fix: database persistent + hash share URL + /s/slug/hash SEO route

main.py

@@ -1,8 +1,8 @@
-"""VNEWS - FastAPI backend."""
-import hashlib, re, time
+"""VNEWS - FastAPI backend with persistent DB + hash-based share URLs."""
+import hashlib, re, time, json, os, threading, unicodedata
 from concurrent.futures import ThreadPoolExecutor, as_completed
-from fastapi import FastAPI, Query
-from fastapi.responses import HTMLResponse, JSONResponse
+from fastapi import FastAPI, Query, Body
+from fastapi.responses import HTMLResponse, JSONResponse, FileResponse
 from fastapi.staticfiles import StaticFiles
 from urllib.parse import unquote, quote
 import requests
@@ -13,10 +13,122 @@ HEADERS = {"User-Agent":"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/5
 BASE_BDP = "https://bongdaplus.vn"
 BASE_24H = "https://www.24h.com.vn"
 SPACE_URL = "https://bep40-vnews.hf.space"
-_cache = {}
-_cache_ttl = 300
 
-def _cached(key, fn):
+# ═══════════════════════════════════════════════════════════════
+# DATABASE: persistent JSON on /data volume
+# ═══════════════════════════════════════════════════════════════
+_DATA_DIR = "/data" if os.path.isdir("/data") else "."
+DB_PATH = os.path.join(_DATA_DIR, "db.json")
+SHARE_PATH = os.path.join(_DATA_DIR, "shares.json")
+_db_lock = threading.Lock()
+_share_lock = threading.Lock()
+
+def _load_json(path, default):
+    try:
+        with open(path,"r") as f: return json.load(f)
+    except: return default
+
+def _save_json(path, data):
+    try:
+        with open(path,"w") as f: json.dump(data, f, ensure_ascii=False)
+    except: pass
+
+# Stats DB
+def db_inc(category, aid):
+    with _db_lock:
+        db = _load_json(DB_PATH, {"views":{},"shares":{},"comments":{}})
+        db[category][aid] = db[category].get(aid, 0) + 1
+        _save_json(DB_PATH, db)
+        return db[category][aid]
+
+def db_add_comment(aid, name, text):
+    with _db_lock:
+        db = _load_json(DB_PATH, {"views":{},"shares":{},"comments":{}})
+        if aid not in db["comments"]: db["comments"][aid] = []
+        db["comments"][aid].append({"n":name,"t":text,"d":time.strftime("%d/%m %H:%M")})
+        _save_json(DB_PATH, db)
+        return db["comments"][aid]
+
+def db_get(aid):
+    db = _load_json(DB_PATH, {"views":{},"shares":{},"comments":{}})
+    return {"views":db["views"].get(aid,0),"shares":db["shares"].get(aid,0),"comments":db["comments"].get(aid,[])}
+
+# Share registry
+def _slug(title):
+    s = unicodedata.normalize('NFKD', title.lower())
+    s = ''.join(c for c in s if not unicodedata.combining(c))
+    s = re.sub(r'[^a-z0-9\s-]','',s)
+    s = re.sub(r'[\s]+','-',s.strip())
+    return re.sub(r'-+','-',s).strip('-')[:60]
+
+def _hash(url):
+    return hashlib.md5(url.encode()).hexdigest()[:12]
+
+def register_share(url, title, img=""):
+    h = _hash(url)
+    sl = _slug(title)
+    with _share_lock:
+        db = _load_json(SHARE_PATH, {})
+        db[h] = {"url":url,"title":title,"img":img,"slug":sl}
+        _save_json(SHARE_PATH, db)
+    return sl, h
+
+def get_share(h):
+    db = _load_json(SHARE_PATH, {})
+    return db.get(h)
+
+# ═══════════════════════════════════════════════════════════════
+# API: Database endpoints
+# ═══════════════════════════════════════════════════════════════
+@app.post("/api/view")
+def api_view(aid:str=Body(...,embed=True)):
+    return JSONResponse({"views": db_inc("views", aid)})
+
+@app.post("/api/share_inc")
+def api_share_inc(aid:str=Body(...,embed=True)):
+    return JSONResponse({"shares": db_inc("shares", aid)})
+
+@app.post("/api/comment")
+def api_comment(aid:str=Body(...), name:str=Body(...), text:str=Body(...)):
+    return JSONResponse({"comments": db_add_comment(aid, name, text)})
+
+@app.get("/api/stats/{aid}")
+def api_stats(aid:str):
+    return JSONResponse(db_get(aid))
+
+# Register share → returns hash URL
+@app.post("/api/register_share")
+def api_register_share(url:str=Body(...), title:str=Body(...), img:str=Body("")):
+    sl, h = register_share(url, title, img)
+    db_inc("shares", _hash(url))
+    return JSONResponse({"slug":sl,"hash":h,"share_url":f"{SPACE_URL}/#/{sl}/{h}","social_url":f"{SPACE_URL}/s/{sl}/{h}"})
+
+# SEO share page: og:image for crawlers, JS redirect for users
+@app.get("/s/{slug}/{hash_id}")
+async def share_page(slug:str, hash_id:str):
+    info = get_share(hash_id)
+    if not info:
+        return HTMLResponse(f'<meta http-equiv="refresh" content="0;url={SPACE_URL}">')
+    title = info["title"]
+    img = info.get("img") or "https://s1.vnecdn.net/vnexpress/restruct/i/v9505/logo_default.jpg"
+    hash_url = f"{SPACE_URL}/#/{slug}/{hash_id}"
+    return HTMLResponse(f'''<!DOCTYPE html><html><head><meta charset="utf-8">
+<title>{title}</title>
+<meta property="og:title" content="{title}">
+<meta property="og:image" content="{img}">
+<meta property="og:url" content="{hash_url}">
+<meta property="og:type" content="article">
+<meta property="og:site_name" content="VNEWS">
+<meta name="twitter:card" content="summary_large_image">
+<meta name="twitter:title" content="{title}">
+<meta name="twitter:image" content="{img}">
+</head><body><script>localStorage.setItem("pending_article",'{json.dumps(info["url"])}');location.href="{SPACE_URL}";</script></body></html>''')
+
+# ═══════════════════════════════════════════════════════════════
+# SCRAPERS (unchanged from working version)
+# ═══════════════════════════════════════════════════════════════
+_cache={}; _cache_ttl=300
+def _cached(key,fn):
     now=time.time()
     if key in _cache and now-_cache[key]["t"]<_cache_ttl:return _cache[key]["d"]
     try:data=fn()
@@ -36,7 +148,7 @@ def scrape_vne(cat_url):
             t=a.get("title","") or a.get_text(strip=True);lk=a.get("href","")
             if not t or not lk:continue
             im=it.find("img");img=(im.get("data-src") or im.get("src","")) if im else ""
-            if img and'blank'in img:
+            if img and 'blank' in img:
                 src=it.find("source")
                 if src:img=src.get("srcset","").split(",")[0].strip().split(" ")[0]
             desc=it.select_one("p.description")
@@ -107,10 +219,7 @@ def scrape_24h_highlights():
             arts.append({"title":title,"link":href,"img":img_src,"source":"24h"})
         def _og(art):
             if art["img"]:return
-            try:
-                r=requests.get(art["link"],headers=HEADERS,timeout=8);r.encoding="utf-8"
-                og=BeautifulSoup(r.text,"lxml").find("meta",property="og:image")
-                if og:art["img"]=og.get("content","")
+            try:r=requests.get(art["link"],headers=HEADERS,timeout=8);r.encoding="utf-8";og=BeautifulSoup(r.text,"lxml").find("meta",property="og:image");art["img"]=og.get("content","") if og else ""
             except:pass
         need=[a for a in arts if not a["img"]][:10]
         if need:
@@ -128,8 +237,7 @@ def scrape_24h_shorts():
             href=a.get("href","")
             if not href.startswith("http"):href=BASE_24H+href
             if href in seen:continue
-            img_tag=art.find("img")
-            title=(img_tag.get("alt","") if img_tag else "") or a.get("title","") or a.get_text(strip=True)
+            img_tag=art.find("img");title=(img_tag.get("alt","") if img_tag else "") or a.get("title","") or a.get_text(strip=True)
             if not title or len(title)<10:continue
             img_src=""
             if img_tag:img_src=img_tag.get("data-original") or img_tag.get("data-src") or img_tag.get("src","")
@@ -162,8 +270,7 @@ def scrape_bbc_vietnamese():
             if not href.startswith("http"):href="https://www.bbc.com"+href
             if href in seen:continue
             title=a.get_text(strip=True)
-            if not title or len(title)<15:continue
-            if any(x in title.lower() for x in["đăng nhập","trang chủ","bbc news"]):continue
+            if not title or len(title)<15 or any(x in title.lower() for x in["đăng nhập","trang chủ","bbc news"]):continue
             img="";container=a.parent
             for _ in range(3):
                 if container:
@@ -184,9 +291,6 @@ def scrape_bbc_article(url):
         for p in soup.select("[data-component='text-block'] p, article p, main p"):
             t=p.get_text(strip=True)
             if t and len(t)>20:body.append({"type":"p","text":t})
-        for img in soup.select("main img, article img"):
-            src=img.get("src","")
-            if src and("ichef" in src or"bbci" in src):body.append({"type":"img","src":src,"alt":img.get("alt","")})
         return{"title":h1.get_text(strip=True) if h1 else "","summary":"","og_image":og_img,"body":body,"source":"bbc","url":url}
     except:return None
 
@@ -199,8 +303,7 @@ def extract_video_url(article_url):
         primary=full or p720
         if not primary:return None
         soup=BeautifulSoup(r.text,"lxml");og=soup.find("meta",property="og:image");poster=og.get("content","") if og else ""
-        base_url=primary[0];parts=[base_url]
-        is_720p='_720p.m3u8' in base_url;name=base_url.replace('_720p.m3u8','').replace('.m3u8','')
+        base_url=primary[0];parts=[base_url];is_720p='_720p.m3u8' in base_url;name=base_url.replace('_720p.m3u8','').replace('.m3u8','')
         m=re.search(r'(\D)(\d{1,2})$',name)
         if m:
             cur=int(m.group(2));w=len(m.group(2));bn=name[:m.start(2)]
@@ -238,7 +341,9 @@ def extract_bdp_video(url):
         return{"src":source.get("src","") if source else "","poster":video.get("poster","")}
     except:return None
 
-# API
+# ════���══════════════════════════════════════════════════════════
+# API Routes
+# ═══════════════════════════════════════════════════════════════
 VNE_CATS={"thoi-su":("https://vnexpress.net/thoi-su","Thời Sự"),"the-gioi":("https://vnexpress.net/the-gioi","Thế Giới"),"kinh-doanh":("https://vnexpress.net/kinh-doanh","Kinh Doanh"),"cong-nghe":("https://vnexpress.net/so-hoa","Công Nghệ"),"the-thao":("https://vnexpress.net/the-thao","Thể Thao"),"giai-tri":("https://vnexpress.net/giai-tri","Giải Trí"),"suc-khoe":("https://vnexpress.net/suc-khoe","Sức Khỏe")}
 BDP_CATS={"ngoai-hang-anh":("https://bongdaplus.vn/ngoai-hang-anh","Ngoại Hạng Anh"),"la-liga":("https://bongdaplus.vn/la-liga","La Liga"),"champions-league":("https://bongdaplus.vn/champions-league-cup-c1","Champions League"),"bong-da-vn":("https://bongdaplus.vn/bong-da-viet-nam","Bóng Đá VN")}
 
@@ -281,14 +386,12 @@ def api_highlights():return JSONResponse(_cached("highlights",scrape_24h_highlig
 def api_shorts():return JSONResponse(_cached("shorts",scrape_24h_shorts))
 @app.get("/api/bdp_videos")
 def api_bdp_videos():return JSONResponse(_cached("bdp_videos",scrape_bdp_videos))
-
 @app.get("/api/video_url")
 def api_video_url(url:str=Query(...)):
     if"24h.com.vn" in url:v=extract_video_url(url)
     elif"bongdaplus.vn" in url:v=extract_bdp_video(url)
     else:v=None
     return JSONResponse(v if v else{"error":"not found"})
-
 @app.get("/api/article")
 def api_article(url:str=Query(...)):
     if"vnexpress.net" in url:data=scrape_vne_article(url)
@@ -296,28 +399,8 @@ def api_article(url:str=Query(...)):
     else:data=None
     return JSONResponse(data if data else{"error":"not supported"})
 
-# Share page: SEO with og:image + redirect to article
-@app.get("/s")
-async def share_redirect(url:str=Query(default=""),title:str=Query(default="VNEWS"),img:str=Query(default="")):
-    """Share URL with proper og:image. Format: /s?url=ENCODED_ARTICLE_URL&title=...&img=..."""
-    og_image=unquote(img) if img else "https://s1.vnecdn.net/vnexpress/restruct/i/v9505/logo_default.jpg"
-    decoded_url=unquote(url)
-    redirect_script=f'<script>localStorage.setItem("pending_article","{decoded_url}");location.href="{SPACE_URL}";</script>' if decoded_url else f'<script>location.href="{SPACE_URL}";</script>'
-    return HTMLResponse(f'''<!DOCTYPE html><html><head>
-<meta charset="utf-8"><title>{unquote(title)}</title>
-<meta property="og:title" content="{unquote(title)}">
-<meta property="og:image" content="{og_image}">
-<meta property="og:url" content="{SPACE_URL}/s?url={url}&title={title}&img={img}">
-<meta property="og:type" content="article">
-<meta property="og:site_name" content="VNEWS">
-<meta property="og:description" content="{unquote(title)} - Đọc trên VNEWS">
-<meta name="twitter:card" content="summary_large_image">
-<meta name="twitter:title" content="{unquote(title)}">
-<meta name="twitter:image" content="{og_image}">
-</head><body>{redirect_script}</body></html>''')
-
 @app.get("/")
 async def index():
-    with open("/app/static/index.html","r",encoding="utf-8") as f:return HTMLResponse(content=f.read())
+    return FileResponse("/app/static/index.html")
 
 app.mount("/static",StaticFiles(directory="/app/static"),name="static")