v1.0定版

backend/app/crawler/rss_fetcher.py

@@ -25,6 +25,31 @@ def _parse_date(raw: str) -> Optional[datetime]:
         return None
 
 
+def _extract_image(entry) -> Optional[str]:
+    """Try to pull an image URL from common RSS media extensions."""
+    # <media:thumbnail>
+    thumbnails = getattr(entry, "media_thumbnail", [])
+    if thumbnails:
+        url = thumbnails[0].get("url", "").strip()
+        if url:
+            return url
+    # <media:content medium="image">
+    for mc in getattr(entry, "media_content", []):
+        mc_type = mc.get("type", "")
+        mc_medium = mc.get("medium", "")
+        if mc_medium == "image" or mc_type.startswith("image/"):
+            url = mc.get("url", "").strip()
+            if url:
+                return url
+    # <enclosure type="image/...">
+    for enc in getattr(entry, "enclosures", []):
+        if enc.get("type", "").startswith("image/"):
+            url = (enc.get("href") or enc.get("url") or "").strip()
+            if url:
+                return url
+    return None
+
+
 async def fetch_rss(url: str, max_items: int = 30) -> list[dict]:
     try:
         async with httpx.AsyncClient(headers=HEADERS, timeout=30, follow_redirects=True) as client:
@@ -54,6 +79,7 @@ async def fetch_rss(url: str, max_items: int = 30) -> list[dict]:
             "url": link,
             "content": content[:3000],
             "published_at": _parse_date(published_raw),
+            "image_url": _extract_image(entry),
         })
 
     logger.info(f"RSS {url}: got {len(items)} items")