filename from the redirected URL

[AlttiRi]

Is it possible to parse filename/ext from the final URL after all redirects?

For example, the mastodon extractor have the follow bug when API sometimes returns a media URL with /media_proxy/ (when I download files from instance b by using instance a (https://a.org/@[email protected])):

urllib3.connectionpool: https://a.org:443 "GET /media_proxy/111222333444555666/original HTTP/11" 302 0
urllib3.connectionpool: https://media.a.org:443 "GET /cache/media_attachments/files/111/222/333/444/555/666/original/a1b2c3d4e5f6.jpg HTTP/11" 200 123456

In this case the filename key is original, while it should be a1b2c3d4e5f6.

[mikf]

No, that's (currently) not possible.
The redirect happens at downloader level, which is too late for any proper metadata extraction and handling.

You could give metadata-http a try, but that's a buggy mess.

[AlttiRi]

gdl_http['filename']/gdl_http[filename] is None with added "metadata-http": "http",.