VNEWS

Running

App Files Files Community

bep40 commited on Jul 10

Commit

6e354ed

verified ·

1 Parent(s): bf2ca14

Upload app_v2_entry.py with huggingface_hub

Browse files

Files changed (1) hide show

app_v2_entry.py +69 -31

app_v2_entry.py CHANGED Viewed

@@ -26,6 +26,7 @@ import asyncio
 HL_LEAGUES['friendly'] = {"path": "giai-khac/friendly", "name": "Giao hữu", "emoji": "🤝"}
 STATIC_DIR = os.path.join(os.path.dirname(os.path.abspath(__file__)), 'static')
 app.router.routes=[r for r in app.router.routes if not(getattr(r,'path',None)=='/' and hasattr(r,'methods') and 'GET' in getattr(r,'methods',set()))]
 app.routes[:]=[r for r in app.routes if not isinstance(r, Mount)]
 app.router.routes=[r for r in app.router.routes if not isinstance(r, Mount)]
@@ -328,7 +329,7 @@ def _search_all(topic,limit=36):
             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])
     return out[:limit]
-for _path in ['/api/article', '/api/hot_topics', '/api/categories', '/api/storage_status', '/s']:
     app.router.routes=[r for r in app.router.routes if not(getattr(r,'path',None)==_path and 'GET' in getattr(r,'methods',set()))]
 _article_cache = {}
@@ -504,8 +505,44 @@ def _ht(topic:str=Query(...),page:int=Query(default=0)):
 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)})
 @app.get('/s')
-async def _sh(url:str='',title:str='',img:str=''):return HTMLResponse(f'<!DOCTYPE html><html><head><meta property="og:title" content="{_clean(title)}"><meta property="og:image" content="{_clean(img)}"><meta http-equiv="refresh" content="0;url={_clean(url) or "/"}"></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
@@ -1075,7 +1112,7 @@ def _scrape_article_for_rewrite(url):
 def _extract_key_points_rw(paragraphs, max_points=5):
-    """Extract key points from paragraphs - extracts ALL sentences, not just first one.
     Fixes: Original regex `^(.+?[.!?])\s` only captured first sentence per paragraph.
     Now splits on all sentence boundaries and takes valid sentences until max_points.
@@ -1326,32 +1363,23 @@ async def _auto_fetch_short(post_id):
     except: pass
     return False
-async def _auto_rewrite_one(topic, slot_label, used_urls=None):
     """Rewrite one topic: find articles, summarize, post to wall, trigger short.
-    used_urls: shared set to avoid duplicate articles across topics."""
     from urllib.parse import quote as _q
-    items = _search_all(topic, limit=6)
     # Skip URLs already used by another topic
     if used_urls is not None:
         filtered = [it for it in items if it.get('url') not in used_urls]
         if filtered:
             items = filtered
-    if not items:
-        # fallback: Google RSS
-        try:
-            r = req.get(f"https://news.google.com/rss/search?q={_q(topic)}&hl=vi&gl=VN&ceid=VN:vi",
-                       headers={'User-Agent':'Mozilla/5.0'}, timeout=8)
-            r.encoding = 'utf-8'
-            soup = BeautifulSoup(r.text, 'xml')
-            for it in soup.find_all('item')[:5]:
-                t = _clean(it.find('title').get_text(' ',strip=True) if it.find('title') else '')
-                lk = it.find('link').get_text(strip=True) if it.find('link') else ''
-                if t and lk: items.append({'title':t,'url':lk,'via':'Google News'})
-        except: pass
-    if not items:
         return False
-    item = items[0]  # best match
     url = item.get('url', '')
     title = item.get('title', topic)
     if url and used_urls is not None:
@@ -1407,7 +1435,7 @@ async def _auto_rewrite_one(topic, slot_label, used_urls=None):
     return True
 async def _do_scheduled_run(slot_label):
-    """Main scheduled run: exactly 3 posts from 3 different HOT topics, no duplicates."""
     print(f"[auto] Starting scheduled rewrite for {slot_label}")
     # Get top hot topics, skip duplicates
@@ -1440,14 +1468,25 @@ async def _do_scheduled_run(slot_label):
     _used_urls = set()
     results = []
     for jt in job_topics:
-        try:
-            ok = await asyncio.wait_for(_auto_rewrite_one(jt, slot_label, _used_urls), timeout=120)
-            results.append((jt, ok))
-        except Exception as e:
-            print(f"[auto] Error on '{jt}': {e}")
-            results.append((jt, False))
-        await asyncio.sleep(2)
     # Log
     from datetime import datetime, timezone, timedelta
@@ -1457,11 +1496,10 @@ async def _do_scheduled_run(slot_label):
     if today_str not in log: log[today_str] = {}
     log[today_str][slot_label] = {
         'time': datetime.now(VN_TZ_SCHED).strftime('%H:%M:%S'),
-        'count': sum(1 for _, ok in results if ok),
-        'total': len(results),
     }
     _save_auto_log(log)
-    print(f"[auto] Done {slot_label}: {sum(1 for _, ok in results if ok)}/{len(results)} posts")
 def _scheduler_loop():
     """Check every 60s; trigger at 7:00, 13:00, 19:00 VN time.

 HL_LEAGUES['friendly'] = {"path": "giai-khac/friendly", "name": "Giao hữu", "emoji": "🤝"}
 STATIC_DIR = os.path.join(os.path.dirname(os.path.abspath(__file__)), 'static')
+SPACE = "https://bep40-vnews.hf.space"  # SEO URL base for share links
 app.router.routes=[r for r in app.router.routes if not(getattr(r,'path',None)=='/' and hasattr(r,'methods') and 'GET' in getattr(r,'methods',set()))]
 app.routes[:]=[r for r in app.routes if not isinstance(r, Mount)]
 app.router.routes=[r for r in app.router.routes if not isinstance(r, Mount)]
             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])
     return out[:limit]
+for _path in ['/api/article', '/api/hot_topics', '/api/categories', '/api/storage_status']:
     app.router.routes=[r for r in app.router.routes if not(getattr(r,'path',None)==_path and 'GET' in getattr(r,'methods',set()))]
 _article_cache = {}
 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)})
+@app.get('/s/{slug}')
+async def _sh_slug(slug: str, url: str = '', title: str = '', img: str = ''):
+    """SEO-friendly share endpoint with slug in URL path."""
+    safe_title = _clean(title) if title else 'VNEWS - Tin tức'
+    safe_img = _clean(img) if img else ''
+    safe_url = _clean(url) if url else '/'
+    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/{slug}">
+<link rel="canonical" href="{SPACE}/s/{slug}">
+<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=''):
+    safe_title = _clean(title) if title else 'VNEWS - Tin tức'
+    safe_img = _clean(img) if img else ''
+    safe_url = _clean(url) if url else '/'
+    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">
+<link rel="canonical" href="{SPACE}/s">
+<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
 def _extract_key_points_rw(paragraphs, max_points=5):
+    r"""Extract key points from paragraphs - extracts ALL sentences, not just first one.
     Fixes: Original regex `^(.+?[.!?])\s` only captured first sentence per paragraph.
     Now splits on all sentence boundaries and takes valid sentences until max_points.
     except: pass
     return False
+async def _auto_rewrite_one(topic, slot_label, used_urls=None, post_index=0):
     """Rewrite one topic: find articles, summarize, post to wall, trigger short.
+    used_urls: shared set to avoid duplicate articles across topics.
+    post_index: 0-based index to create multiple posts per topic (0,1,2 = up to 3 posts)."""
     from urllib.parse import quote as _q
+    # Get MORE items to support 1-3 posts per topic
+    items = _search_all(topic, limit=12)
     # Skip URLs already used by another topic
     if used_urls is not None:
         filtered = [it for it in items if it.get('url') not in used_urls]
         if filtered:
             items = filtered
+    if not items or post_index >= len(items):
         return False
+    # Get article at post_index (0,1,2 for multiple posts)
+    item = items[post_index]  # post_index allows multiple articles per topic
     url = item.get('url', '')
     title = item.get('title', topic)
     if url and used_urls is not None:
     return True
 async def _do_scheduled_run(slot_label):
+    """Main scheduled run: 1-3 posts from 3 different HOT topics (3-9 total), no duplicates."""
     print(f"[auto] Starting scheduled rewrite for {slot_label}")
     # Get top hot topics, skip duplicates
     _used_urls = set()
     results = []
+    # Process each topic, create 1-3 posts per topic
     for jt in job_topics:
+        for post_idx in range(3):  # Try up to 3 posts per topic
+            try:
+                ok = await asyncio.wait_for(_auto_rewrite_one(jt, slot_label, _used_urls, post_idx), timeout=120)
+                if ok:
+                    results.append((jt, post_idx, True))
+                    print(f"[auto] Created post {post_idx+1} for '{jt}'")
+                else:
+                    # No more articles for this topic
+                    break
+            except Exception as e:
+                print(f"[auto] Error on '{jt}' post {post_idx}: {e}")
+                results.append((jt, post_idx, False))
+            await asyncio.sleep(1)  # Small delay between posts
+    # Ensure at least 3 posts total (fallback if needed)
+    successful_posts = sum(1 for _, _, ok in results if ok)
+    print(f"[auto] Done {slot_label}: {successful_posts} posts created")
     # Log
     from datetime import datetime, timezone, timedelta
     if today_str not in log: log[today_str] = {}
     log[today_str][slot_label] = {
         'time': datetime.now(VN_TZ_SCHED).strftime('%H:%M:%S'),
+        'count': successful_posts,
+        'total': len(job_topics),
     }
     _save_auto_log(log)
 def _scheduler_loop():
     """Check every 60s; trigger at 7:00, 13:00, 19:00 VN time.