Class: Deba::Extractor

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

Constant Summary collapse

HEADING_TAGS = 
%w(h1 h2 h3 h4 h5 h6)
BLOCK_INITIATING_TAGS = 
%w(article aside body blockquote div dd dt header li nav ol p pre section td th ul)
ENHANCERS = 
{ %w(b strong) => "*", %w(i em) => "_" }

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(doc, options = {}) ⇒ Extractor 



8
9
10
11
 
# File 'lib/deba/extractor.rb', line 8

def initialize(doc, options = {})
  @node = doc.root
  @options = options
end

Instance Attribute Details

#blocks ⇒ Object (readonly)

Returns the value of attribute blocks.



6
7
8
 
# File 'lib/deba/extractor.rb', line 6

def blocks
  @blocks
end

Instance Method Details

#extract ⇒ Object 



13
14
15
16
17
18
19
20
21
 
# File 'lib/deba/extractor.rb', line 13

def extract
  @just_appended_br = false
  @in_blockquote = false
  @document = Deba::Document.new(self)

  process(@node)

  @document.content.chomp("\n")
end

#in_blockquote? ⇒ Boolean 



128
129
130
 
# File 'lib/deba/extractor.rb', line 128

def in_blockquote?
  @in_blockquote
end

#process(node) ⇒ Object 



23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
 
# File 'lib/deba/extractor.rb', line 23

def process(node)
  if @options.key?(:exclude)
    return if Array(@options[:exclude]).any? { |selector| node.matches?(selector) }
  end

  node_name = node.name.downcase

  return if node_name == 'head'

  #Handle repeated brs by making a paragraph break
  if node_name == 'br'
    if @just_appended_br
      @just_appended_br = false

      @document.break(Deba::Paragraph)

      return
    else
      @just_appended_br = true
    end
  elsif @just_appended_br
    @just_appended_br = false

    @document << "\n"
  end

  if node.text?
    @document << Deba::Span.new(node.inner_text) if Deba::Utils.present?(node.inner_text)

    return
  end

  if ENHANCERS.keys.flatten.include?(node_name)
    ENHANCERS.each_pair do |tags, nsf_rep|
      if tags.include?(node_name)
        @document << nsf_rep
        node.children.each { |n| process(n) }
        @document << nsf_rep
      end
    end

    return
  end

  if node_name == 'blockquote'
    @in_blockquote = true

    @document.break(Deba::Paragraph)
    node.children.each { |n| process(n) }
    @document.break(Deba::Paragraph)

    @in_blockquote = false

    return
  end

  if node_name == 'li'
    last_item = node.xpath('count(following-sibling::li)').to_i == 0
    index = node.xpath('boolean(ancestor::ol)') ? (node.xpath('count(preceding-sibling::li)').to_i + 1) : nil
    
    @document.break(Deba::ListItem, last_item, index)
    node.children.each { |n| process(n) }
    @document.break(Deba::Paragraph)

    return
  end

  if node_name == 'dt'
    @document.break(Deba::DefinitionTerm)
    node.children.each { |n| process(n) }
    @document.break(Deba::Paragraph)

    return
  end

  if node_name == 'dd'
    last_item = node.xpath('count(following-sibling::dd)').to_i == 0
    @document.break(Deba::DefinitionDescription, last_item)
    node.children.each { |n| process(n) }
    @document.break(Deba::Paragraph)

    return
  end

  #These tags terminate the current paragraph, if present, and start a new paragraph
  if BLOCK_INITIATING_TAGS.include?(node_name)
    @document.break(Deba::Paragraph)
    node.children.each { |n| process(n) }
    @document.break(Deba::Paragraph)

    return
  end

  if HEADING_TAGS.include?(node_name)
    @document.break(Deba::Heading, node_name[1..-1].to_i)
    node.children.each { |n| process(n) }
    @document.break(Deba::Paragraph)

    return
  end

  #Pretend that the children of this node were siblings of this node (move them one level up the tree)
  node.children.each { |n| process(n) }
end