents, timeout 20s, Google News RSS fallback, source=google param, fix _UNIQUE_CHARS corruption, fix f-string syntax

app_v2_entry.py

@@ -23,7 +23,7 @@ from concurrent.futures import ThreadPoolExecutor, as_completed
 from urllib.parse import quote
 import asyncio
 
-HL_LEAGUES['friendly'] = {"path": "giai-khac/friendly", "name": "Giao hữu", "emoji": "🤝"}
+HL_LEAGUES['friendly'] = {"path": "giai-khac/friendly", "name": "Giao huu", "emoji": "🤝"}
 
 STATIC_DIR = os.path.join(os.path.dirname(os.path.abspath(__file__)), 'static')
 SPACE = "https://bep40-vnews.hf.space"
@@ -33,7 +33,6 @@ app.router.routes=[r for r in app.router.routes if not isinstance(r, Mount)]
 
 def _clean(s): return re.sub(r"\s+"," ",html_lib.unescape(str(s or""))).strip()
 
-# Cache for match details (5 min TTL)
 _match_cache = {}
 
 def _get_match_detail(event_id, slug=None):
@@ -203,9 +202,9 @@ def api_match_detail(event_id: int, url: str = Query(default=None)):
         return JSONResponse(err)
     return JSONResponse({"event_id": event_id, "found": False})
 
-_STOP=set('và của các những một được trong với cho tại sau trước khi không người việt nam hôm nay mới nhất nóng tin tức cập nhật theo từ đến là có thì này đã để'.split())
+_STOP=set('va cua cac nhung mot duoc trong voi cho tai sau truoc khi khong nguoi viet nam hom nay moi nhat nong tin tuc cap nhat theo tu den la co thi nay da de'.split())
 
-# Rotating User-Agents for search scrapers
+# Rotating User-Agents
 _SEARCH_UAS = [
     'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36',
     'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/123.0.0.0 Safari/537.36',
@@ -221,117 +220,157 @@ def _search_headers():
     return {'User-Agent': _search_rand.choice(_SEARCH_UAS), 'Accept-Language': 'vi-VN,vi;q=0.9,en;q=0.8', 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8'}
 
 def _has_kw(topic,title):
-    tl=topic.lower();tt=(title or'').lower()
-    if tl in tt:return True
-    words=[w for w in re.findall(r'[A-Za-zÀ-ỹ0-9]+',tl) if len(w)>2 and w not in _STOP]
-    if not words:return True
+    tl=topic.lower()
+    tt=(title or'').lower()
+    if tl in tt:
+        return True
+    words=[w for w in re.findall(r'[A-Za-z0-9]+',tl) if len(w)>2 and w not in _STOP]
+    if not words:
+        return True
     return any(w in tt for w in words)
 
 def _s_vnexpress(topic,limit=8):
     items=[]
     try:
-        r=req.get(f"https://timkiem.vnexpress.net/?q={quote(topic)}",headers=_search_headers(),timeout=12);soup=BeautifulSoup(r.text,'lxml')
+        r=req.get(f"https://timkiem.vnexpress.net/?q={quote(topic)}",headers=_search_headers(),timeout=12)
+        soup=BeautifulSoup(r.text,'lxml')
         for art in soup.select('article.item-news')[:limit]:
             a=art.select_one('h2 a, h3 a')
             if a and a.get('href'):
                 t=_clean(a.get('title','') or a.get_text(strip=True))
-                if _has_kw(topic,t):items.append({'title':t,'url':a['href'],'via':'VnExpress'})
-    except:pass
+                if _has_kw(topic,t):
+                    items.append({'title':t,'url':a['href'],'via':'VnExpress'})
+    except:
+        pass
     return items
 
 def _s_dantri(topic,limit=8):
     items=[]
     try:
-        r=req.get(f"https://dantri.com.vn/tim-kiem/{quote(topic)}.htm",headers=_search_headers(),timeout=12);soup=BeautifulSoup(r.text,'lxml')
+        r=req.get(f"https://dantri.com.vn/tim-kiem/{quote(topic)}.htm",headers=_search_headers(),timeout=12)
+        soup=BeautifulSoup(r.text,'lxml')
         for a in soup.select('h3 a[href], .article-title a[href]')[:limit*2]:
-            t=_clean(a.get_text(strip=True));href=a.get('href','')
+            t=_clean(a.get_text(strip=True))
+            href=a.get('href','')
             if t and len(t)>15 and _has_kw(topic,t):
-                if not href.startswith('http'):href='https://dantri.com.vn'+href
-                items.append({'title':t,'url':href,'via':'Dân Trí'})
-            if len(items)>=limit:break
-    except:pass
+                if not href.startswith('http'):
+                    href='https://dantri.com.vn'+href
+                items.append({'title':t,'url':href,'via':'Dan Tri'})
+            if len(items)>=limit:
+                break
+    except:
+        pass
     return items
 
 def _s_vietnamnet(topic,limit=6):
     items=[]
     try:
-        r=req.get(f"https://vietnamnet.vn/tim-kiem?q={quote(topic)}",headers=_search_headers(),timeout=12);soup=BeautifulSoup(r.text,'lxml')
+        r=req.get(f"https://vietnamnet.vn/tim-kiem?q={quote(topic)}",headers=_search_headers(),timeout=12)
+        soup=BeautifulSoup(r.text,'lxml')
         for a in soup.select('h3 a[href], .vnn-title a')[:limit*2]:
-            t=_clean(a.get('title','') or a.get_text(strip=True));href=a.get('href','')
+            t=_clean(a.get('title','') or a.get_text(strip=True))
+            href=a.get('href','')
             if t and len(t)>15 and _has_kw(topic,t):
-                if not href.startswith('http'):href='https://vietnamnet.vn'+href
+                if not href.startswith('http'):
+                    href='https://vietnamnet.vn'+href
                 items.append({'title':t,'url':href,'via':'VietNamNet'})
-            if len(items)>=limit:break
-    except:pass
+            if len(items)>=limit:
+                break
+    except:
+        pass
     return items
 
 def _s_bongda(topic,limit=5):
     items=[]
     try:
-        r=req.get(f"https://bongda.com.vn/tim-kiem.html?q={quote(topic)}",headers=_search_headers(),timeout=12);soup=BeautifulSoup(r.text,'lxml')
+        r=req.get(f"https://bongda.com.vn/tim-kiem.html?q={quote(topic)}",headers=_search_headers(),timeout=12)
+        soup=BeautifulSoup(r.text,'lxml')
         for a in soup.select('h3 a[href], .title a[href]')[:limit*2]:
-            t=_clean(a.get_text(strip=True));href=a.get('href','')
+            t=_clean(a.get_text(strip=True))
+            href=a.get('href','')
             if t and len(t)>15 and _has_kw(topic,t):
-                if not href.startswith('http'):href='https://bongda.com.vn'+href
-                items.append({'title':t,'url':href,'via':'Bóng Đá'})
-            if len(items)>=limit:break
-    except:pass
+                if not href.startswith('http'):
+                    href='https://bongda.com.vn'+href
+                items.append({'title':t,'url':href,'via':'Bong Da'})
+            if len(items)>=limit:
+                break
+    except:
+        pass
     return items
 
 def _s_genk(topic,limit=5):
     items=[]
     try:
-        r=req.get(f"https://genk.vn/tim-kiem?q={quote(topic)}",headers=_search_headers(),timeout=12);soup=BeautifulSoup(r.text,'lxml')
+        r=req.get(f"https://genk.vn/tim-kiem?q={quote(topic)}",headers=_search_headers(),timeout=12)
+        soup=BeautifulSoup(r.text,'lxml')
         for a in soup.select('a[href$=".chn"]')[:limit*3]:
-            t=_clean(a.get('title','') or a.get_text(strip=True));href=a.get('href','')
+            t=_clean(a.get('title','') or a.get_text(strip=True))
+            href=a.get('href','')
             if t and len(t)>15 and _has_kw(topic,t):
-                if href.startswith('/'):href='https://genk.vn'+href
+                if href.startswith('/'):
+                    href='https://genk.vn'+href
                 items.append({'title':t,'url':href,'via':'GenK'})
-            if len(items)>=limit:break
-    except:pass
+            if len(items)>=limit:
+                break
+    except:
+        pass
     return items
 
 def _s_thanhnien(topic,limit=6):
     items=[]
     try:
-        r=req.get(f"https://thanhnien.vn/tim-kiem?q={quote(topic)}",headers=_search_headers(),timeout=12);soup=BeautifulSoup(r.text,'lxml')
+        r=req.get(f"https://thanhnien.vn/tim-kiem?q={quote(topic)}",headers=_search_headers(),timeout=12)
+        soup=BeautifulSoup(r.text,'lxml')
         for a in soup.select('h3 a[href], .box-title a')[:limit*2]:
-            t=_clean(a.get('title','') or a.get_text(strip=True));href=a.get('href','')
+            t=_clean(a.get('title','') or a.get_text(strip=True))
+            href=a.get('href','')
             if t and len(t)>15 and _has_kw(topic,t):
-                if not href.startswith('http'):href='https://thanhnien.vn'+href
-                items.append({'title':t,'url':href,'via':'Thanh Niên'})
-            if len(items)>=limit:break
-    except:pass
+                if not href.startswith('http'):
+                    href='https://thanhnien.vn'+href
+                items.append({'title':t,'url':href,'via':'Thanh Nien'})
+            if len(items)>=limit:
+                break
+    except:
+        pass
     return items
 
 def _s_tuoitre(topic,limit=6):
     items=[]
     try:
-        r=req.get(f"https://tuoitre.vn/tim-kiem.htm?keywords={quote(topic)}",headers=_search_headers(),timeout=12);soup=BeautifulSoup(r.text,'lxml')
+        r=req.get(f"https://tuoitre.vn/tim-kiem.htm?keywords={quote(topic)}",headers=_search_headers(),timeout=12)
+        soup=BeautifulSoup(r.text,'lxml')
         for a in soup.select('h3 a[href], .box-title-text a')[:limit*2]:
-            t=_clean(a.get('title','') or a.get_text(strip=True));href=a.get('href','')
+            t=_clean(a.get('title','') or a.get_text(strip=True))
+            href=a.get('href','')
             if t and len(t)>15 and _has_kw(topic,t):
-                if not href.startswith('http'):href='https://tuoitre.vn'+href
-                items.append({'title':t,'url':href,'via':'Tuổi Trẻ'})
-            if len(items)>=limit:break
-    except:pass
+                if not href.startswith('http'):
+                    href='https://tuoitre.vn'+href
+                items.append({'title':t,'url':href,'via':'Tuoi Tre'})
+            if len(items)>=limit:
+                break
+    except:
+        pass
     return items
 
 def _s_thethaovanhoa(topic,limit=5):
     items=[]
     try:
-        r=req.get(f"https://thethaovanhoa.vn/tim-kiem.htm?keyword={quote(topic)}",headers=_search_headers(),timeout=12);soup=BeautifulSoup(r.text,'lxml')
+        r=req.get(f"https://thethaovanhoa.vn/tim-kiem.htm?keyword={quote(topic)}",headers=_search_headers(),timeout=12)
+        soup=BeautifulSoup(r.text,'lxml')
         for a in soup.select('h3 a[href], .title a[href]')[:limit*2]:
-            t=_clean(a.get('title','') or a.get_text(strip=True));href=a.get('href','')
+            t=_clean(a.get('title','') or a.get_text(strip=True))
+            href=a.get('href','')
             if t and len(t)>15 and _has_kw(topic,t):
-                if not href.startswith('http'):href='https://thethaovanhoa.vn'+href
+                if not href.startswith('http'):
+                    href='https://thethaovanhoa.vn'+href
                 items.append({'title':t,'url':href,'via':'TT&VH'})
-            if len(items)>=limit:break
-    except:pass
+            if len(items)>=limit:
+                break
+    except:
+        pass
     return items
 
 def _s_google_news(topic,limit=8):
-    """Google News RSS fallback"""
     items=[]
     try:
         r=req.get(f"https://news.google.com/rss/search?q={quote(topic)}+Viet+Nam&hl=vi&gl=VN&ceid=VN:vi",headers=_search_headers(),timeout=10)
@@ -342,8 +381,9 @@ def _s_google_news(topic,limit=8):
             link=item.find('link').get_text() if item.find('link') else ''
             source=item.find('source').get_text() if item.find('source') else 'Google News'
             if title and link and len(title)>15:
-                items.append({'title':title,'url':link,'via':f'Google {source}'})
-    except:pass
+                items.append({'title':title,'url':link,'via':'Google '+source})
+    except:
+        pass
     return items
 
 def _search_all(topic,limit=36):
@@ -351,17 +391,24 @@ def _search_all(topic,limit=36):
     with ThreadPoolExecutor(8) as ex:
         futs={ex.submit(_s_vnexpress,topic,8):'vne',ex.submit(_s_dantri,topic,8):'dt',ex.submit(_s_vietnamnet,topic,6):'vnn',ex.submit(_s_bongda,topic,5):'bd',ex.submit(_s_genk,topic,5):'gk',ex.submit(_s_thanhnien,topic,6):'tn',ex.submit(_s_tuoitre,topic,6):'tt',ex.submit(_s_thethaovanhoa,topic,5):'tvh'}
         for f in as_completed(futs,timeout=20):
-            try:results[futs[f]]=f.result()
-            except:results[futs[f]]=[]
-    srcs=list(results.values());out=[];seen=set()
+            try:
+                results[futs[f]]=f.result()
+            except:
+                results[futs[f]]=[]
+    srcs=list(results.values())
+    out=[]
+    seen=set()
     for i in range(max((len(s) for s in srcs),default=0)):
         for s in srcs:
-            if i<len(s) and s[i].get('url') and s[i]['url'] not in seen:seen.add(s[i]['url']);out.append(s[i])
+            if i<len(s) and s[i].get('url') and s[i]['url'] not in seen:
+                seen.add(s[i]['url'])
+                out.append(s[i])
     if not out:
         try:
             google_items=_s_google_news(topic,limit)
             out.extend(google_items)
-        except:pass
+        except:
+            pass
     return out[:limit]
 
 for _path in ['/api/article', '/api/hot_topics', '/api/categories', '/api/storage_status']:
@@ -461,7 +508,7 @@ def _scrape_article_fast(url):
                             'body': fallback, 'source': domain, 'url': url, 'fallback': True}
             if title:
                 return {'title': _clean(title), 'summary': '', 'og_image': '',
-                        'body': [{'type': 'p', 'text': 'Nội dung đang được tải...'}],
+                        'body': [{'type': 'p', 'text': 'Noi dung dang duoc tai...'}],
                         'source': domain, 'url': url, 'fallback': True}
             break
         except Exception:
@@ -485,7 +532,7 @@ def api_article_v2(url: str = Query(...)):
             resp = JSONResponse(data)
             resp.headers["Cache-Control"] = "public, max-age=1800"
             return resp
-        result = {'error': 'Không đọc được', 'url': safe_url}
+        result = {'error': 'Khong doc duoc', 'url': safe_url}
         resp = JSONResponse(result)
         resp.headers["Cache-Control"] = "public, max-age=60"
         return resp
@@ -495,134 +542,133 @@ def api_article_v2(url: str = Query(...)):
 _hot_cache={'t':0,'d':[]}
 def _get_hot_topics():
     now=time.time()
-    if _hot_cache['d'] and now-_hot_cache['t']<600:return _hot_cache['d']
-    freq={};display={}
+    if _hot_cache['d'] and now-_hot_cache['t']<600:
+        return _hot_cache['d']
+    freq={}
+    display={}
     feeds=['https://vnexpress.net/rss/tin-moi-nhat.rss','https://dantri.com.vn/rss/home.rss','https://vietnamnet.vn/rss/tin-moi-nhat.rss','https://thanhnien.vn/rss/home.rss','https://tuoitre.vn/rss/tin-moi-nhat.rss','https://genk.vn/rss','https://vnexpress.net/rss/the-thao.rss','https://thethaovanhoa.vn/rss/tin-nong.rss']
     for feed_url in feeds:
         try:
-            r=req.get(feed_url,headers={'User-Agent':'Mozilla/5.0'},timeout=6);r.encoding='utf-8';soup=BeautifulSoup(r.text,'xml')
+            r=req.get(feed_url,headers={'User-Agent':'Mozilla/5.0'},timeout=6)
+            r.encoding='utf-8'
+            soup=BeautifulSoup(r.text,'xml')
             for item in soup.find_all('item')[:12]:
                 title=_clean(item.find('title').get_text() if item.find('title') else '')
-                if not title:continue
-                title=re.sub(r'\s*[-|].*$','',title);words=[w for w in re.findall(r'[A-Za-zÀ-ỹ0-9]+',title) if len(w)>2 and w.lower() not in _STOP]
-                if len(words)<2:continue
+                if not title:
+                    continue
+                title=re.sub(r'\s*[-|].*$','',title)
+                words=[w for w in re.findall(r'[A-Za-z0-9]+',title) if len(w)>2 and w.lower() not in _STOP]
+                if len(words)<2:
+                    continue
                 for n in(3,4,2):
                     for i in range(max(0,len(words)-n+1)):
                         phrase=' '.join(words[i:i+n])
-                        if 8<=len(phrase)<=45:key=phrase.lower();freq[key]=freq.get(key,0)+1;display[key]=phrase
-        except:continue
-    ranked=sorted(freq.items(),key=lambda x:x[1],reverse=True);topics=[];seen=set()
+                        if 8<=len(phrase)<=45:
+                            key=phrase.lower()
+                            freq[key]=freq.get(key,0)+1
+                            display[key]=phrase
+        except:
+            continue
+    ranked=sorted(freq.items(),key=lambda x:x[1],reverse=True)
+    topics=[]
+    seen=set()
     for key,count in ranked:
         is_dup=any(len(set(e.split())&set(key.split()))/max(len(set(e.split())),len(set(key.split())),1)>0.6 for e in seen)
-        if is_dup:continue
-        seen.add(key);topics.append({'label':'#'+re.sub(r'\s+','',display[key].title()),'topic':display[key],'count':count})
-        if len(topics)>=20:break
-    for kw in['World Cup 2026','Kinh tế Việt Nam','Bóng đá châu Âu','Công nghệ AI','Giá vàng','Thời tiết']:
-        if len(topics)>=24:break
-        if not any(kw.lower() in s for s in seen):topics.append({'label':'#'+re.sub(r'\s+','',kw.title()),'topic':kw,'count':0})
-    _hot_cache.update({'t':now,'d':topics[:24]});return topics[:24]
+        if is_dup:
+            continue
+        seen.add(key)
+        topics.append({'label':'#'+re.sub(r'\s+','',display[key].title()),'topic':display[key],'count':count})
+        if len(topics)>=20:
+            break
+    for kw in['World Cup 2026','Kinh te Viet Nam','Bong da chau Au','Cong nghe AI','Gia vang','Thoi tiet']:
+        if len(topics)>=24:
+            break
+        if not any(kw.lower() in s for s in seen):
+            topics.append({'label':'#'+re.sub(r'\s+','',kw.title()),'topic':kw,'count':0})
+    _hot_cache.update({'t':now,'d':topics[:24]})
+    return topics[:24]
 
 @app.get('/api/hot_topics')
 def api_hot_topics():
     resp = JSONResponse({'topics':_get_hot_topics()})
     resp.headers["Cache-Control"] = "public, max-age=120"
     return resp
+
 @app.get('/')
 async def serve_index():
     p=os.path.join(STATIC_DIR,'index_v2.html')
-    if os.path.exists(p):return FileResponse(p,media_type='text/html')
+    if os.path.exists(p):
+        return FileResponse(p,media_type='text/html')
     return HTMLResponse('<h1>VNEWS</h1>')
+
 @app.get('/api/hashtag/sources')
 def _ht(topic:str=Query(...),page:int=Query(default=0),source:str=Query(default=None)):
     if source == 'google':
         items=_s_google_news(topic,36)
     else:
         items=_search_all(topic,36)
-    per_page=8;start=page*per_page;end=start+per_page
+    per_page=8
+    start=page*per_page
+    end=start+per_page
     return JSONResponse({'sources':items[start:end],'topic':topic,'page':page,'has_more':end<len(items),'total':len(items)})
+
 @app.get('/api/categories')
-def _cat():return JSONResponse([])
+def _cat():
+    return JSONResponse([])
+
 @app.get('/api/storage_status')
-def _st():return JSONResponse({'persistent':os.path.isdir('/data') and os.access('/data',os.W_OK)})
+def _st():
+    return JSONResponse({'persistent':os.path.isdir('/data') and os.access('/data',os.W_OK)})
 
 def _render_slides_page(post, safe_title, safe_img, safe_url):
     slides = post.get('slides', [])
     if not safe_img and slides and slides[0].get('image'):
         safe_img = slides[0].get('image', '')
-    description = _clean((post.get('text') or '')[:200]) or "Tin tức tóm tắt, AI rewrite, World Cup 2026"
+    description = _clean((post.get('text') or '')[:200]) or "Tin tuc tom tat, AI rewrite, World Cup 2026"
     if safe_url and safe_url != '/':
         canonical_url = f"{SPACE}/s?url={quote(safe_url)}&title={quote(safe_title[:100])}"
     else:
         canonical_url = f"{SPACE}/s?post_id={post.get('id') or ''}"
-    h = f'''<!DOCTYPE html>
-<html lang="vi">
-<head>
-<meta charset="utf-8">
-<meta name="viewport" content="width=device-width,initial-scale=1">
-<title>{_clean(safe_title)}</title>
-<meta property="og:title" content="{_clean(safe_title)}">
-<meta property="og:image" content="{_clean(safe_img)}">
-<meta property="og:description" content="{description}">
-<meta property="og:url" content="{canonical_url}">
-<link rel="canonical" href="{canonical_url}">
-<style>
-*{{box-sizing:border-box;margin:0;padding:0}}body{{background:#111;color:#eee;font-family:-apple-system,BlinkMacSystemFont,'Segoe UI',Roboto,sans-serif;padding:12px}}
-.slide-card{{background:#1a1a1a;border:1px solid #2a2a2a;border-radius:12px;padding:16px;margin-bottom:12px;max-width:600px;margin-left:auto;margin-right:auto}}
-.slide-num{{color:#5cb87a;font-size:12px;font-weight:700;margin-bottom:6px}}
-.slide-img{{width:100%;max-height:300px;object-fit:cover;border-radius:8px;margin-bottom:8px}}
-.slide-text{{color:#ddd;font-size:14px;line-height:1.6;margin:0}}
-</style>
-</head>
-<body>'''
+    h = '<!DOCTYPE html>\n<html lang="vi">\n<head>\n<meta charset="utf-8">\n'
+    h += '<title>'+_clean(safe_title)+'</title>\n'
+    h += '<meta property="og:title" content="'+_clean(safe_title)+'">\n'
+    h += '<meta property="og:image" content="'+_clean(safe_img)+'">\n'
+    h += '<meta property="og:description" content="'+description+'">\n'
+    h += '<meta property="og:url" content="'+canonical_url+'">\n'
+    h += '<link rel="canonical" href="'+canonical_url+'">\n'
+    h += '<style>*{box-sizing:border-box;margin:0;padding:0}body{background:#111;color:#eee;padding:12px}.slide-card{background:#1a1a1a;border:1px solid #2a2a2a;border-radius:12px;padding:16px;margin-bottom:12px}.slide-num{color:#5cb87a;font-size:12px;font-weight:700}.slide-img{width:100%;max-height:300px;object-fit:cover;border-radius:8px}.slide-text{color:#ddd;font-size:14px;line-height:1.6}</style>\n</head><body>'
     for s in slides:
         img_src = s.get('image', '')
-        if img_src and ('cdnphoto.dantri' in img_src or 'refooty' in img_src or 'vnexpress' in img_src or 'vcdn' in img_src):
-            img_tag = f'<img src="/api/proxy/img?url={quote(img_src, safe="")}" class="slide-img" loading="lazy" onerror="this.style.display=&#39;none&#39;">'
-        else:
-            img_tag = f'<img src="{_clean(img_src)}" class="slide-img" loading="lazy" onerror="this.style.display=&#39;none&#39;">' if img_src else ''
-        h += f'<div class="slide-card"><div class="slide-num">Slide {s.get("index",1)}/{len(slides)}</div>{img_tag}<p class="slide-text">{_clean(s.get("text",""))}</p></div>'
+        img_tag = ''
+        if img_src:
+            if 'cdnphoto.dantri' in img_src or 'refooty' in img_src or 'vnexpress' in img_src or 'vcdn' in img_src:
+                img_tag = '<img src="/api/proxy/img?url='+quote(img_src, safe="")+'" class="slide-img" loading="lazy" onerror="this.style.display=\'none\'">'
+            else:
+                img_tag = '<img src="'+_clean(img_src)+'" class="slide-img" loading="lazy" onerror="this.style.display=\'none\'">'
+        h += '<div class="slide-card"><div class="slide-num">Slide '+str(s.get("index",1))+'/'+str(len(slides))+'</div>'+img_tag+'<p class="slide-text">'+_clean(s.get("text",""))+'</p></div>'
     h += '</body></html>'
     return HTMLResponse(h)
 
 def _render_video_page(post, safe_title, safe_img, safe_url):
     video_url = post.get('video', '')
-    description = _clean((post.get('text') or '')[:200]) or "Tin tức tóm tắt, AI rewrite, World Cup 2026"
     if safe_url and safe_url != '/':
         canonical_url = f"{SPACE}/s?url={quote(safe_url)}&title={quote(safe_title[:100])}"
     else:
         canonical_url = f"{SPACE}/s?post_id={post.get('id') or ''}"
-    h = f'''<!DOCTYPE html>
-<html lang="vi">
-<head>
-<meta charset="utf-8">
-<meta name="viewport" content="width=device-width,initial-scale=1,maximum-scale=1,user-scalable=no">
-<title>{_clean(safe_title)}</title>
-<meta property="og:title" content="{_clean(safe_title)}">
-<meta property="og:image" content="{_clean(safe_img)}">
-<meta property="og:description" content="{description}">
-<meta property="og:url" content="{canonical_url}">
-<link rel="canonical" href="{canonical_url}">
-<meta name="twitter:card" content="player">
-<meta name="twitter:player" content="{video_url}">
-<style>
-*{{box-sizing:border-box;margin:0;padding:0}}body{{background:#111;color:#eee;font-family:-apple-system,BlinkMacSystemFont,'Segoe UI',Roboto,sans-serif;padding:0;overflow:hidden}}
-.video-container{{width:100vw;height:100vh;display:flex;flex-direction:column;align-items:center;justify-content:center;background:#000}}
-video{{width:100%;height:100%;max-height:100vh;object-fit:contain;background:#000}}
-.title-bar{{position:fixed;bottom:0;left:0;right:0;background:linear-gradient(transparent,rgba(0,0,0,.8));padding:40px 16px 16px;text-align:center}}
-.title-text{{color:#fff;font-size:13px;line-height:1.4;max-width:600px;margin:0 auto}}
-</style>
-</head>
-<body>
-<div class="video-container">
-<video src="{_clean(video_url)}" controls autoplay playsinline loop></video>
-<div class="title-bar"><div class="title-text">{_clean(safe_title)}</div></div>
-</div>
-</body></html>'''
+    h = '<!DOCTYPE html>\n<html lang="vi">\n<head>\n<meta charset="utf-8">\n<meta name="viewport" content="width=device-width,initial-scale=1,maximum-scale=1,user-scalable=no">\n'
+    h += '<title>'+_clean(safe_title)+'</title>\n'
+    h += '<meta property="og:title" content="'+_clean(safe_title)+'">\n'
+    h += '<meta property="og:image" content="'+_clean(safe_img)+'">\n'
+    h += '<meta property="og:url" content="'+canonical_url+'">\n'
+    h += '<link rel="canonical" href="'+canonical_url+'">\n'
+    h += '<style>*{margin:0;padding:0;box-sizing:border-box}body{background:#000;overflow:hidden}video{width:100vw;height:100vh;object-fit:contain}</style>\n</head><body>\n'
+    h += '<video src="'+_clean(video_url)+'" controls autoplay playsinline loop></video>\n'
+    h += '</body></html>'
     return HTMLResponse(h)
 
 @app.get('/s/{slug}')
 async def _sh_slug(slug: str, request: Request, url: str = '', title: str = '', img: str = ''):
-    safe_title = _clean(title) if title else 'VNEWS - Tin tức'
+    safe_title = _clean(title) if title else 'VNEWS'
     safe_img = _clean(img) if img else ''
     safe_url = _clean(url) if url else '/'
     post = None
@@ -642,23 +688,11 @@ async def _sh_slug(slug: str, request: Request, url: str = '', title: str = '',
         return _render_slides_page(post, safe_title, safe_img, safe_url)
     if post and post.get('video'):
         return _render_video_page(post, safe_title, safe_img, safe_url)
-    return HTMLResponse(f'''<!DOCTYPE html>
-<html lang="vi">
-<head>
-<meta charset="utf-8">
-<meta name="viewport" content="width=device-width,initial-scale=1">
-<title>{_clean(safe_title)}</title>
-<meta property="og:title" content="{_clean(safe_title)}">
-<meta property="og:image" content="{_clean(safe_img)}">
-<meta property="og:description" content="Tin tức tóm tắt, AI rewrite, World Cup 2026">
-<meta property="og:url" content="{SPACE}/s/{slug}">
-<link rel="canonical" href="{SPACE}/s/{slug}">
-<meta http-equiv="refresh" content="0;url={safe_url}">
-</head><body></body></html>''')
+    return HTMLResponse(f'<!DOCTYPE html><html lang="vi"><head><meta charset="utf-8"><title>{_clean(safe_title)}</title><meta property="og:title" content="{_clean(safe_title)}"><meta property="og:image" content="{_clean(safe_img)}"><meta http-equiv="refresh" content="0;url={safe_url}"></head><body></body></html>')
 
 @app.get('/s')
 async def _sh(url:str='',title:str='',img:str='',post_id:str=''):
-    safe_title = _clean(title) if title else 'VNEWS - Tin tức'
+    safe_title = _clean(title) if title else 'VNEWS'
     safe_img = _clean(img) if img else ''
     safe_url = _clean(url) if url else '/'
     post = None
@@ -694,19 +728,7 @@ async def _sh(url:str='',title:str='',img:str='',post_id:str=''):
         return _render_slides_page(post, safe_title, safe_img, safe_url)
     if post and post.get('video'):
         return _render_video_page(post, safe_title, safe_img, safe_url)
-    return HTMLResponse(f'''<!DOCTYPE html>
-<html lang="vi">
-<head>
-<meta charset="utf-8">
-<meta name="viewport" content="width=device-width,initial-scale=1">
-<title>{safe_title}</title>
-<meta property="og:title" content="{safe_title}">
-<meta property="og:image" content="{safe_img}">
-<meta property="og:description" content="Tin tức tóm tắt, AI rewrite, World Cup 2026">
-<meta property="og:url" content="{SPACE}/s?url={quote(safe_url)}">
-<link rel="canonical" href="{SPACE}/s?url={quote(safe_url)}">
-<meta http-equiv="refresh" content="0;url={safe_url}">
-</head><body></body></html>''')
+    return HTMLResponse(f'<!DOCTYPE html><html lang="vi"><head><meta charset="utf-8"><title>{safe_title}</title><meta property="og:title" content="{safe_title}"><meta property="og:image" content="{safe_img}"><meta http-equiv="refresh" content="0;url={safe_url}"></head><body></body></html>')
 
 from wc2026_scraper import scrape_summary,scrape_fixtures,scrape_standings,scrape_stats,scrape_wc_news,scrape_road_to_wc,get_wc2026_all,scrape_history,scrape_h2h,scrape_lineups,scrape_match_detail
 
@@ -786,25 +808,35 @@ def proxy_xlb(path: str = Query(default="")):
         return JSONResponse({"videos": [], "count": 0, "error": str(e)}, status_code=500)
 
 @app.get('/api/wc2026')
-def _w():return JSONResponse(get_wc2026_all())
+def _w():
+    return JSONResponse(get_wc2026_all())
 @app.get('/api/wc2026/fixtures')
-def _wf():return JSONResponse(scrape_fixtures())
+def _wf():
+    return JSONResponse(scrape_fixtures())
 @app.get('/api/wc2026/standings')
-def _ws():return JSONResponse(scrape_standings())
+def _ws():
+    return JSONResponse(scrape_standings())
 @app.get('/api/wc2026/stats')
-def _wst():return JSONResponse(scrape_stats())
+def _wst():
+    return JSONResponse(scrape_stats())
 @app.get('/api/wc2026/history')
-def _whi():return JSONResponse(scrape_history())
+def _whi():
+    return JSONResponse(scrape_history())
 @app.get('/api/wc2026/news')
-def _wn():return JSONResponse(scrape_wc_news())
+def _wn():
+    return JSONResponse(scrape_wc_news())
 @app.get('/api/wc2026/road')
-def _wr():return JSONResponse(scrape_road_to_wc())
+def _wr():
+    return JSONResponse(scrape_road_to_wc())
 @app.get('/api/wc2026/h2h/{eid}')
-def _wh2(eid:int):return JSONResponse(scrape_h2h(eid))
+def _wh2(eid:int):
+    return JSONResponse(scrape_h2h(eid))
 @app.get('/api/wc2026/lineups/{eid}')
-def _wl(eid:int):return JSONResponse(scrape_lineups(eid))
+def _wl(eid:int):
+    return JSONResponse(scrape_lineups(eid))
 @app.get('/api/wc2026/match/{eid}')
-def _wm(eid:int):return JSONResponse(scrape_match_detail(eid))
+def _wm(eid:int):
+    return JSONResponse(scrape_match_detail(eid))
 
 DATA_DIR='/data' if os.path.isdir('/data') else os.path.join(os.path.dirname(os.path.abspath(__file__)),'data')
 os.makedirs(DATA_DIR,exist_ok=True)
@@ -817,34 +849,62 @@ os.makedirs(WALL_VIDEO_DIR,exist_ok=True)
 _il=threading.Lock();_cl=threading.Lock();_wl_lock=threading.Lock()
 def _lj(p):
     try:
-        if os.path.exists(p):return json.load(open(p,'r',encoding='utf-8'))
-    except:pass
-    return{}
+        if os.path.exists(p):
+            return json.load(open(p,'r',encoding='utf-8'))
+    except:
+        pass
+    return {}
 def _sj(p,d):
-    try:open(p+'.tmp','w',encoding='utf-8').write(json.dumps(d,ensure_ascii=False));os.replace(p+'.tmp',p)
-    except:pass
+    try:
+        open(p+'.tmp','w',encoding='utf-8').write(json.dumps(d,ensure_ascii=False))
+        os.replace(p+'.tmp',p)
+    except:
+        pass
 
 @app.post('/api/v2/interact')
 async def _int(request:Request):
-    b=await request.json();v=str(b.get('id','')).strip();t=str(b.get('type','')).strip()
-    if not v or t not in('view','like'):return JSONResponse({'error':'x'},status_code=400)
-    with _il:db=_lj(IF);db.setdefault(v,{'views':0,'likes':0,'comments':0});db[v][t+'s']+=1;_sj(IF,db);return JSONResponse(db[v])
+    b=await request.json()
+    v=str(b.get('id','')).strip()
+    t=str(b.get('type','')).strip()
+    if not v or t not in('view','like'):
+        return JSONResponse({'error':'x'},status_code=400)
+    with _il:
+        db=_lj(IF)
+        db.setdefault(v,{'views':0,'likes':0,'comments':0})
+        db[v][t+'s']+=1
+        _sj(IF,db)
+        return JSONResponse(db[v])
 
 @app.get('/api/v2/interactions')
 def _gi(id:str=Query(...)):
-    with _il:return JSONResponse(_lj(IF).get(id.strip(),{'views':0,'likes':0,'comments':0}))
+    with _il:
+        return JSONResponse(_lj(IF).get(id.strip(),{'views':0,'likes':0,'comments':0}))
 
 @app.get('/api/v2/comments')
 def _gc(id:str=Query(...)):
-    with _cl:return JSONResponse({'comments':_lj(CF).get(id.strip(),[])})
+    with _cl:
+        return JSONResponse({'comments':_lj(CF).get(id.strip(),[])})
 
 @app.post('/api/v2/comment')
 async def _pc(request:Request):
-    b=await request.json();v=str(b.get('id','')).strip();tx=str(b.get('text','')).strip()[:500]
-    if not v or not tx:return JSONResponse({'error':'x'},status_code=400)
+    b=await request.json()
+    v=str(b.get('id','')).strip()
+    tx=str(b.get('text','')).strip()[:500]
+    if not v or not tx:
+        return JSONResponse({'error':'x'},status_code=400)
     c={'text':tx,'time':time.strftime('%H:%M %d/%m',time.localtime()),'ts':int(time.time())}
-    with _cl:db=_lj(CF);db.setdefault(v,[]);db[v].append(c);db[v]=db[v][-200:];_sj(CF,db);cms=db[v]
-    with _il:idb=_lj(IF);idb.setdefault(v,{'views':0,'likes':0,'comments':0});idb[v]['comments']=len(cms);_sj(IF,idb)
+    with _cl:
+        db=_lj(CF)
+        db.setdefault(v,[])
+        db[v].append(c)
+        db[v]=db[v][-200:]
+        _sj(CF,db)
+        cms=db[v]
+    with _il:
+        idb=_lj(IF)
+        idb.setdefault(v,{'views':0,'likes':0,'comments':0})
+        idb[v]['comments']=len(cms)
+        _sj(IF,idb)
     return JSONResponse({'comments':cms})
 
 def _load_wall_posts():
@@ -870,7 +930,7 @@ async def api_wall_post(request: Request):
             form = await request.form()
         except Exception as e:
             return JSONResponse({"error": f"Form parse error: {str(e)}"}, status_code=400)
-        title = form.get('title', 'Video mới') or 'Video mới'
+        title = form.get('title', 'Video moi') or 'Video moi'
         text = form.get('text', '') or ''
         source = form.get('source', 'vtv_recorder') or 'vtv_recorder'
         video_file = form.get('video')
@@ -884,7 +944,7 @@ async def api_wall_post(request: Request):
                 ext = '.webm'
             else:
                 ext = '.webm'
-            video_filename = f"wall_{post_id}{ext}"
+            video_filename = "wall_"+post_id+ext
             video_path = os.path.join(WALL_VIDEO_DIR, video_filename)
             try:
                 content = await video_file.read()
@@ -895,10 +955,10 @@ async def api_wall_post(request: Request):
                 file_size_mb = len(content) / 1024 / 1024
                 if file_size_mb > 50:
                     os.remove(video_path)
-                    return JSONResponse({"error": f"Video qua lon ({file_size_mb:.1f}MB). Toi da 50MB."}, status_code=400)
-                video_url = f"/api/wall/video/{video_filename}"
+                    return JSONResponse({"error": f"Video too large ({file_size_mb:.1f}MB). Max 50MB."}, status_code=400)
+                video_url = "/api/wall/video/"+video_filename
             except Exception as e:
-                return JSONResponse({"error": f"Loi luu video: {str(e)}"}, status_code=500)
+                return JSONResponse({"error": "Video save error: "+str(e)}, status_code=500)
         post = {
             "id": post_id,
             "title": title[:200],
@@ -973,24 +1033,5 @@ def api_wall_delete(post_id: str):
             return JSONResponse({"ok": True})
     return JSONResponse({"error": "Post not found"}, status_code=404)
 
-# ===== LANGUAGE & EMOTION DETECTION =====
-import random as _random2
-from urllib.parse import quote as _quote2
-
-_UA_RW = {'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36', 'Accept-Language': 'vi-VN,vi;q=0.9'}
-
-_UNIQUE_CHARS = {
-    'vietnamese': set('daeooua a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a a'),
-    'spanish': set('n ye'),
-    'portuguese': set('aocaeiua'),
-}
-
-_STOPWORDS = {
-    'english': {'the', 'is', 'at', 'which', 'on', 'a', 'an', 'and', 'or', 'but', 'in', 'with', 'to', 'for', 'of', 'not', 'no', 'can', 'had', 'have', 'has', 'was', 'were', 'are', 'be', 'been', 'this', 'that', 'it', 'he', 'she', 'they', 'his', 'her', 'my', 'your', 'our', 'we', 'you', 'i'},
-    'vietnamese': {'la', 'cua', 'va', 'co', 'duoc', 'cho', 'khong', 'voi', 'nay', 'do', 'tu', 'trong', 'da', 'se', 'mot', 'cac', 'nhung', 've', 'tai', 'nguoi', 'nam', 'den', 'ra', 'lai', 'nhu', 'khi', 'de', 'rat', 'cung', 'ma', 'neu', 'sau', 'tren', 'theo', 'vi', 'do', 'nen', 'thi', 'minh', 'toi', 'ban', 'anh', 'chi', 'em'},
-    'portuguese': {'de', 'um', 'que', 'e', 'do', 'da', 'em', 'para', 'com', 'nao', 'uma', 'os', 'no', 'se', 'na', 'por', 'mais', 'as', 'dos', 'como', 'mas', 'ao', 'ele', 'das', 'tem', 'seu', 'sua', 'ou', 'quando', 'muito', 'nos', 'ja', 'eu', 'tambem', 'so', 'pelo', 'pela', 'ate', 'isso', 'ela', 'entre', 'depois', 'sem', 'mesmo', 'aos', 'sao', 'esta', 'ter', 'ser', 'foi', 'era', 'ha', 'estao', 'voce', 'nos', 'eles', 'elas'},
-    'spanish': {'de', 'que', 'el', 'en', 'y', 'a', 'los', 'del', 'se', 'las', 'por', 'un', 'para', 'con', 'no', 'una', 'su', 'al', 'es', 'lo', 'como', 'mas', 'pero', 'sus', 'le', 'ya', 'o', 'fue', 'este', 'ha', 'si', 'porque', 'esta', 'son', 'entre', 'esta', 'cuando', 'muy', 'sin', 'sobre', 'ser', 'tambien', 'me', 'hasta', 'hay', 'donde', 'han', 'quien', 'estan', 'desde', 'todo', 'nos', 'durante', 'todos', 'uno', 'les', 'ni', 'contra', 'otros', 'fueron', 'ese', 'eso', 'ante', 'ellos', 'yo', 'tu', 'el', 'ella', 'nosotros', 'usted', 'ustedes'},
-}
-
 print(f"[VNEWS] app_v2_entry loaded. DATA_DIR={DATA_DIR}, STATIC_DIR={STATIC_DIR}")
 print(f"[VNEWS] Persistent storage: {os.path.isdir('/data') and os.access('/data', os.W_OK)}")