panchira 1.3.0 → 1.3.4

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 9f1b155399ed2e9caef218bd5599d72c3d6102c5b7e0da50fc297fe4db99614c
-  data.tar.gz: 92e31687ac7d5edbbd94df6eff606cabaa3dab891b57fda1a3b9a4af3815949c
+  metadata.gz: 8c708ae7d488bb68419b16ba734bca6e31c023913104453373269168780781f1
+  data.tar.gz: 7ed26206ae6a2c87c662dfef7f2b9817232142bbdb474160150f25348d8f5fc6
 SHA512:
-  metadata.gz: 8b8a5e274a776a2a384c4901b754fa14025c1f4a3b735708d5e692656b2462f1fbdbc8df63edde2600c154b440195e3c4183a42a1a9d9245668e35b0494ae41d
-  data.tar.gz: 727acf463987e381620e0a6d5200c22063df44d4d5fa4c3f086abf950ea7543f105fab7077446cc6cdfcc3beb9c90b8da4a617ddb5146bd3930c4ab91c67f4e0
+  metadata.gz: 5576bf4e990a86737e2f8a520c619a1028c4fa32d2f372e35e21e3a0ae914ebfdff03fc5ffcbe29d85c4332dfe2223b56af71a97065b365c575ae3ad97b24abb
+  data.tar.gz: 28b1c641f95a3cff9097032d4695d1eda1b9f759c4bc598778e6949bf63cf3d711a7526b5736ed7391fe17d30e71517676f6c7ae4a076375c917ed7d3f2afa8e

data/.github/workflows/ruby.yml

@@ -16,7 +16,7 @@ on:
 jobs:
   test:
 
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-18.04
 
     steps:
     - uses: actions/checkout@v2

data/CHANGELOG.md

@@ -4,6 +4,29 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](http://keepachangelog.com/)
 and this project adheres to [Semantic Versioning](http://semver.org/).
 
+## 1.3.4 - 2021-07-26
+### Fixed
+- Fixed an issue where Iwara Resolver failed when a description was present.
+
+## 1.3.3 - 2021-07-25
+### Added
+- Added Support for Iwara.
+
+### Fixed
+- Fixed an issue where DLsite Resolver was retrieving wrong tags.
+
+## 1.3.2 - 2021-05-23
+### Fixed
+- Fixed an issue where Fanza Resolver was retrieving incorrect canonical URLs.
+- Fixed an issue where Narou Resolver was retrieving wrong descriptions.
+
+### Changed
+- Updated dependencies.
+
+## 1.3.1 - 2021-02-17
+### Added
+- Added support for Fanza Video.
+
 ## 1.3.0 - 2021-02-06
 ### Added
 - Added support for multiple authors. PanchiraResult#authors now returns an array of authors.
@@ -67,6 +90,8 @@ and this project adheres to [Semantic Versioning](http://semver.org/).
 ### Added
 - Released Panchira gem. At this time we can parse only 5 websites.
 
+[1.3.2]: https://github.com/nuita/panchira/releases/tag/v1.3.2
+[1.3.1]: https://github.com/nuita/panchira/releases/tag/v1.3.1
 [1.3.0]: https://github.com/nuita/panchira/releases/tag/v1.3.0
 [1.2.0]: https://github.com/nuita/panchira/releases/tag/v1.2.0
 [1.1.0]: https://github.com/nuita/panchira/releases/tag/v1.1.0

data/Gemfile.lock

@@ -1,41 +1,41 @@
 PATH
   remote: .
   specs:
-    panchira (1.3.0)
+    panchira (1.3.4)
       fastimage (~> 2.1.7)
       nokogiri (>= 1.10.9, < 1.12.0)
 
 GEM
   remote: https://rubygems.org/
   specs:
-    ast (2.4.1)
+    ast (2.4.2)
     fastimage (2.1.7)
-    minitest (5.14.2)
-    nokogiri (1.11.0-x86_64-darwin)
+    minitest (5.14.4)
+    nokogiri (1.11.7-x86_64-darwin)
       racc (~> 1.4)
     parallel (1.20.1)
-    parser (3.0.0.0)
+    parser (3.0.1.1)
       ast (~> 2.4.1)
     racc (1.5.2)
     rainbow (3.0.0)
     rake (12.3.3)
-    regexp_parser (2.0.3)
-    rexml (3.2.4)
-    rubocop (1.7.0)
+    regexp_parser (2.1.1)
+    rexml (3.2.5)
+    rubocop (1.15.0)
       parallel (~> 1.10)
-      parser (>= 2.7.1.5)
+      parser (>= 3.0.0.0)
       rainbow (>= 2.2.2, < 4.0)
       regexp_parser (>= 1.8, < 3.0)
       rexml
-      rubocop-ast (>= 1.2.0, < 2.0)
+      rubocop-ast (>= 1.5.0, < 2.0)
       ruby-progressbar (~> 1.7)
-      unicode-display_width (>= 1.4.0, < 2.0)
-    rubocop-ast (1.4.0)
-      parser (>= 2.7.1.5)
-    rubocop-minitest (0.10.2)
-      rubocop (>= 0.87, < 2.0)
+      unicode-display_width (>= 1.4.0, < 3.0)
+    rubocop-ast (1.5.0)
+      parser (>= 3.0.1.1)
+    rubocop-minitest (0.12.1)
+      rubocop (>= 0.90, < 2.0)
     ruby-progressbar (1.11.0)
-    unicode-display_width (1.7.0)
+    unicode-display_width (2.0.0)
 
 PLATFORMS
   ruby

data/README.md

@@ -6,7 +6,7 @@
 
 Due to some legal or ethical issues, most hentai and NSFW platforms don't clarify their content on meta tags. As a result, most hentai platforms are rendered poorly on the card previews on social media.
 
-To solve this issue, Panchira is made to parse correct and uncensored metadata from such web platforms (at this time we cover **DLSite, Komiflo, Melonbooks, Nijie, Pixiv, Shousetsuka ni narou, Fanza and Twitter**).
+To solve this issue, Panchira is made to parse correct and uncensored metadata from such web platforms (at this time we cover **DLSite, Komiflo, Melonbooks, Nijie, Pixiv, Shousetsuka ni narou, Fanza, Iwara and Twitter**).
 
 If you need card previews of hentai on your web application, but can't get them with simply parsing metatags, then it is time for Panchira.
 

data/lib/panchira/resolvers/dlsite_resolver.rb

@@ -39,7 +39,13 @@ module Panchira
       end
 
       def parse_tags
-        @page.css('.main_genre').children.children.map(&:text)
+        @page.css('table[id*="work_"] tr').each do |tr|
+          next unless tr.css('th').text =~ /ジャンル/
+
+          return tr.css('td a').map do |node|
+            node.text.strip
+          end
+        end
       end
   end
 

data/lib/panchira/resolvers/fanza_resolver.rb

@@ -41,6 +41,12 @@ module Panchira
 
       private
 
+        # canonical urlに別サービス(FANZA GAMES)のURLが設定されていることがあるため、
+        # 別サービスの場合はとりあえず元URLを設定する
+        def parse_canonical_url
+          @url
+        end
+
         def parse_circle
           @page.css('a.circleName__txt').first.text
         end
@@ -49,8 +55,24 @@ module Panchira
           @page.css('.genreTag__item').map { |t| t.text.strip }
         end
     end
+
+    class FanzaVideoResolver < FanzaResolver
+      URL_REGEXP = %r{www.dmm.co.jp/digital/}.freeze
+
+      private
+
+        def parse_title
+          # og:titleは文字数制限で短く切られてる
+          @page.title.match(/(.+)- \S+ - FANZA動画/)[1]&.strip || super
+        end
+
+        def parse_image_url
+          super.sub(/(pr|ps).jpg$/, 'pl.jpg')
+        end
+    end
   end
 
   ::Panchira::Extensions.register(Panchira::Fanza::FanzaBookResolver)
   ::Panchira::Extensions.register(Panchira::Fanza::FanzaDoujinResolver)
+  ::Panchira::Extensions.register(Panchira::Fanza::FanzaVideoResolver)
 end

data/lib/panchira/resolvers/iwara_resolver.rb

@@ -0,0 +1,39 @@
+# frozen_string_literal: true
+
+module Panchira
+  class IwaraResolver < Resolver
+    URL_REGEXP = /(www|ecchi)\.iwara\.tv\//.freeze
+
+    private
+      def parse_title
+        super.split(' | ')[0]
+      end
+
+      def parse_image_url
+        url = @page.at_css('#video-player')&.attributes['poster']&.value
+        'https:' + url if url
+      end
+
+      def parse_author
+        @page.at_css('.node-info .username')&.children&.[](0)&.text
+      end
+
+      def parse_tags
+        @page.css('.field-name-field-categories .field-item').map { |e| e.children&.text }
+      end
+
+      def parse_description
+        @page.at_css('.field-name-body')&.text
+      end
+
+      def cookie
+        'show_adult=1'
+      end
+
+      def parse_canonical_url
+        @url # canonical has relative path. ignore it
+      end
+  end
+
+  ::Panchira::Extensions.register(Panchira::IwaraResolver)
+end

data/lib/panchira/resolvers/narou_resolver.rb

@@ -25,6 +25,10 @@ module Panchira
         Nokogiri::HTML.parse(res.body, uri)
       end
 
+      def parse_description
+        @desc&.xpath('//*[@id="noveltable1"]/tr/td')&.first&.text&.strip
+      end
+
       def parse_author
         @desc&.xpath('//*[@id="noveltable1"]/tr[2]/td')&.text&.strip
       end
@@ -33,6 +37,11 @@ module Panchira
         # つらい。
         @desc&.xpath('//*[@id="noveltable1"]/tr[3]')&.text&.split("\n\n\n")&.dig(1)&.split(' ')
       end
+
+      # og:urlで指定されたncode.syosetu.com/~~~にアクセスすると301で戻されるので何もしない
+      def parse_canonical_url
+        @url
+      end
     end
 
     class NcodeResolver < Resolver
@@ -47,6 +56,10 @@ module Panchira
         end
       end
 
+      def parse_description
+        @desc&.xpath('//*[@id="noveltable1"]/tr/td')&.first&.text&.strip
+      end
+
       def parse_author
         @desc&.xpath('//*[@id="noveltable1"]/tr[2]/td')&.text&.strip
       end
@@ -55,6 +68,11 @@ module Panchira
         # めっちゃつらい。
         @desc&.xpath('//*[@id="noveltable1"]/tr[3]')&.text&.split("\n\n\n")&.dig(1)&.delete("\u00A0")&.split(' ')&.grep_v('')
       end
+
+      # og:urlで指定されたncode.syosetu.com/~~~にアクセスすると301で戻されるので何もしない
+      def parse_canonical_url
+        @url
+      end
     end
   end
 

data/lib/panchira/resolvers/resolver.rb

@@ -67,17 +67,20 @@ module Panchira
         # fetch page and refresh canonical_url until canonical_url converges.
         loop do
           url_in_res = @page.css('//link[rel="canonical"]/@href').to_s
+          if url_in_res.empty?
+            url_in_res = @page.css('//meta[property="og:url"]/@content').to_s
+          end
 
           if url_in_res.empty?
             return history.last || @url
-          else
-            if history.include?(url_in_res) || history.length > 5
-              return url_in_res
-            else
-              history.push(url_in_res)
-              @page = fetch_page(url_in_res)
-            end
           end
+
+          if history.include?(url_in_res) || history.length > 5
+            return url_in_res
+          end
+
+          history.push(url_in_res)
+          @page = fetch_page(url_in_res)
         end
       end
 

data/lib/panchira/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Panchira
-  VERSION = '1.3.0'
+  VERSION = '1.3.4'
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: panchira
 version: !ruby/object:Gem::Version
-  version: 1.3.0
+  version: 1.3.4
 platform: ruby
 authors:
 - kyp
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2021-02-06 00:00:00.000000000 Z
+date: 2021-07-26 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -141,6 +141,7 @@ files:
 - lib/panchira/resolvers/dlsite_resolver.rb
 - lib/panchira/resolvers/fanza_resolver.rb
 - lib/panchira/resolvers/image_resolver.rb
+- lib/panchira/resolvers/iwara_resolver.rb
 - lib/panchira/resolvers/komiflo_resolver.rb
 - lib/panchira/resolvers/melonbooks_resolver.rb
 - lib/panchira/resolvers/narou_resolver.rb