How do I filter Pixiv content?

[Roccobot]

Sometimes I like an artist so much I download their entire gallery. NSFW is okay, but I don't want 'extreme' stuff so I tried to add a filter to my gallery-dl.config, but looks I'm getting it completely wrong.

I just tried to add a line to the pixiv extractor:
"filter": "tag not in ('beastiality', 'guro', 'scat')"

but all I get is an error while launching the script:
(NameError: name 'tag' is not defined)

Could anybody help me with this? Thanks a lot

[mikf]

"not contains(tags, ('beastiality', 'guro', 'scat'))"

There are other, more Pythonic ™️ methods, but they are either a lot more verbose or do not work due to how filter expressions are implemented in gallery-dl.

Be sure to use translated tags (-o tags=translated), otherwise you will have to deal mostly Japanese tags.

[Roccobot]

@mikf collateral question: if a tag contains the character ' (apostrophe), should I escape it like \'?

[mikf]

It seems you have to backslash escape it twice, once for the JSON parser and again for the Python parser:

    "image-filter": "not contains(tags, ('foo\\'s', 'bar\\'s'))"

so foo\\'s -> foo's, etc
(I wasn't sure myself, so I played around with it until it worked ...)

It should also be possible to use triple quotes inside the Python expression:

    "image-filter": "not contains(tags, ('''foo's''', '''bar's'''))"

[Roccobot]

Sorry to bother you again @mikf but until yeesterday the filter was working. Do you know if I need to escape spaces as well?

[mikf]

You do not need to escape spaces or really anything other than '.

contains(…) only does exact matches, so maybe there's an extra character here or there, or the case does not match (Foo != foO etc), but I can't really help you without an example.

[Roccobot]

aaaah, as usual I'm too relaxed with uppercase/lowercase, I probably didn't consider the filter is case sensitive. Thanks again!