Add date_vn to article scraper, add date extraction to search functions

app_v2_entry.py

@@ -1,6 +1,7 @@
 """VNEWS v2 Entry Point - with fast bongda proxy + rewrite endpoints + multilingual TTS"""
 import sys, os
-from main import app, HEADERS, BONGDA_HEADERS, fetch_bongda_api, HL_LEAGUES
+from datetime import timedelta
+from main import app, HEADERS, BONGDA_HEADERS, fetch_bongda_api, HL_LEAGUES, VN_TZ
 
 try:
     import ai_ext
@@ -213,6 +214,11 @@ def _has_kw(topic,title):
     if not words:return True
     return any(w in tt for w in words)
 
+def _s_date(el):
+    dt = el.find('time') or el.select_one('.date, .time, span.date, span.time, .meta-date')
+    if dt: return _clean(dt.get_text(strip=True))[:20]
+    return ''
+
 def _s_vnexpress(topic,limit=8):
     items=[]
     try:
@@ -221,7 +227,7 @@ def _s_vnexpress(topic,limit=8):
             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'})
+                if _has_kw(topic,t):items.append({'title':t,'url':a['href'],'via':'VnExpress','date':_s_date(art)})
     except:pass
     return items
 
@@ -233,7 +239,7 @@ def _s_dantri(topic,limit=8):
             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'})
+                items.append({'title':t,'url':href,'via':'Dân Trí','date':_s_date(a.parent or a)})
             if len(items)>=limit:break
     except:pass
     return items
@@ -246,7 +252,7 @@ def _s_vietnamnet(topic,limit=6):
             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
-                items.append({'title':t,'url':href,'via':'VietNamNet'})
+                items.append({'title':t,'url':href,'via':'VietNamNet','date':_s_date(a.parent or a)})
             if len(items)>=limit:break
     except:pass
     return items
@@ -259,7 +265,7 @@ def _s_bongda(topic,limit=5):
             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'})
+                items.append({'title':t,'url':href,'via':'Bóng Đá','date':_s_date(a.parent or a)})
             if len(items)>=limit:break
     except:pass
     return items
@@ -272,7 +278,7 @@ def _s_genk(topic,limit=5):
             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
-                items.append({'title':t,'url':href,'via':'GenK'})
+                items.append({'title':t,'url':href,'via':'GenK','date':_s_date(a.parent or a)})
             if len(items)>=limit:break
     except:pass
     return items
@@ -285,7 +291,7 @@ def _s_thanhnien(topic,limit=6):
             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'})
+                items.append({'title':t,'url':href,'via':'Thanh Niên','date':_s_date(a.parent or a)})
             if len(items)>=limit:break
     except:pass
     return items
@@ -298,7 +304,7 @@ def _s_tuoitre(topic,limit=6):
             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'})
+                items.append({'title':t,'url':href,'via':'Tuổi Trẻ','date':_s_date(a.parent or a)})
             if len(items)>=limit:break
     except:pass
     return items
@@ -311,7 +317,7 @@ def _s_thethaovanhoa(topic,limit=5):
             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
-                items.append({'title':t,'url':href,'via':'TT&VH'})
+                items.append({'title':t,'url':href,'via':'TT&VH','date':_s_date(a.parent or a)})
             if len(items)>=limit:break
     except:pass
     return items
@@ -376,6 +382,39 @@ def _scrape_article_fast(url):
             if ogi:
                 og_img = ogi.get('content', '')
                 if og_img.startswith('//'): og_img = 'https:' + og_img
+            # Extract date
+            pub_date = ""
+            for meta_name in ['article:published_time', 'datePublished', 'date', 'pubdate', 'og:updated_time']:
+                meta = soup.find('meta', property=meta_name) or soup.find('meta', attrs={'name': meta_name})
+                if meta and meta.get('content'):
+                    pub_date = meta.get('content')[:19]
+                    break
+            if not pub_date:
+                time_el = soup.find('time')
+                if time_el and time_el.get('datetime'):
+                    pub_date = time_el.get('datetime')[:19]
+            # Convert to VN time
+            date_vn = ""
+            if pub_date:
+                try:
+                    from datetime import datetime as _dt
+                    from dateutil import parser
+                    dt = parser.parse(pub_date)
+                    vn = dt + timedelta(hours=7)
+                    now = _dt.now(VN_TZ)
+                    diff = (now - vn).total_seconds()
+                    if diff < 3600:
+                        mins = int(diff / 60)
+                        date_vn = f"{mins} phút trước" if mins > 0 else "Vừa xong"
+                    elif diff < 86400:
+                        hrs = int(diff / 3600)
+                        date_vn = f"{hrs} giờ trước"
+                    elif diff < 172800:
+                        date_vn = "Hôm qua"
+                    else:
+                        date_vn = vn.strftime("%d/%m/%Y %H:%M")
+                except:
+                    date_vn = pub_date[:10]
             h1 = soup.find('h1')
             if not title and h1: title = h1.get_text(strip=True)[:200]
             body = []
@@ -416,18 +455,18 @@ def _scrape_article_fast(url):
                                 if ct: body.append({'type': 'p', 'text': ct})
                     if len(body) >= 2:
                         return {'title': _clean(title), 'summary': _clean(summary), 'og_image': og_img,
-                                'body': body[:50], 'source': domain, 'url': url}
+                                'body': body[:50], 'source': domain, 'url': url, 'date_vn': date_vn}
             if title and (summary or og_img):
                 fallback = []
                 if og_img: fallback.append({'type': 'img', 'src': og_img})
                 if summary: fallback.append({'type': 'p', 'text': summary})
                 if fallback:
                     return {'title': _clean(title), 'summary': _clean(summary), 'og_image': og_img,
-                            'body': fallback, 'source': domain, 'url': url, 'fallback': True}
+                            'body': fallback, 'source': domain, 'url': url, 'fallback': True, 'date_vn': date_vn}
             if title:
                 return {'title': _clean(title), 'summary': '', 'og_image': '',
                         'body': [{'type': 'p', 'text': 'Nội dung đang được tải...'}],
-                        'source': domain, 'url': url, 'fallback': True}
+                        'source': domain, 'url': url, 'fallback': True, 'date_vn': date_vn}
             break
         except Exception:
             continue