Class: Datasets::Wikipedia::ArticlesListener

Inherits:
Object
  • Object
show all
Includes:
REXML::StreamListener
Defined in:
lib/datasets/wikipedia.rb

Instance Method Summary collapse

Constructor Details

#initialize(block) ⇒ ArticlesListener

Returns a new instance of ArticlesListener.



78
79
80
81
82
83
84
85
86
87
 
# File 'lib/datasets/wikipedia.rb', line 78

def initialize(block)
  @block = block
  @page = nil
  @revision = nil
  @contributor = nil
  @current_tag = nil
  @tag_stack = []
  @text_stack = [""]
  @first_page = true
end

Instance Method Details

#cdata(contnet) ⇒ Object 



156
157
158
 
# File 'lib/datasets/wikipedia.rb', line 156

def cdata(contnet)
  @text_stack.last << content
end

#tag_end(name) ⇒ Object 



103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
 
# File 'lib/datasets/wikipedia.rb', line 103

def tag_end(name)
  case name
  when "page"
    on_page(@page)
    @page = nil
  when "title"
    @page.title = @text_stack.last
  when "ns"
    @page.namespace = Integer(@text_stack.last)
  when "id"
    id = Integer(@text_stack.last)
    case @tag_stack[-2]
    when "page"
      @page.id = id
    when "revision"
      @revision.id = id
    when "contributor"
      @contributor.id = id
    end
  when "restrictions"
    @page.restrictions = @text_stack.last.split(":")
  when "revision"
    @page.revision = @revision
    @revision = nil
  when "parentid"
    @revision.parent_id = Integer(@text_stack.last)
  when "timestamp"
    @revision.timestamp = Time.iso8601(@text_stack.last)
  when "contributor"
    @revision.contributor = @contributor
    @contributor = nil
  when "username"
    @contributor.user_name = @text_stack.last
  when "minor"
    # TODO
  when "comment"
    @revision.comment = @text_stack.last
  when "model"
    @revision.model = @text_stack.last
  when "format"
    @revision.format = @text_stack.last
  when "text"
    @revision.text = @text_stack.last
  when "sha1"
    @revision.sha1 = @text_stack.last
  end
  pop_stacks
end

#tag_start(name, attributes) ⇒ Object 



89
90
91
92
93
94
95
96
97
98
99
100
101
 
# File 'lib/datasets/wikipedia.rb', line 89

def tag_start(name, attributes)
  push_stacks(name)
  case name
  when "page"
    @page = Page.new
  when "revision"
    @revision = Revision.new
  when "contributor"
    @contributor = Contributor.new
  when "redirect"
    @page.redirect = attributes["title"]
  end
end

#text(data) ⇒ Object 



152
153
154
 
# File 'lib/datasets/wikipedia.rb', line 152

def text(data)
  @text_stack.last << data
end