Fix rewrite URL, short JSON error, add missing endpoints, wc2026 timeout

ai_ext.py

@@ -197,4 +197,136 @@ def _fallback_summary_from_prompt(prompt: str, max_units: int = 6) -> str:
                     break
         if chunks:
             return "\n".join("• " + c for c in chunks)
-    return "• Không có đủ nội dung để tóm tắt."
+    return "• Không có đủ nội dung để tóm tắt."
+
+# ===== URL scraping & article processing =====
+HEADERS = {"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36", "Accept-Language": "vi-VN,vi;q=0.9,en;q=0.8"}
+
+try:
+    _shorts_base = "/data" if os.path.isdir("/data") else os.path.join(os.path.dirname(os.path.abspath(__file__)), "data")
+except Exception:
+    _shorts_base = os.path.join(os.path.dirname(os.path.abspath(__file__)), "data")
+SHORTS_DIR = os.path.join(_shorts_base, "ai_shorts")
+os.makedirs(SHORTS_DIR, exist_ok=True)
+
+import random as _random2
+from datetime import datetime, timezone, timedelta
+_VN_TZ = timezone(timedelta(hours=7))
+
+
+def _safe_name(filename: str) -> str:
+    """Sanitize filename."""
+    return re.sub(r"[^a-zA-Z0-9_.-]", "_", filename)[:120]
+
+
+def pollinations_image_url(topic: str) -> str:
+    """Generate a placeholder image URL via Pollinations."""
+    try:
+        return "https://image.pollinations.ai/prompt/" + quote("Vietnamese editorial illustration, " + topic, safe="") + "?width=1024&height=576&nologo=true"
+    except Exception:
+        return ""
+
+
+def _download_image(url: str, fallback_title: str, out_path: str) -> str:
+    """Download an image from URL or create a placeholder."""
+    if url:
+        try:
+            r = requests.get(url, headers=HEADERS, timeout=15)
+            if r.status_code == 200 and len(r.content) > 1200:
+                os.makedirs(os.path.dirname(out_path), exist_ok=True)
+                with open(out_path, "wb") as f:
+                    f.write(r.content)
+                return out_path
+        except Exception:
+            pass
+    # Fallback: create a placeholder image
+    try:
+        from PIL import Image, ImageDraw, ImageFont
+        img = Image.new("RGB", (1080, 760), (24, 24, 24))
+        draw = ImageDraw.Draw(img)
+        try:
+            font = ImageFont.truetype("/usr/share/fonts/truetype/dejavu/DejaVuSans-Bold.ttf", 48)
+        except Exception:
+            font = None
+        text = (fallback_title or "VNEWS")[:40]
+        try:
+            bbox = draw.textbbox((0, 0), text, font=font)
+            tw = bbox[2] - bbox[0]
+        except Exception:
+            tw = len(text) * 24
+        draw.text(((1080 - tw) // 2, 330), text, fill=(255, 255, 255), font=font)
+        os.makedirs(os.path.dirname(out_path), exist_ok=True)
+        img.save(out_path, quality=90)
+        return out_path
+    except Exception:
+        return out_path
+
+
+def scrape_any_url(url: str) -> dict:
+    """Scrape article content from any URL."""
+    if not url or not url.startswith("http"):
+        return {"title": "", "text": "", "summary": "", "image": "", "og_image": "", "via": ""}
+    try:
+        r = requests.get(url, headers=HEADERS, timeout=15, allow_redirects=True)
+        if r.status_code != 200 or not r.text:
+            return {"title": "", "text": "", "summary": "", "image": "", "og_image": "", "via": _domain(url)}
+        r.encoding = "utf-8"
+        soup = BeautifulSoup(r.text, "lxml")
+        for tag in soup.find_all(["script", "style", "nav", "footer", "aside", "form", "noscript", "iframe", ".ads", ".ad", ".banner-ads", ".fb-comments", ".fb-root", ".social-share", ".related-news", ".breadcrumb"]):
+            tag.decompose()
+        title = ""
+        ogt = soup.find("meta", property="og:title")
+        if ogt:
+            title = ogt.get("content", "")
+        h1 = soup.find("h1")
+        if not title and h1:
+            title = h1.get_text(strip=True)
+        if not title:
+            t = soup.find("title")
+            if t:
+                title = t.get_text(strip=True)
+        og_image = ""
+        ogi = soup.find("meta", property="og:image")
+        if ogi:
+            og_image = ogi.get("content", "")
+            if og_image.startswith("//"):
+                og_image = "https:" + og_image
+        summary = ""
+        ogd = soup.find("meta", property="og:description") or soup.find("meta", attrs={"name": "description"})
+        if ogd:
+            summary = ogd.get("content", "")[:500]
+        body_text = []
+        for sel in ["article", ".singular-content", ".detail-content", ".fck_detail", ".content-detail", ".knc-content", "main", ".cms-body", ".article__body", ".post-content", ".entry-content"]:
+            el = soup.select_one(sel)
+            if el and len(el.find_all("p")) >= 2:
+                for p in el.find_all("p"):
+                    t = _clean_text(p.get_text(strip=True))
+                    if t and len(t) > 30:
+                        body_text.append(t)
+                break
+        if not body_text and soup.body:
+            for p in soup.body.find_all("p"):
+                t = _clean_text(p.get_text(strip=True))
+                if t and len(t) > 30:
+                    body_text.append(t)
+        text = "\n".join(body_text)
+        return {"title": _clean_text(title), "text": text, "summary": _clean_text(summary), "image": og_image, "og_image": og_image, "via": _domain(url), "url": url}
+    except Exception as e:
+        return {"title": "", "text": "", "summary": "", "image": "", "og_image": "", "via": _domain(url)}
+
+
+def make_post(title: str, text: str, img: str, url: str, kind: str = "auto", sources: list = None) -> dict:
+    """Create a wall post dict."""
+    import random as _r2
+    now = int(time.time() * 1000)
+    return {
+        "id": str(now) + str(_r2.randint(100, 999)),
+        "title": (title or "Bài viết")[:200],
+        "text": (text or "")[:5000],
+        "img": img or "",
+        "url": url or "",
+        "kind": kind or "auto",
+        "sources": sources or [],
+        "created": now,
+        "created_str": datetime.now(_VN_TZ).strftime("%H:%M %d/%m/%Y"),
+    }

app_v2_entry.py

@@ -226,7 +226,7 @@ _STOP=set('và của các những một được trong với cho tại sau trư
 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]
+    words=[w for w in re.findall(r'[A-Za-zÀ-ỹ0-9]+',tl) if len(w)>2 and w.lower() not in _STOP]
     if not words:return True
     return any(w in tt for w in words)
 
@@ -859,6 +859,31 @@ def _save_wall_posts(posts):
     with _wl_lock:
         _sj(WALL_FILE, posts)
 
+@app.get("/api/ai/wall")
+def api_ai_wall():
+    return JSONResponse({"posts": _load_wall_posts()})
+
+@app.post("/api/ai/short/{post_id}")
+def api_ai_short(post_id: str):
+    """Generate a short video for a wall post.
+    Returns clear JSON error when video not available yet.
+    """
+    post_id_s = str(post_id)
+    posts = _load_wall_posts()
+    if not isinstance(posts, list):
+        posts = []
+    post = None
+    for p in posts:
+        pid = str(p.get("id", ""))
+        if pid == post_id_s or pid.startswith(post_id_s):
+            post = p
+            break
+    if not post:
+        return JSONResponse({"error": "Không tìm thấy bài viết", "post_id": post_id}, status_code=404)
+    if post.get("video"):
+        return JSONResponse({"video": post["video"], "post": post})
+    return JSONResponse({"error": "Chưa có video cho bài này. Vui lòng upload video trước."}, status_code=409)
+
 @app.get('/api/wall')
 def api_wall():
     posts = _load_wall_posts()