Ruby4Misawa 0.1.2 → 0.1.4

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA1:
+  metadata.gz: 3b5ec5d9fb4d7d600949d79b09af92fa0af852fa
+  data.tar.gz: 4e387642d93550fdf83a372cfbc141ae2d8fc968
+SHA512:
+  metadata.gz: 9fa104901863b91789f0c440a8b7ea2a5462860ab5e588887596e613a7e565a7f6863593cb2c448f3b8d5dae8e61557c27383a8b1b1b116558d5891bb1601136
+  data.tar.gz: 8f8104fd574b0f0a64549de8a764fd6570882c737e02a5589566d109fa6ff646df691830bdf72888362a372fd7c5f2593a8c95aeff9a5b8c55e3f81c24355f79

data/Gemfile

@@ -1,3 +1,3 @@
-source :rubygems
-gemspec :name => 'Ruby4Misawa'
-gem 'nokogiri', '1.5.5'
+source 'http://rubygems.org'
+
+gemspec :name => "Ruby4Misawa"

data/README.md

@@ -0,0 +1,58 @@
+# Ruby4Misawa
+
+## About
+
+A scraping library for 地獄のミサワ
+
+This is not compatible with old version.
+
+## Installation
+
+add this line to your Gemfile.
+
+`gem 'Ruby4Misawa'`
+
+or
+
+`$ gem install Ruby4Misawa`
+
+## Usage
+
+### cid
+
+```ruby
+require 'Ruby4Misawa'
+
+misawa = Misawa.new(32)
+misawa.scrape ##> [{ :image => ..., :title => ..., :date => ..., :identifier => ..., :body => ... }, ...]
+```
+
+### name
+
+```ruby
+require 'Ruby4Misawa'
+
+misawa = Misawa.new("あつしさん")
+misawa.scrape ##> [{ :image => ..., :title => ..., :date => ..., :identifier => ..., :body => ... }, ...]
+```
+
+### designating page numebr
+
+```ruby
+require 'Ruby4Misawa'
+
+misawa = Misawa.new("あつしさん", 2)
+misawa.scrape ##> [{ :image => ..., :title => ..., :date => ..., :identifier => ..., :body => ... }, ...]
+```
+
+## Contributing
+
+1. fork the project.
+2. create your feature branch. (`git checkout -b my-feature`)
+3. commit your changes. (`git commit -am 'commit message.'`)
+4. push to the branch. (`git push origin my-feature`)
+5. send pull request.
+
+## License
+
+MIT

data/Ruby4Misawa.gemspec

@@ -0,0 +1,13 @@
+require File.expand_path("../lib/version", __FILE__)
+
+Gem::Specification.new "Ruby4Misawa", Misawa::VERSION do |spec|
+  spec.summary     = 'A scraping library for Jigoku no Misawa'
+  spec.homepage    = 'https://github.com/namusyaka/Ruby4Misawa'
+  spec.authors     = ['namusyaka']
+  spec.email       = 'namusyaka@gmail.com'
+  spec.files       = `git ls-files`.split("\n") - %w(.gitignore)
+  spec.test_files  = spec.files.select { |path| path =~ /^spec\/.*_spec\.rb/ }
+  spec.license     = "MIT"
+
+  spec.add_dependency('nokogiri')
+end

data/lib/Ruby4Misawa.rb

@@ -1,102 +1,62 @@
-# -*- encoding: UTF-8 -*-
-
-%w(
-  uri
-  open-uri
-  kconv
-  optparse
-  rubygems
-  nokogiri
-).each { | name | require name }
+require 'open-uri'
+require 'kconv'
+require 'rubygems'
+require 'nokogiri'
 
 class Misawa
-
-  class NotFoundError < ArgumentError; end;
-
-  DOMAIN = 'http://jigokuno.com/'
-
+  DOMAIN       = 'http://jigokuno.com/'
   @@categories = {}
 
+  class NotFoundError < ArgumentError; end
+
   Nokogiri.HTML(open(DOMAIN).read).css('dl').each do |dl|
     if dl.children[0].inner_html.toutf8 == '惚れさせ男子達'
-      dl.children[2].children[1].children.to_a.delete_if{|node| node.class == Nokogiri::XML::Text}.each do |li|
+      dl.children[2].children[1].children.to_a.delete_if{|node|node.class == Nokogiri::XML::Text}.each do |li|
         a = li.child
         @@categories[a.text.scan(/(.+?)(?:\([0-9]+\))?$/)[0][0]] = a[:href].scan(/cid=([0-9]+)/)[0][0]
       end
     end
   end
 
-  attr_reader :category, :page
-
-  def initialize(category, page = 0)
-    raise NotFoundError unless @@categories[category]
-    @category = @@categories[category]
-    @page = page =~ /^[0-9]+?\.\.[0-9]+?$/ ? eval(page) : page
+  def initialize(name, page = 0)
+    @name = name
+    @cid  = @name.is_a?(Integer) ? @name : name_to_cid(@name)
+    @page = page
   end
 
   def scrape
-    create_uri.inject([]) do | result, uri |
-      body = get_body(uri)
-      Nokogiri.HTML(body).css('img.pict').map { | img | result << img['src'] } if body
-      result
-    end
-  end
-
-  class << self;
+    data = []
 
-    def save_misawa(category, page)
-      misawa = self.new(category, page)
-      misawa.scrape.each_with_index do | src, i |
-        File.open("#{misawa.category}-#{i}.gif", 'w') { | f | f.write(misawa.get_body(src)) }
-      end
+    begin
+      nokogiri = Nokogiri.HTML(open(misawa_uri).read)
+    rescue OpenURI::HTTPError
+      raise NotFoundError
     end
 
-    def uri_misawa(category, page)
-      misawa = self.new(category, page)
-      misawa.scrape.each_with_index do | src, i |
-        puts "#{misawa.category}-#{i} : #{src} "
+    # parse some attributes
+    nokogiri.xpath('//comment()[contains(., "rdf")]').each do |entry|
+      attributes = Nokogiri.XML(entry.to_s.toutf8.gsub(/^<!--|-->$/, "")).child.css('rdf|Description')[0].attributes
+      data << %w[title date identifier].inject({}) do |result, key|
+        result[key.to_sym] = attributes[key].value
+        result
       end
     end
 
-    def method_missing(name, *args)
-      raise ArgumentError;
-    end
+    # parse images
+    nokogiri.css('img.pict').to_a.each_with_index { |image, i|
+      data[i].merge!(:image => image['src'], :body => image['alt'])
+    }
 
+    data
   end
 
-  def get_body(uri)
-    open(URI.encode(uri)).read
-  end
+  private
 
-  def create_uri
-    if @page.is_a?(Range)
-      base = "#{DOMAIN}/?cid=#{@category}&page="
-      @page.map { | i | "#{base}#{i}" }
-    else
-      ["#{DOMAIN}/?cid=#{@category}&page=#{@page}"]
-    end
+  def misawa_uri
+    "#{DOMAIN}?cid=#{@cid}&page=#{@page}"
   end
 
-end
-
-if $0 === __FILE__
-
-  OptionParser.new do | opt |
-
-    arguments = {}
-
-    opt.on('-t [TYPE]', '--type [TYPE]') do | value |
-      arguments[:type] = value
-    end
-    opt.on('-p [PAGE]', '--page [PAGE]') do | value |
-      arguments[:page] = value
-    end
-    opt.on('-c [CATEGORY]', '--category [CATEGORY]') do | value |
-      arguments[:category] = value
-    end
-
-    opt.parse!(ARGV)
-    Misawa.send("#{arguments[:type] || 'uri'}_misawa", arguments[:category], arguments[:page])
+  def name_to_cid(name)
+    @@categories[name]
   end
-
 end

data/lib/version.rb

@@ -0,0 +1,3 @@
+class Misawa
+  VERSION = '0.1.4'
+end

metadata

@@ -1,88 +1,63 @@
---- !ruby/object:Gem::Specification 
+--- !ruby/object:Gem::Specification
 name: Ruby4Misawa
-version: !ruby/object:Gem::Version 
-  hash: 31
-  prerelease: 
-  segments: 
-  - 0
-  - 1
-  - 2
-  version: 0.1.2
+version: !ruby/object:Gem::Version
+  version: 0.1.4
 platform: ruby
-authors: 
+authors:
 - namusyaka
 autorequire: 
 bindir: bin
 cert_chain: []
-
-date: 2012-09-29 00:00:00 Z
-dependencies: 
-- !ruby/object:Gem::Dependency 
+date: 2013-08-08 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
   name: nokogiri
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
   prerelease: false
-  requirement: &id001 !ruby/object:Gem::Requirement 
-    none: false
-    requirements: 
-    - - ">="
-      - !ruby/object:Gem::Version 
-        hash: 3
-        segments: 
-        - 0
-        version: "0"
-  type: :development
-  version_requirements: *id001
-description: !binary |
-  5Zyw542E44Gu44Of44K144Ov44OW44Ot44Kw44Gr44GC44KLVVJM44GM44Go
-  44KM44Gf44KK44CB44Gd44Gu44G+44G+55S75YOP44KS5L+d5a2Y44GX44Gf
-  44KK44Gn44GN44G+44GZ44CC
-
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+description: 
 email: namusyaka@gmail.com
 executables: []
-
 extensions: []
-
 extra_rdoc_files: []
-
-files: 
-- README.rdoc
-- Ruby4Misawa-0.1.2.gemspec
+files:
 - Gemfile
+- README.md
+- Ruby4Misawa.gemspec
 - lib/Ruby4Misawa.rb
+- lib/version.rb
 homepage: https://github.com/namusyaka/Ruby4Misawa
-licenses: []
-
+licenses:
+- MIT
+metadata: {}
 post_install_message: 
 rdoc_options: []
-
-require_paths: 
+require_paths:
 - lib
-required_ruby_version: !ruby/object:Gem::Requirement 
-  none: false
-  requirements: 
-  - - ">="
-    - !ruby/object:Gem::Version 
-      hash: 3
-      segments: 
-      - 0
-      version: "0"
-required_rubygems_version: !ruby/object:Gem::Requirement 
-  none: false
-  requirements: 
-  - - ">="
-    - !ruby/object:Gem::Version 
-      hash: 3
-      segments: 
-      - 0
-      version: "0"
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - '>='
+    - !ruby/object:Gem::Version
+      version: '0'
 requirements: []
-
 rubyforge_project: 
-rubygems_version: 1.8.1
+rubygems_version: 2.0.2
 signing_key: 
-specification_version: 3
-summary: !binary |
-  44K544Kv44Os44Kk44OU44Oz44Kw44Op44Kk44OW44Op44OqIGZvciDlnLDn
-  jYTjga7jg5/jgrXjg68=
-
+specification_version: 4
+summary: A scraping library for Jigoku no Misawa
 test_files: []
-
+has_rdoc: 

data/README.rdoc

@@ -1,24 +0,0 @@
-= Ruby4Misawa
-
-== これはなにか
-
-- 地獄のミサワの画像を取ってくるやつ
-- 画像はURLで出力するか、保存するかを選ぶ
-
-== インストール
-
-gem install Ruby4Misawa でおk
-
-== つかいかた
-
-libにあるRuby4Misawa.rbを適当に移動して、
-ruby Ruby4Misawa.rb [オプション] [値]
-
-[-t [uri, save]] uriは画像のURIを出力、saveは保存しまくる。permission関係でこける可能性はありけり。
-[-c [カテゴリ名]] ルシフェル や バイトリーダーのように指定。一つしか指定できないので注意。
-[-p [対象ページ]] 1 や 0のような数値か、0..5のようにRange形式で指定
-
---type, --category, --pageでも可
-
-ちなみに何も指定しないと地獄のミサワのトップページにあるミサワ画像(と思われるもの)を取ってくる
-

data/Ruby4Misawa-0.1.2.gemspec

@@ -1,13 +0,0 @@
-# -*- encoding: UTF-8 -*-
-
-Gem::Specification.new do | spec |
-  spec.name = 'Ruby4Misawa'
-  spec.version = '0.1.2'
-  spec.summary = 'スクレイピングライブラリ for 地獄のミサワ'
-  spec.description = '地獄のミサワブログにあるURLがとれたり、そのまま画像を保存したりできます。'
-  spec.homepage = 'https://github.com/namusyaka/Ruby4Misawa'
-  spec.author = 'namusyaka'
-  spec.email = 'namusyaka@gmail.com'
-  spec.files = %w( README.rdoc Ruby4Misawa-0.1.2.gemspec Gemfile lib lib/Ruby4Misawa.rb )
-  spec.add_development_dependency('nokogiri')
-end