Files
danbooru/app/logical/source/extractor/gelbooru.rb
evazion c19fc16885 sources: don't escape Unicode characters in tag search URLs.
Fix it so that Unicode characters aren't unnecessarily percent-encoded when generating tag search
URLs. For example, generate URLs like this:

* https://www.pixiv.net/tags/オリジナル/artworks

Not like this:

* https://www.pixiv.net/tags/%E3%82%AA%E3%83%AA%E3%82%B8%E3%83%8A%E3%83%AB/artworks
2022-12-02 16:35:49 -06:00

84 lines
2.6 KiB
Ruby

# frozen_string_literal: true
# Source extractor for Gelbooru-based sites, including Gelbooru, Safebooru.org, TBIB.org, and Rule34.xxx. The commentary
# and artist information are pulled from the booru post's source, while the translated tags include both the booru tags
# and the source's tags.
#
# @see Source::URL::Gelbooru
# @see https://gelbooru.com/index.php?page=wiki&s=view&id=18780 (howto:api)
# @see https://safebooru.org/index.php?page=help&topic=dapi
# @see https://tbib.org/
# @see https://rule34.xxx/index.php?page=help&topic=dapi
module Source
class Extractor
class Gelbooru < Source::Extractor
delegate :artist_name, :profile_url, :profile_urls, :other_names, :tag_name, :artist_commentary_title, :artist_commentary_desc, :dtext_artist_commentary_title, :dtext_artist_commentary_desc, to: :sub_extractor, allow_nil: true
def match?
Source::URL::Gelbooru === parsed_url
end
def image_urls
if parsed_url.full_image_url.present?
[parsed_url.full_image_url]
else
[api_response[:file_url]].compact
end
end
def page_url
"https://#{domain}/index.php?page=post&s=view&id=#{post_id}" if post_id.present?
end
def tags
site_tags + source_tags
end
def site_tags
return [] if api_response.blank?
tags = api_response[:tags].split + ["rating:#{api_response[:rating]}"]
tags.map do |tag|
[tag, "https://#{domain}/index.php?page=post&s=list&tags=#{Danbooru::URL.escape(tag)}"]
end
end
def source_tags
sub_extractor&.tags.to_a
end
def other_names
sub_extractor&.other_names.to_a
end
def domain
parsed_url.domain
end
def post_id
parsed_url.post_id || parsed_referer&.post_id || api_response[:id]
end
def api_url
# https://gelbooru.com//index.php?page=dapi&s=post&q=index&tags=md5:338078144fe77c9e5f35dbb585e749ec
# https://gelbooru.com//index.php?page=dapi&s=post&q=index&tags=id:7903922
parsed_url.api_url || parsed_referer&.api_url
end
memoize def api_response
return {} unless api_url.present?
response = http.cache(1.minute).get(api_url)
return {} unless response.status == 200
response.parse.dig("posts", "post").to_h.with_indifferent_access
end
def sub_extractor
return nil if !api_response[:source].to_s.match?(%r{\Ahttps?://}i)
@sub_extractor ||= Source::Extractor.find(api_response[:source], default: nil)
end
end
end
end