wikibot 0.2.0

data/Manifest

@@ -0,0 +1,9 @@
+README.textile
+Rakefile
+lib/category.rb
+lib/class_ext.rb
+lib/hash_ext.rb
+lib/openhash.rb
+lib/page.rb
+lib/wikibot.rb
+Manifest

data/README.textile

@@ -0,0 +1,17 @@
+h1=(). WikiBot
+
+h2. About
+
+WikiBot was originally a PHP-based framework for bots I run on Wikipedia, however when it broke due to changes in mediawiki code, I decided to rewrite it in Ruby, using the MediaWiki API instead of screen-scraping. This is the result.
+
+As you'll notice, the features it provides are rather sparse at the moment, as I've only been adding features that I require. When I get more time, I'll flesh it out more.
+
+h2. Gem Requirements
+
+* "taf2-curb":taf2 (a fork of curb; using the base curb gem might work but is untested)
+* "xml-simple":xml
+* "deep_merge":dm
+
+[taf2]http://github.com/taf2/curb/tree/master
+[xml]http://xml-simple.rubyforge.org/
+[dm]http://rubyforge.org/projects/deepmerge/

data/Rakefile

@@ -0,0 +1,15 @@
+require 'rubygems'
+require 'rake'
+require 'echoe'
+
+Echoe.new('wikibot', '0.2.0') do |p|
+  p.description    = "Mediawiki Bot framework"
+  p.url            = "http://github.com/dvandersluis/wiki_bot"
+  p.author         = "Daniel Vandersluis"
+  p.email          = "daniel@codexed.com"
+  p.ignore_pattern = ["tmp/*", "script/*"]
+  p.development_dependencies = []
+  p.runtime_dependencies = ['taf2-curb', 'xml-simple', 'deep_merge']
+end
+
+Dir["#{File.dirname(__FILE__)}/tasks/*.rake"].sort.each { |ext| load ext }

data/lib/category.rb

@@ -0,0 +1,48 @@
+module WikiBot
+  class Category < Page
+    def category_info
+      data = {
+        :action => :query,
+        :titles => @name,
+        :prop => :categoryinfo
+      }
+
+      # The query API returns nothing for an empty cat, so we'll return a hash with all the normal
+      # properties set to 0 instead
+      empty_cat = { "pages" => 0, "size" => 0, "files" => 0, "subcats" => 0, "hidden" => "" }
+      @wiki_bot.query_api(:get, data).query.pages.page.categoryinfo || empty_cat
+    end
+
+    def members(sort = :sortkey, dir = :desc, namespace = nil)
+      data = {
+        :action => :query,
+        :list => :categorymembers,
+        :cmtitle => @name,
+        :cmsort => sort,
+        :cmdir => dir,
+        :cmnamespace => namespace
+      }
+
+      @wiki_bot.query_api(:get, data).query.categorymembers.cm
+    end
+
+    # Returns a hash of how many pages live in a category
+    def count(include_subcats = false)
+      out = {}
+      ci = category_info
+
+      out[@name] = {
+        :pages => ci.pages.to_i
+      }
+
+      if include_subcats and ci.subcats.to_i > 0
+        out[@name][:subcats] = {}
+        members.each do |m|
+          out[@name][:subcats].merge! WikiPage.new(@wiki_bot, m.title).count(include_subcats)
+        end
+      end
+
+      out
+    end
+  end
+end

data/lib/class_ext.rb

@@ -0,0 +1,40 @@
+# Borrowed from Rails 
+class Class
+  def cattr_reader(*syms)
+    syms.flatten.each do |sym|
+      next if sym.is_a?(Hash)
+      class_eval(<<-EOS, __FILE__, __LINE__)
+        unless defined? @@#{sym}
+          @@#{sym} = nil
+        end
+        
+        def self.#{sym}
+          @@#{sym}
+        end
+        
+        def #{sym}
+          @@#{sym}
+        end
+      EOS
+    end
+  end
+
+  def cattr_writer(*syms)
+    syms.flatten.each do |sym|
+      class_eval(<<-EOS, __FILE__, __LINE__)
+        unless defined? @@#{sym}
+          @@#{sym} = nil
+        end
+        
+        def self.#{sym}=(obj)
+          @@#{sym} = obj
+        end
+      EOS
+    end
+  end
+
+  def cattr_accessor(*syms)
+    cattr_reader(*syms)
+    cattr_writer(*syms)
+  end
+end

data/lib/hash_ext.rb

@@ -0,0 +1,17 @@
+require 'cgi'
+
+class Hash
+  def to_post_fields
+    inject([]) do |memo, pair|
+      key, val = pair
+      memo.push Curl::PostField.content(key.to_s, val.to_s)
+    end
+  end
+
+  def to_querystring
+    inject([]) do |memo, pair|
+      key, val = pair
+      memo.push "#{CGI::escape(key.to_s)}=#{CGI::escape(val.to_s)}"
+    end.join("&")
+  end
+end

data/lib/openhash.rb

@@ -0,0 +1,23 @@
+class OpenHash < Hash
+  undef_method :id
+
+  def initialize(hash = {})
+    super()
+    update(hash)
+  end
+
+  def [](key)
+    h = super(key)
+    h.is_a?(Hash) ? OpenHash.new(h) : h
+  end
+
+  # Allow hash properties to be referenced by dot notation
+  def method_missing(name, *args)
+    name = name.to_s
+    if self.include? name
+      self[name]
+    elsif self.include? name.to_sym
+      self[name.to_sym]
+    end
+  end
+end

data/lib/page.rb

@@ -0,0 +1,108 @@
+module WikiBot
+  class Page
+    class WriteError < StandardError; end
+
+    attr_writer :wiki_bot
+    attr_reader :name
+
+    def initialize(wiki_bot, name)
+      @wiki_bot = wiki_bot
+      @name = name
+    end
+
+    ###
+    # Read from page
+    def content
+      @content ||= begin
+        data = {
+          :action => :query,
+          :titles => @name,
+          :prop => :revisions,
+          :rvprop => :content
+        }
+
+        @wiki_bot.query_api(:get, data).query.pages.page.revisions.rev.content
+      end
+    end
+
+    # Parse page content
+    def text
+      @text ||= begin
+        data = {
+          :action => :parse,
+          :page => @name
+        }
+
+        @wiki_bot.query_api(:get, data).parse.text.content
+      end
+    end
+
+    ###
+    # Get page categories
+    def categories(show = :all)
+      # Cache hidden and non-hidden categories separately
+      @categories ||= begin
+        puts "Loading category data"
+        data = {
+          :action => :query,
+          :titles => @name,
+          :prop => :categories,
+          :clshow => "!hidden"
+        }
+
+        categories = @wiki_bot.query_api(:get, data).query.pages.page.categories.cl
+        categories = categories.inject([]) do |memo, category|
+          memo.push(WikiBot::Category.new(@wiki_bot, category.title))
+        end
+        
+        data = {
+          :action => :query,
+          :titles => @name,
+          :prop => :categories,
+          :clshow => "hidden"
+        }
+
+        hidden_categories = @wiki_bot.query_api(:get, data).query.pages.page.categories.cl
+        hidden_categories = hidden_categories.inject([]) do |memo, category|
+          memo.push(WikiBot::Category.new(@wiki_bot, category.title))
+        end
+
+        {:nonhidden => categories, :hidden => hidden_categories}
+      end
+
+      show = :all unless [:all, :hidden, :nonhidden].include? show
+      return @categories[:nonhidden] + @categories[:hidden] if show == :all
+      @categories[show]
+    end
+
+    def category_names(show = :all)
+      categories(show).map{ |c| c.name }
+    end
+
+    ###
+    # Write to page
+    def write(text, summary, section = nil, minor = false)
+      return if @wiki_bot.debug or @wiki_bot.readonly
+
+      data = {
+        :action => :edit,
+        :title => @name,
+
+        :text => text,
+        :token => @wiki_bot.edit_token,
+        :summary => summary,
+        :recreate => 1,
+        :bot => 1
+      }
+
+      data[:section] = section if !section.nil?
+      data[:minor] = 1 if minor
+      data[:notminor] = 1 if !minor
+    
+      result = @wiki_bot.query_api(:post, data)
+      status = result.edit.result
+      @wiki_bot.page_writes += 1
+      raise WriteError, status unless status == "Success"
+    end
+  end
+end

data/lib/wikibot.rb

@@ -0,0 +1,208 @@
+require 'class_ext'
+require 'hash_ext'
+require 'openhash'
+require 'page'
+require 'category'
+
+require 'rubygems'
+require 'curb'
+require 'xmlsimple'
+require 'deep_merge'
+
+module WikiBot
+  class CurbError < StandardError
+    attr_accessor :curb
+    def initialize(curb)
+      @curb = curb 
+    end
+  end
+
+  class APIError < StandardError
+    attr_accessor :code, :info
+    def initialize(code, info)
+      @code = code
+      @info = info
+      end
+  end
+
+  class Bot
+    class LoginError < StandardError; end
+
+    @@version = "0.2.0"       # WikiBot version
+
+    cattr_reader :version
+    cattr_accessor :cookiejar # Filename where cookies will be stored
+
+    attr_reader :config
+    attr_reader :api_hits
+    attr_accessor :page_writes
+    attr_accessor :debug      # In debug mode, no writes will be made to the wiki
+    attr_accessor :readonly   # Writes will not be made
+
+    def initialize(username, password, options = {})
+      @config = Hash.new
+      @cookies = Hash.new
+      @api_hits = 0
+      @page_writes = 0
+
+      api = options[:api] || "http://en.wikipedia.org/w/api.php"
+      auto_login = options[:auto_login] || false
+      @readonly = options[:readonly] || false
+      @debug = options[:debug] || false
+
+      @config = {
+        :username   => username,
+        :password   => password,
+        :api        => api,
+        :logged_in  => false
+      }
+
+      # Set up cURL:
+      @curl = Curl::Easy.new do |c|
+        c.headers["User-Agent"] = "Mozilla/5.0 Curb/Taf2/0.2.8 WikiBot/#{config[:username]}/#{@@version}"
+        #c.enable_cookies        = true
+        #c.cookiejar             = @@cookiejar
+       end
+
+      login if auto_login
+    end
+
+    def query_api(method, raw_data = {})
+      url = @config[:api]
+
+      raw_data[:format] = :xml if !(raw_data.include? :format and raw_data.include? 'format')
+
+      @curl.headers["Cookie"] = @cookies.inject([]) do |memo, pair|
+        key, val = pair
+        memo.push(CGI::escape(key) + "=" + CGI::escape(val))
+      end.join("; ") unless @cookies.nil?
+
+      response_xml = {}
+
+      while true
+        if method == :post
+          data = raw_data.to_post_fields
+        elsif method == :get
+          url = url.chomp("?") + "?" + raw_data.to_querystring
+          data = nil
+        end
+
+        @curl.url = url
+        @curl.headers["Expect"] = nil # MediaWiki will give a 417 error if Expect is set
+
+        if @debug
+          @curl.on_debug do |type, data|
+             p data
+          end
+        end
+        
+        # If Set-Cookie headers are given in the response, set the cookies
+        @curl.on_header do |data|
+          header, text = data.split(":").map(&:strip)
+          if header == "Set-Cookie"
+            parts = text.split(";")
+            cookie_name, cookie_value = parts[0].split("=")
+            @cookies[cookie_name] = cookie_value
+          end
+          data.length
+        end
+
+        if data.nil? or (data.is_a? Array and data.empty?)
+          @curl.send("http_#{method}".to_sym)
+        else
+          @curl.send("http_#{method}".to_sym, *data)
+        end
+        @api_hits += 1
+
+        raise CurbError.new(@curl) unless @curl.response_code == 200
+        
+        xml = XmlSimple.xml_in(@curl.body_str, {'ForceArray' => false})
+        raise APIError.new(xml['error']['code'], xml['error']['info']) if xml['error']
+
+        response_xml.deep_merge! xml
+        if xml['query-continue']
+          raw_data.merge! xml['query-continue'][xml['query-continue'].keys.first]
+        else
+          break
+        end
+      end
+
+      OpenHash.new(response_xml)
+    end
+    
+    def login
+      return if @config[:logged_in]
+
+      data = {
+        :action     => :login, 
+        :lgname     => @config[:username],
+        :lgpassword => @config[:password]
+      }
+
+      response = query_api(:post, data).login
+
+      if response.result == "NeedToken"
+        data = {
+          :action     => :login, 
+          :lgname     => @config[:username],
+          :lgpassword => @config[:password],
+          :lgtoken    => response.token
+        }
+
+        response = query_api(:post, data).login
+      end
+
+      raise LoginError, response.result unless response.result == "Success"
+
+      @config[:cookieprefix] = response.cookieprefix 
+      @config[:logged_in] = true
+    end
+
+    def logout
+      return if !@config[:logged_in]
+
+      query_api(:post, { :action => :logout })
+      @config[:logged_in] = false
+      @config[:edit_token] = nil
+    end
+
+    def edit_token(page = "Main Page")
+      @config[:edit_token] ||= begin
+        data = {
+          :action     => :query, 
+          :prop       => :info,
+          :intoken    => :edit,
+          :titles     => page 
+        }
+
+        query_api(:get, data).query.pages.page.edittoken
+      end
+    end
+
+    # Get wiki stats
+    def stats
+      data = {
+        :action       => :query,
+        :meta         => :siteinfo,
+        :siprop       => :statistics
+      }
+
+      query_api(:get, data).query.statistics
+    end
+
+    def page(name)
+      WikiBot::Page.new(self, name)
+    end
+
+    def category(name)
+      WikiBot::Category.new(self, name)
+    end
+
+    def format_date(date)
+      # Formats a date into the Wikipedia format
+      time = date.strftime("%H:%M")
+      month = date.strftime("%B")
+      "#{time}, #{date.day} #{month} #{date.year} (UTC)"
+    end
+  end
+end

data/wikibot.gemspec

@@ -0,0 +1,39 @@
+# -*- encoding: utf-8 -*-
+
+Gem::Specification.new do |s|
+  s.name = %q{wikibot}
+  s.version = "0.2.0"
+
+  s.required_rubygems_version = Gem::Requirement.new(">= 1.2") if s.respond_to? :required_rubygems_version=
+  s.authors = ["Daniel Vandersluis"]
+  s.date = %q{2010-04-09}
+  s.description = %q{Mediawiki Bot framework}
+  s.email = %q{daniel@codexed.com}
+  s.extra_rdoc_files = ["README.textile", "lib/category.rb", "lib/class_ext.rb", "lib/hash_ext.rb", "lib/openhash.rb", "lib/page.rb", "lib/wikibot.rb"]
+  s.files = ["README.textile", "Rakefile", "lib/category.rb", "lib/class_ext.rb", "lib/hash_ext.rb", "lib/openhash.rb", "lib/page.rb", "lib/wikibot.rb", "Manifest", "wikibot.gemspec"]
+  s.homepage = %q{http://github.com/dvandersluis/wiki_bot}
+  s.rdoc_options = ["--line-numbers", "--inline-source", "--title", "Wikibot", "--main", "README.textile"]
+  s.require_paths = ["lib"]
+  s.rubyforge_project = %q{wikibot}
+  s.rubygems_version = %q{1.3.5}
+  s.summary = %q{Mediawiki Bot framework}
+
+  if s.respond_to? :specification_version then
+    current_version = Gem::Specification::CURRENT_SPECIFICATION_VERSION
+    s.specification_version = 3
+
+    if Gem::Version.new(Gem::RubyGemsVersion) >= Gem::Version.new('1.2.0') then
+      s.add_runtime_dependency(%q<taf2-curb>, [">= 0"])
+      s.add_runtime_dependency(%q<xml-simple>, [">= 0"])
+      s.add_runtime_dependency(%q<deep_merge>, [">= 0"])
+    else
+      s.add_dependency(%q<taf2-curb>, [">= 0"])
+      s.add_dependency(%q<xml-simple>, [">= 0"])
+      s.add_dependency(%q<deep_merge>, [">= 0"])
+    end
+  else
+    s.add_dependency(%q<taf2-curb>, [">= 0"])
+    s.add_dependency(%q<xml-simple>, [">= 0"])
+    s.add_dependency(%q<deep_merge>, [">= 0"])
+  end
+end

metadata

@@ -0,0 +1,104 @@
+--- !ruby/object:Gem::Specification 
+name: wikibot
+version: !ruby/object:Gem::Version 
+  version: 0.2.0
+platform: ruby
+authors: 
+- Daniel Vandersluis
+autorequire: 
+bindir: bin
+cert_chain: []
+
+date: 2010-04-09 00:00:00 -06:00
+default_executable: 
+dependencies: 
+- !ruby/object:Gem::Dependency 
+  name: taf2-curb
+  type: :runtime
+  version_requirement: 
+  version_requirements: !ruby/object:Gem::Requirement 
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        version: "0"
+    version: 
+- !ruby/object:Gem::Dependency 
+  name: xml-simple
+  type: :runtime
+  version_requirement: 
+  version_requirements: !ruby/object:Gem::Requirement 
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        version: "0"
+    version: 
+- !ruby/object:Gem::Dependency 
+  name: deep_merge
+  type: :runtime
+  version_requirement: 
+  version_requirements: !ruby/object:Gem::Requirement 
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        version: "0"
+    version: 
+description: Mediawiki Bot framework
+email: daniel@codexed.com
+executables: []
+
+extensions: []
+
+extra_rdoc_files: 
+- README.textile
+- lib/category.rb
+- lib/class_ext.rb
+- lib/hash_ext.rb
+- lib/openhash.rb
+- lib/page.rb
+- lib/wikibot.rb
+files: 
+- README.textile
+- Rakefile
+- lib/category.rb
+- lib/class_ext.rb
+- lib/hash_ext.rb
+- lib/openhash.rb
+- lib/page.rb
+- lib/wikibot.rb
+- Manifest
+- wikibot.gemspec
+has_rdoc: true
+homepage: http://github.com/dvandersluis/wiki_bot
+licenses: []
+
+post_install_message: 
+rdoc_options: 
+- --line-numbers
+- --inline-source
+- --title
+- Wikibot
+- --main
+- README.textile
+require_paths: 
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement 
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      version: "0"
+  version: 
+required_rubygems_version: !ruby/object:Gem::Requirement 
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      version: "1.2"
+  version: 
+requirements: []
+
+rubyforge_project: wikibot
+rubygems_version: 1.3.5
+signing_key: 
+specification_version: 3
+summary: Mediawiki Bot framework
+test_files: []
+