[gelbooru] regex for extractor not allowing alternate parameter order in url.

[KJNeko]

The regex currently implemented wouldn't support urls where the s, page, and id orders were different.

placing page=post and s=view into look-aheads from the ? forward allows for any order while also checking if present.

URLs tested

1: https://gelbooru.com/index.php?page=post&s=view&id=7583325
2: https://gelbooru.com/index.php?page=post&s=view&id=7583325&tags=cat
3: https://gelbooru.com/index.php?id=7583325&page=post&s=view

Match 2 was gotten by searching for 'cat' and then clicking on an image. Seems gelbooru embeds search parameters

Old regex

(?:https?://)?(?:www\.)?gelbooru\.com/(?:index\.php)?\?page=post&s=view&id=(?P<post>\d+)

Results:
1: Match: https://gelbooru.com/index.php?page=post&s=view&id=7586232, Group post: 7586232
2: Match: https://gelbooru.com/index.php?page=post&s=view&id=7586232, Group post: 7586232
3: No match

New regex

(?:https?:\/\/)?(?:www\.)?gelbooru\.com\/(?:index\.php)?\?(?=.*page=post)(?=.*s=view).*id=(?P<post>\d+).*

Results:
1: Match: https://gelbooru.com/index.php?page=post&s=view&id=7586232, Group post: 7586232
2: Match: https://gelbooru.com/index.php?page=post&s=view&id=7582526&tags=cat, Group post 7586232
3: Match: https://gelbooru.com/index.php?id=7586232&page=post&s=view, Group post: 7586232

I was unable to determine if the old regex was made with the intention of sanitizing additional arguments ( such as removing &tag=cat) or if it was just a byproduct or a 'bonus' of the regex

[thatfuckingbird]

@mikf Can you take a look at this sometime? I can confirm this parameter order issue. I think SauceNAO and possibly some other sites return such rearranged URLs sometimes, at least I semi-regularly encounter them.

[mikf]

Using regular expressions to "parse" URL query strings is generally a bad idea, as is using .* everywhere, which is why I kind of didn't want to merge this.
And then there is the "problem" that this only fixes the issue for post URLs, and only for Gelbooru.