More robust handling of the Content-Type header for thumbnail generation

Signed-off-by: Sergey Shpikin <rkfg@rkfg.me>

changelog.d/9788.bugfix

@@ -0,0 +1 @@
+Fix thumbnail generation for some sites with non-standard content types.

synapse/config/repository.py

@@ -71,6 +71,7 @@ def parse_thumbnail_requirements(thumbnail_sizes):
         jpeg_thumbnail = ThumbnailRequirement(width, height, method, "image/jpeg")
         png_thumbnail = ThumbnailRequirement(width, height, method, "image/png")
         requirements.setdefault("image/jpeg", []).append(jpeg_thumbnail)
+        requirements.setdefault("image/jpg", []).append(jpeg_thumbnail)
         requirements.setdefault("image/webp", []).append(jpeg_thumbnail)
         requirements.setdefault("image/gif", []).append(png_thumbnail)
         requirements.setdefault("image/png", []).append(png_thumbnail)

synapse/rest/media/v1/preview_url_resource.py

@@ -513,6 +513,10 @@ async def _download_url(self, url: str, user: str) -> Dict[str, Any]:
 
                 if b"Content-Type" in headers:
                     media_type = headers[b"Content-Type"][0].decode("ascii")
+                    if media_type.startswith("image/"):
+                        scpos = media_type.find(";")
+                        if scpos > 0:
+                            media_type = media_type[:scpos]
                 else:
                     media_type = "application/octet-stream"