[atfforum/xenforo] Support user media tab

[BARNvq]

Some users post content as reference to the "media" tab on their profile.

/forum/index.php?members/<username>.<integer>/
/forum/index.php?members/<username>.<integer>/#xfmgMedia

Downloading images from this tab is not supported, when using either the media tab URL or linking directly to a specific image with the format below.

/forum/index.php?media/<image-name>.<integer>/
/forum/index.php?media/<image-name>.<integer>/full

gallery-dl --version
1.31.1

[mikf]

The "actual" URL for these files appears to be

/forum/index.php?media/users/<username>.<integer>/

[BARNvq]

Both /forum/index.php?members/<username>.<integer>/#xfmgMedia or /forum/index.php?media/users/<username>.<integer>/ lead to a similar media overview, but the individual images are still located at /forum/index.php?media/<image-name>.<integer>/full regardless of which overview you use to navigate to them. None of these three URLs are supported.

[mikf]

60104f2 supports all three mentioned URL types

Enable metadata if you want "full" metadata.

[BARNvq]

Excellent, normal downloading works perfectly. I noticed however that -o metadata=true returns 403 (at least on atfforum). This is because the URL used to extract the metadata contains "forum" twice. Removing one should fix it.

  File "/usr/lib/python3.13/site-packages/gallery_dl/extractor/xenforo.py", line 263, in _extract_media_ex
    page = self.request(f"{self.root}{path}/").text

Current:
/forum/forum/index.php?media/<image-name>.<integer>/
Fixed:
/forum/index.php?media/<image-name>.<integer>/

[mikf]

This is because the URL used to extract the metadata contains "forum" twice. Removing one should fix it.

32e95f4

I've also added support for category listings: bb7d540

[BARNvq]

Thanks, everything seems to be working now. Categories too.
One more thing I noticed regarding media gallery downloads is that for users with a capitalized name (let's say "BoB") the author field is bob when metadata=false, but BoB when metadata=true. This results in separate directories ("BoB" or "bob") depending on if metadata is on, although I am not sure if this is fixable.

[hrxn]

Either add formatting conversions to those specific fields, or nuke it all directly from orbit:
https://gdl-org.github.io/docs/configuration.html#extractor-path-convert

[mikf]

The author name with "metadata": false is just the URL slug since that's all that's available without any extra HTTP requests, while in any other situations it is the full name of a user.

We could add a user_slug field to have a consistent name format, since the slug (as well as the ID) is always available. Or you use {author!g} and convert to to a slug as HRXN suggested.

[BARNvq]

The user_slug is probably a good idea to keep the author field consistent, but I just added the overwrite for now. Thank you both.

"directory": {
	"'media-user' in subcategory": ["{category}","Media","{author!g}"]
},