Class: PdfExtract::Receiver

Inherits:
Object
  • Object
show all
Defined in:
lib/pdf.rb

Instance Method Summary collapse

Constructor Details

#initialize(pdf) ⇒ Receiver

Returns a new instance of Receiver.



54
55
56
57
58
 
# File 'lib/pdf.rb', line 54

def initialize pdf
  @pdf = pdf
  @listeners = {}
  @object_listeners = {}
end

Instance Method Details

#add_spatial_objects(default_type, objs) ⇒ Object 



113
114
115
116
117
118
119
120
121
122
123
124
 
# File 'lib/pdf.rb', line 113

def add_spatial_objects default_type, objs
  if objs.class != Array
    objs = [objs] unless objs.nil?
    objs = [] if objs.nil?
  end

  objs.each do |obj|
    type = obj.delete(:group) || default_type
    @pdf.spatial_objects[type] ||= []
    @pdf.spatial_objects[type] << obj
  end
end

#after(&block) ⇒ Object 



73
74
75
 
# File 'lib/pdf.rb', line 73

def after &block
  @after = {:type => @pdf.operating_type, :fn => block}
end

#before(&block) ⇒ Object 



69
70
71
 
# File 'lib/pdf.rb', line 69

def before &block
  @before = block
end

#call_after ⇒ Object 



97
98
99
 
# File 'lib/pdf.rb', line 97

def call_after
  self.add_spatial_objects @after[:type], @after[:fn].call unless @after.nil?
end

#call_before ⇒ Object 



101
102
103
 
# File 'lib/pdf.rb', line 101

def call_before
  @before.call unless @before.nil?
end

#call_object_listeners(spatial_objects) ⇒ Object 



89
90
91
92
93
94
95
 
# File 'lib/pdf.rb', line 89

def call_object_listeners spatial_objects
  @object_listeners.each_pair do |type, listeners|
    listeners.each do |listener|
        spatial_objects[type].each { |obj| listener.call obj }
    end
  end
end

#expand_listeners_to_callback_methods ⇒ Object 



77
78
79
80
81
82
83
84
85
86
87
 
# File 'lib/pdf.rb', line 77

def expand_listeners_to_callback_methods
  # TODO merge on callback_name
  @listeners.each_pair do |callback_name, callback_handler|
    p = proc do |*args|
      spatial_objects = callback_handler[:fn].call args
      self.add_spatial_objects callback_handler[:type], spatial_objects
    end
    
    self.class.send :define_method, callback_name, p
  end
end

#for(callback_name, &block) ⇒ Object 



60
61
62
 
# File 'lib/pdf.rb', line 60

def for callback_name, &block
  @listeners[callback_name] = {:type => @pdf.operating_type, :fn => block}
end

#for_calls? ⇒ Boolean

Returns:

  • (Boolean) 


105
106
107
 
# File 'lib/pdf.rb', line 105

def for_calls?
  @listeners.size > 0
end

#invoke_calls(filename, spatial_options) ⇒ Object 



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
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
 
# File 'lib/pdf.rb', line 126

def invoke_calls filename, spatial_options
  if spatial_options[:paged]
      
    paged_objs = {}
    @object_listeners.each_pair do |type, _|
      @pdf.paged_objects(type).each_pair do |page, objs|
          paged_objs[page] ||= {}
        paged_objs[page][type] = objs
      end
    end
    
    paged_objs.each_pair do |page, objs|
      call_before

      if object_calls?
        @object_listeners.each_pair do |type, listeners|
          listeners.each do |listener|
            if objs[type].nil?
              raise "#{@pdf.operating_type} is missing a dependency on #{type}"
            end
            objs[type].each { |obj| listener.call obj }
          end
        end
      end
      
      call_after
    end
    
  else

    call_before
    if object_calls?
      call_object_listeners @pdf.spatial_objects
    end
    call_after

  end
  
  if for_calls?
    expand_listeners_to_callback_methods
    #PDF::Reader.file filename, self, :raw_text => true

    reader = PDF::Reader.new filename, :raw_text => true
    reader.pages.each do |page|
      begin_page page
      page.walk self
      end_page page
    end
  end
end

#object_calls? ⇒ Boolean

Returns:

  • (Boolean) 


109
110
111
 
# File 'lib/pdf.rb', line 109

def object_calls?
  @object_listeners.size > 0
end

#objects(type_name, &block) ⇒ Object 



64
65
66
67
 
# File 'lib/pdf.rb', line 64

def objects type_name, &block
  @object_listeners[type_name] ||= []
  @object_listeners[type_name] << block
end