[pixiv] Series with 18+ works isn't getting fully downloaded

[SunSerega]

In #2964:

you might need to supply your cookies for series with R18 works, haven't tested that yet. Your example downloads fine with just the usual Pixiv refresh token.

I've just created a series to reproduce more easily:

gallery-dl.exe -d "some_dir" --cookies-from-browser opera "https://www.pixiv.net/user/38349769/series/216957"

This command only downloads the first work, while the second one (which I arbitrarily marked 18+) is not seen by gallery-dl.
If every work in a series is 18+, I get this message:

[pixiv][info] No results for <URL>

[thatfuckingbird]

Can confirm, seeing the following behavior with refresh token + cookies:

1. The series link from OP only gets the first file.
2. If I download their user gallery, then it gets both (further confirming it's not a problem with r18 but specifically something with series links).

[mikf]

This is my fault. Cookies are generally disabled for Pixiv (ce54b8c) and I didn't re-enable them for series. And even with them enabled, you get a Cloudflare challenge. Thankfully -o browser=firefox -o tls12=false works, at least for me.

[SunSerega]

-o tls12=false

Are there any docs for this?

[mikf]

Now that you mention it, no there isn't any.
You won't have to use it manually, though. This will all be set automatically by the changes in commit dd14adc, which hasn't been git pushed yet.

edit: fixed, but you still need to provide logged in cookies.

[SunSerega]

Hmm, this workflow didn't seem to run even though you pushed that commit to master. And neither did the test workflow.

[SunSerega]

Ah, nevermind, I can swear I saw GitHub show this commit was 1 hour old, while workflow run was 2 hours old. But now I see the commit is older.

Regardless, everything works with the latest binaries. Thanks for fixing it so quickly.

Can I also delete the pixiv series I created for this issue? Can keep it if you need it as a consistent test reference or something...

[mikf]

Can I also delete the pixiv series I created for this issue?

Yeah, that's fine.

[Sherman-Liu]

@mikf Hi, I've already provided refresh token in the config file, and cookies in the cookie file, for some unknown reason, it's still not working for me, example:

I:\Pictures\Sparrowl>gallery-dl https://www.pixiv.net/artworks/101636041 -v
[gallery-dl][debug] Version 1.26.5
[gallery-dl][debug] Python 3.11.5 - Windows-10-10.0.19045-SP0
[gallery-dl][debug] requests 2.31.0 - urllib3 2.1.0
[gallery-dl][debug] Configuration Files ['%USERPROFILE%\\gallery-dl.conf']
[gallery-dl][debug] Starting DownloadJob for 'https://www.pixiv.net/artworks/101636041'
[pixiv][debug] Using PixivWorkExtractor for 'https://www.pixiv.net/artworks/101636041'
[urllib3.connectionpool][debug] Starting new HTTPS connection (1): app-api.pixiv.net:443
[urllib3.connectionpool][debug] https://app-api.pixiv.net:443 "GET /v1/illust/detail?illust_id=101636041 HTTP/1.1" 200 396
[postprocessor.ugoira][debug] using concat demuxer
[pixiv][debug] Active postprocessor modules: [UgoiraPP]
[pixiv][debug] Skipping 'sanity_level' warning (101636041) 

It's been working fine for a very long time, but start from someday it just suddenly doesn't retrieve any images, I tried to get a new refresh token, but it's still not working properly.

The config file (cropped):

        "pixiv": {
            "filename": "{id}_{num}_{title}.{extension}",
            "directory": [
                "{category}",
                "{user[name]}"
            ],
            "refresh-token": "_____",
            "tags": "translated",
            "ugoira": true
        },