sources: add Gelbooru support.

Add support for uploading posts from Gelbooru. Note that the translated
tags will include both the Gelbooru tags and the tags from the Gelbooru
post's source. The commentary and artist information will also be taken
from the Gelbooru post's source. The source of the Danbooru post however
will be left as the Gelbooru post itself, not as the Gelbooru post's source.

app/logical/source/extractor.rb

@@ -39,6 +39,7 @@ module Source
       Source::Extractor::Moebooru,
       Source::Extractor::Nijie,
       Source::Extractor::ArtStation,
+      Source::Extractor::Gelbooru,
       Source::Extractor::HentaiFoundry,
       Source::Extractor::Fanbox,
       Source::Extractor::Mastodon,

app/logical/source/extractor/gelbooru.rb

@@ -0,0 +1,76 @@
+# frozen_string_literal: true
+
+# Source extractor for Gelbooru. The commentary and artist information are
+# pulled from the Gelbooru post's source, while the translated tags include
+# both the Gelbooru tags and the source's tags.
+#
+# @see Source::URL::Gelbooru
+# @see https://gelbooru.com/index.php?page=wiki&s=view&id=18780 (howto:api)
+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
+        [api_response[:file_url]].compact
+      end
+
+      def page_url
+        "https://gelbooru.com/index.php?page=post&s=view&id=#{post_id}" if post_id.present?
+      end
+
+      def tags
+        gelbooru_tags + source_tags
+      end
+
+      def gelbooru_tags
+        return [] if api_response.nil?
+
+        tags = api_response[:tags].split + ["rating:#{api_response[:rating]}"]
+        tags.map do |tag|
+          [tag, "https://gelbooru.com/index.php?page=post&s=list&tags=#{CGI.escape(tag)}"]
+        end
+      end
+
+      def source_tags
+        sub_extractor&.tags.to_a
+      end
+
+      def post_id
+        parsed_url.post_id || parsed_referer&.post_id || post_id_from_md5
+      end
+
+      def api_url
+        # https://gelbooru.com/index.php?page=dapi&s=post&q=index&id=779812&json=1
+        "https://gelbooru.com/index.php?page=dapi&s=post&q=index&id=#{post_id}&json=1" if post_id.present?
+      end
+
+      memoize def api_response
+        return nil unless api_url.present?
+
+        response = http.cache(1.minute).get(api_url)
+        return nil unless response.status == 200
+
+        response.parse["post"].first.with_indifferent_access
+      end
+
+      memoize def post_id_from_md5
+        return nil unless parsed_url.image_url? && parsed_url.page_url.present?
+
+        response = http.cache(1.minute).head(parsed_url.page_url)
+        return nil unless response.status == 200
+
+        Source::URL.parse(response.uri).post_id
+      end
+
+      def sub_extractor
+        return nil if api_response[:source].nil?
+        @sub_extractor ||= Source::Extractor.find(api_response[:source], default: nil)
+      end
+    end
+  end
+end

app/logical/source/url.rb

@@ -40,6 +40,7 @@ module Source
       Source::URL::Fantia,
       Source::URL::Fc2,
       Source::URL::Foundation,
+      Source::URL::Gelbooru,
       Source::URL::HentaiFoundry,
       Source::URL::Instagram,
       Source::URL::Lofter,

app/logical/source/url/gelbooru.rb

@@ -0,0 +1,56 @@
+# frozen_string_literal: true
+
+class Source::URL::Gelbooru < Source::URL
+  attr_reader :post_id, :md5, :full_image_url
+
+  def self.match?(url)
+    url.domain.in?(%w[gelbooru.com])
+  end
+
+  def parse
+    case [domain, *path_segments]
+
+    # https://gelbooru.com/index.php?page=post&s=view&id=7798045
+    # https://www.gelbooru.com/index.php?page=post&s=view&id=7798045
+    in "gelbooru.com", "index.php" if params[:page] == "post" && params[:s] == "view" && params[:id].present?
+      @post_id = params[:id].to_i
+
+    # https://gelbooru.com/index.php?page=post&s=list&md5=99d9977d6c3aa185083a2da22bd8acfb
+    in "gelbooru.com", "index.php" if params[:page] == "post" && params[:s] == "list" && params[:md5].present?
+      @md5 = params[:md5]
+
+    # https://gelbooru.com/index.php?page=dapi&s=post&q=index&id=7798045&json=1
+    in "gelbooru.com", "index.php" if params[:page] == "dapi" && params[:q] == "index" && params[:id].present?
+      @post_id = params[:id].to_i
+
+    # https://gelbooru.com//images/ee/5c/ee5c9a69db9602c95debdb9b98fb3e3e.jpeg
+    # http://simg.gelbooru.com//images/2003/edd1d2b3881cf70c3acf540780507531.png
+    # https://simg3.gelbooru.com//samples/0b/3a/sample_0b3ae5e225072b8e391c827cb470d29c.jpg
+    # https://video-cdn3.gelbooru.com/images/62/95/6295154d082f04009160261b90e7176e.mp4
+    # https://img2.gelbooru.com//images/a9/64/a96478bbf9bc3f0584f2b5ddf56025fa.webm
+    # https://gelbooru.com/thumbnails/08/06/thumbnail_08066c138e7e138a47489a0934c29156.jpg
+    in "gelbooru.com", ("images" | "samples" | "thumbnails"), h1, h2, /\A(?:\w+_)?(\h{32})\.(jpeg|jpg|png|gif|mp4|webm)\z/i
+      @md5 = $1
+      @full_image_url = "https://#{host}/images/#{h1}/#{h2}/#{md5}.#{file_ext}"
+
+    else
+      nil
+    end
+  end
+
+  def image_url?
+    full_image_url.present?
+  end
+
+  def page_url
+    if post_id.present?
+      "https://gelbooru.com/index.php?page=post&s=view&id=#{post_id}"
+    elsif md5.present?
+      "https://gelbooru.com/index.php?page=post&s=list&md5=#{md5}"
+    end
+  end
+
+  def api_url
+    "https://gelbooru.com/index.php?page=dapi&s=post&q=index&id=#{post_id}&json=1" if post_id.present?
+  end
+end

app/logical/source/url/null.rb

@@ -131,15 +131,6 @@ class Source::URL::Null < Source::URL
       @work_id = $1
       @page_url = "https://www.facebook.com/photo.php?fbid=#{work_id}"
 
-    # https://gelbooru.com//images/ee/5c/ee5c9a69db9602c95debdb9b98fb3e3e.jpeg
-    # http://simg.gelbooru.com//images/2003/edd1d2b3881cf70c3acf540780507531.png
-    # https://simg3.gelbooru.com//samples/0b/3a/sample_0b3ae5e225072b8e391c827cb470d29c.jpg
-    # https://video-cdn3.gelbooru.com/images/62/95/6295154d082f04009160261b90e7176e.mp4
-    # https://img2.gelbooru.com//images/a9/64/a96478bbf9bc3f0584f2b5ddf56025fa.webm
-    in _, "gelbooru.com", ("images" | "samples"), *subdirs, /^(?:sample_)?(\h{32})\.(jpeg|jpg|png|gif|mp4|webm)$/
-      @md5 = $1
-      @page_url = "https://gelbooru.com/index.php?page=post&s=list&tags=md5:#{@md5}"
-
     # https://a.hitomi.la/galleries/907838/1.png
     # https://0a.hitomi.la/galleries/1169701/23.png
     # https://aa.hitomi.la/galleries/990722/003_01_002.jpg

test/functional/uploads_controller_test.rb

@@ -313,6 +313,8 @@ class UploadsControllerTest < ActionDispatch::IntegrationTest
         should_upload_successfully("https://konachan.com/post/show/270916")
         should_upload_successfully("https://konachan.com/image/ca12cdb79a66d242e95a6f958341bf05/Konachan.com%20-%20270916.png")
 
+        should_upload_successfully("https://gelbooru.com/index.php?page=post&s=view&id=7798121")
+
         should_upload_successfully("http://lohas.nicoseiga.jp/o/910aecf08e542285862954017f8a33a8c32a8aec/1433298801/4937663")
         should_upload_successfully("http://seiga.nicovideo.jp/seiga/im4937663")
         should_upload_successfully("https://seiga.nicovideo.jp/image/source/9146749")

test/unit/sources/gelbooru_test.rb

@@ -0,0 +1,58 @@
+require "test_helper"
+
+module Sources
+  class GelbooruTest < ActiveSupport::TestCase
+    context "A Gelbooru direct image url without a referer" do
+      strategy_should_work(
+        "https://img3.gelbooru.com/images/04/f2/04f2767c64593c3030ce74ecc2528704.jpg",
+        image_urls: ["https://img3.gelbooru.com/images/04/f2/04f2767c64593c3030ce74ecc2528704.jpg"],
+        artist_name: "灰色灰烬bot",
+        profile_url: "https://www.pixiv.net/users/3330425",
+        tags: %w[1girl back_bow bangs black_pantyhose blue_bow blue_hair blue_ribbon boots bow cape chibi chinese_commentary closed_eyes full_body hair_between_eyes hair_ribbon hat hatsune_miku indai_(3330425) on_ground pantyhose pom_pom_(clothes) rabbit rabbit_yukine rating:general ribbon simple_background sitting solo twintails vocaloid white_background white_cape white_headwear witch_hat yuki_miku yuki_miku_(2014) 初音ミク 雪ミク],
+        artist_commentary_title: "2010~2021雪ミク",
+        artist_commentary_desc: "动作参考@速写班长",
+        download_size: 480_621,
+      )
+    end
+
+    context "A Gelbooru direct image url with a referer" do
+      strategy_should_work(
+        "https://img3.gelbooru.com/images/04/f2/04f2767c64593c3030ce74ecc2528704.jpg",
+        referer: "https://gelbooru.com/index.php?page=post&s=view&id=7798121",
+        image_urls: ["https://img3.gelbooru.com/images/04/f2/04f2767c64593c3030ce74ecc2528704.jpg"],
+        artist_name: "灰色灰烬bot",
+        profile_url: "https://www.pixiv.net/users/3330425",
+        tags: %w[1girl back_bow bangs black_pantyhose blue_bow blue_hair blue_ribbon boots bow cape chibi chinese_commentary closed_eyes full_body hair_between_eyes hair_ribbon hat hatsune_miku indai_(3330425) on_ground pantyhose pom_pom_(clothes) rabbit rabbit_yukine rating:general ribbon simple_background sitting solo twintails vocaloid white_background white_cape white_headwear witch_hat yuki_miku yuki_miku_(2014) 初音ミク 雪ミク],
+        artist_commentary_title: "2010~2021雪ミク",
+        artist_commentary_desc: "动作参考@速写班长",
+        download_size: 480_621,
+      )
+    end
+
+    context "A Gelbooru sample image url" do
+      strategy_should_work(
+        "https://img3.gelbooru.com/samples/04/f2/sample_04f2767c64593c3030ce74ecc2528704.jpg",
+        image_urls: ["https://img3.gelbooru.com/images/04/f2/04f2767c64593c3030ce74ecc2528704.jpg"],
+        artist_name: "灰色灰烬bot",
+        profile_url: "https://www.pixiv.net/users/3330425",
+        tags: %w[1girl back_bow bangs black_pantyhose blue_bow blue_hair blue_ribbon boots bow cape chibi chinese_commentary closed_eyes full_body hair_between_eyes hair_ribbon hat hatsune_miku indai_(3330425) on_ground pantyhose pom_pom_(clothes) rabbit rabbit_yukine rating:general ribbon simple_background sitting solo twintails vocaloid white_background white_cape white_headwear witch_hat yuki_miku yuki_miku_(2014) 初音ミク 雪ミク],
+        artist_commentary_title: "2010~2021雪ミク",
+        artist_commentary_desc: "动作参考@速写班长",
+        download_size: 480_621,
+      )
+    end
+
+    context "A Gelbooru page url" do
+      strategy_should_work(
+        "https://gelbooru.com/index.php?page=post&s=view&id=7798121",
+        image_urls: ["https://img3.gelbooru.com/images/04/f2/04f2767c64593c3030ce74ecc2528704.jpg"],
+        artist_name: "灰色灰烬bot",
+        profile_url: "https://www.pixiv.net/users/3330425",
+        tags: %w[1girl back_bow bangs black_pantyhose blue_bow blue_hair blue_ribbon boots bow cape chibi chinese_commentary closed_eyes full_body hair_between_eyes hair_ribbon hat hatsune_miku indai_(3330425) on_ground pantyhose pom_pom_(clothes) rabbit rabbit_yukine rating:general ribbon simple_background sitting solo twintails vocaloid white_background white_cape white_headwear witch_hat yuki_miku yuki_miku_(2014) 初音ミク 雪ミク],
+        artist_commentary_title: "2010~2021雪ミク",
+        artist_commentary_desc: "动作参考@速写班长",
+        download_size: 480_621,
+      )
+    end
+  end
+end