fixes #3100

app/logical/nico_seiga_api_client.rb

@@ -0,0 +1,42 @@
+class NicoSeigaApiClient
+  BASE_URL = "http://seiga.nicovideo.jp/api"
+  attr_reader :user_id, :moniker, :image_id, :title, :desc
+
+  def initialize(illust_id)
+    get_illust(illust_id)
+    get_artist(user_id)
+  end
+
+  def get_illust(id)
+    uri = URI.parse("#{BASE_URL}/illust/info?id=#{id}")
+    Net::HTTP.start(uri.host, uri.port, :use_ssl => uri.is_a?(URI::HTTPS)) do |http|
+      resp = http.request_get(uri.request_uri)
+      if resp.is_a?(Net::HTTPSuccess)
+        parse_illust_xml_response(resp.body)
+      end
+    end
+  end
+
+  def get_artist(id)
+    uri = URI.parse("#{BASE_URL}/user/info?id=#{id}")
+    Net::HTTP.start(uri.host, uri.port, :use_ssl => uri.is_a?(URI::HTTPS)) do |http|
+      resp = http.request_get(uri.request_uri)
+      if resp.is_a?(Net::HTTPSuccess)
+        parse_artist_xml_response(resp.body)
+      end
+    end
+  end
+
+  def parse_artist_xml_response(text)
+    doc = Nokogiri::Slop(text)
+    @moniker = doc.response.user.nickname.content
+  end
+
+  def parse_illust_xml_response(text)
+    doc = Nokogiri::Slop(text)
+    @image_id = doc.response.image.id.content.to_i
+    @user_id = doc.response.image.user_id.content.to_i
+    @title = doc.response.image.title.content
+    @desc = [doc.response.image.description.try(:content), doc.response.image.summary.try(:content)].compact.join("\n\n")
+  end
+end

app/logical/sources/strategies/nico_seiga.rb

@@ -1,6 +1,8 @@
 module Sources
   module Strategies
     class NicoSeiga < Base
+      extend Memoist
+      
       def self.url_match?(url)
         url =~ /^https?:\/\/(?:\w+\.)?nico(?:seiga|video)\.jp/
       end
@@ -25,8 +27,9 @@ module Sources
       def get
         page = load_page
 
-        @artist_name, @profile_url = get_profile_from_page(page)
+        @artist_name, @profile_url = get_profile_from_api
         @image_url = get_image_url_from_page(page)
+        @artist_commentary_title, @artist_commentary_desc = get_artist_commentary_from_api
 
         # Log out before getting the tags.
         # The reason for this is that if you're logged in and viewing a non-adult-rated work, the tags will be added with javascript after the page has loaded meaning we can't extract them easily.
@@ -47,12 +50,25 @@ module Sources
 
       def normalize_for_artist_finder!
         page = load_page
-        @artist_name, @profile_url = get_profile_from_page(page)
+        @illust_id = get_illust_id_from_url
-        profile_url
+        @artist_name, @profile_url = get_profile_from_api
+        @profile_url + "/"
       end
 
     protected
 
+      def api_client
+        NicoSeigaApiClient.new(get_illust_id_from_url)
+      end
+
+      def get_illust_id_from_url
+        if normalized_url =~ %r!http://seiga.nicovideo.jp/seiga/im(\d+)!
+          $1.to_i
+        else
+          nil
+        end
+      end
+
       def load_page
         page = agent.get(normalized_url)
 
@@ -66,18 +82,8 @@ module Sources
         page
       end
 
-      def get_profile_from_page(page)
+      def get_profile_from_api
-        links = page.search("li a").select {|x| x["href"] =~ /user\/illust/}
+        return [api_client.moniker, "http://seiga.nicovideo.jp/user/illust/#{api_client.user_id}"]
-
-        if links.any?
-          profile_url = "http://seiga.nicovideo.jp" + links[0]["href"]
-          artist_name = links[0].search("span")[0].children[0].text
-        else
-          profile_url = nil
-          artist_name = nil
-        end
-
-        return [artist_name, profile_url].compact
       end
 
       def get_image_url_from_page(page)
@@ -108,6 +114,10 @@ module Sources
         end
       end
 
+      def get_artist_commentary_from_api
+        [api_client.title, api_client.desc]
+      end
+
       def normalized_url
         @normalized_url ||= begin
           if url =~ %r!\Ahttp://lohas\.nicoseiga\.jp/o/[a-f0-9]+/\d+/(\d+)!
@@ -163,6 +173,8 @@ module Sources
           mech
         end
       end
+
+      memoize :api_client
     end
   end
 end

app/models/artist.rb

@@ -44,6 +44,7 @@ class Artist < ActiveRecord::Base
           url = File.dirname(url) + "/"
           break if url =~ /pixiv\.net\/(?:img\/)?$/i
           break if url =~ /lohas\.nicoseiga\.jp\/priv\/$/i
+          break if url =~ /nicovideo\.jp\/user\/illust/
           break if url =~ /(?:data|media)\.tumblr\.com\/[a-z0-9]+\/$/i
           break if url =~ /deviantart\.net\//i
           break if url =~ %r!\Ahttps?://(?:mobile\.)?twitter\.com/\Z!i

script/fixes/046_fix_nicovideo_artist_urls.rb

@@ -0,0 +1,11 @@
+#!/usr/bin/env ruby
+
+require File.expand_path(File.join(File.dirname(__FILE__), '..', '..', 'config', 'environment'))
+
+ActiveRecord::Base.without_timeout do
+  ArtistUrl.where("normalized_url like ?", "\%nico\%").find_each do |url|
+    before = url.normalized_url
+    url.normalized_url = Sources::Site.new(before).normalize_for_artist_finder!
+    puts "#{before} -> #{url.normalized_url}" if before != url.normalized_url
+  end
+end

test/unit/artist_test.rb

@@ -250,6 +250,16 @@ class ArtistTest < ActiveSupport::TestCase
       end
     end
 
+    context "when finding nico seiga artists" do
+      setup do
+        FactoryGirl.create(:artist, :name => "osamari", :url_string => "http://seiga.nicovideo.jp/user/illust/7017777")
+      end
+
+      should "find the artist by the profile" do
+        assert_artist_found("osamari", "http://seiga.nicovideo.jp/seiga/im4937663")
+      end
+    end
+
     context "when finding twitter artists" do
       setup do
         FactoryGirl.create(:artist, :name => "hammer_(sunset_beach)", :url_string => "http://twitter.com/hamaororon")

test/unit/sources/nico_seiga_test.rb

@@ -21,6 +21,11 @@ module Sources
         assert_equal("osamari", @site_2.artist_name)
       end
 
+      should "get the artist commentary" do
+        assert_equal("コジコジ", @site_2.artist_commentary_title)
+        assert_equal("懐かしいですよね。テ また懐かしいものを ", @site_2.artist_commentary_desc)
+      end
+
       should "get the image url" do
         assert_match(/^http:\/\/lohas\.nicoseiga\.jp\/priv\//, @site_1.image_url)
         assert_match(/^http:\/\/lohas\.nicoseiga\.jp\/priv\//, @site_2.image_url)