Class: Rypper::Extractor

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

Instance Attribute Summary collapse

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(selector) ⇒ Extractor

Returns a new instance of Extractor.



20
21
22
 
# File 'lib/rypper/extractor.rb', line 20

def initialize(selector)
  @selector = selector
end

Instance Attribute Details

#selector ⇒ Object (readonly)

Returns the value of attribute selector.



5
6
7
 
# File 'lib/rypper/extractor.rb', line 5

def selector
  @selector
end

Class Method Details

.dirname(uri) ⇒ Object 



7
8
9
10
11
12
13
14
15
16
17
18
 
# File 'lib/rypper/extractor.rb', line 7

def self.dirname(uri)
  uri = uri.to_s
  dirname = nil
  if File.extname(uri).empty?
    dirname = uri
  else
    dirname = File.dirname(uri)
  end
  dirname.chomp!('/')
  dirname.concat('/')
  dirname
end

Instance Method Details

#extract!(uri, html) ⇒ Object 



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
 
# File 'lib/rypper/extractor.rb', line 24

def extract!(uri, html)
  uri = uri.to_s
  unless html.kind_of?(Nokogiri::HTML::Document)
    html = Nokogiri::HTML(html.to_s)
  end
  res = []
  elems = html.search(self.selector)
  if elems.count == 1
    elem = elems.first
    if elem.name == 'img'
      res << elem[:src]
    elsif elem.name == 'a'
      res << elem[:href]
    else
      res << elem
    end
  else
    res = elems
  end
  res.map! do |elem|
    if elem.is_a?(String)
      elem_uri = ::URI.parse(elem)
      if !elem_uri.absolute?
        elem_uri = ::URI.join(self.dirname(uri), elem)
      elsif elem_uri.instance_of?(::URI::Generic) # absolute path only
        elem_uri = ::URI.parse(uri)
        elem_uri.path = elem
      end
      elem_uri.to_s
    else
      elem
    end
  end
  res
end