feat: import F95 screenshot galleries with AVIF support

backend/app/f95.py

@@ -82,7 +82,7 @@ def fetch_image(url: str, thread_url: str, cookie: str = "", user_agent: str = "
     except urllib.error.HTTPError as e:
         if e.code in (401, 403):
             raise F95AuthError("F95zone не принял cookie или ограничил доступ к превью") from e
-        raise F95Error("Не удалось скачать превью F95zone") from e
+        raise F95Error(f"F95zone не отдал изображение (HTTP {e.code})") from e
     except OSError as e:
         raise F95Error("Не удалось скачать превью F95zone") from e
 
@@ -133,6 +133,8 @@ class _Parser(HTMLParser):
         self.og_image = None
         self.post_image = None
         self.cover_image = None
+        self.gallery_urls: list[str] = []
+        self._gallery_paths: set[str] = set()
         self.tags: list[str] = []
         self.post: list[str] = []
         self.template = ""
@@ -157,6 +159,14 @@ class _Parser(HTMLParser):
             self._in_tag, self._tag_buf = True, ""
         if self._done:
             return
+        # Screenshot lightboxes are often inside collapsed spoilers. Their href is
+        # the full-sized attachment; img src/data-src below it is only a thumbnail.
+        if self._depth and tag == "a" and "js-lbImage" in cls:
+            if candidate := image_url(a.get("href") or ""):
+                identity = urlparse(candidate)._replace(query="", fragment="").geturl()
+                if identity not in self._gallery_paths:
+                    self._gallery_paths.add(identity)
+                    self.gallery_urls.append(candidate)
         if self._depth == 0:
             if tag == "div" and "bbWrapper" in cls and not self.post:
                 self._depth = 1
@@ -215,8 +225,12 @@ def parse(html: str) -> dict:
     # "VN - Ren'Py - Hooked! [v0.6.0] [BlenderKnight]" -> "Hooked!"
     title = re.sub(r"\s*\[[^\]]*\]", "", title)
     title = re.sub(r"^(?:[A-Za-z0-9'’ ]{1,20} - )+", "", title).strip()
+    cover = p.cover_image or p.post_image or p.og_image
+    cover_path = urlparse(cover)._replace(query="", fragment="").geturl() if cover else None
+    gallery = [url for url in p.gallery_urls
+               if urlparse(url)._replace(query="", fragment="").geturl() != cover_path]
     return {"title": title, "description": _description("".join(p.post)),
-            "tags": sorted(set(p.tags)), "image_url": p.cover_image or p.post_image or p.og_image,
+            "tags": sorted(set(p.tags)), "image_url": cover, "gallery_urls": gallery,
             "logged_in": p.logged_in}