Class: Doko

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

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(str) ⇒ Doko

Returns a new instance of Doko.



39
40
41
42
43
44
45
46
47
48
49
50
 
# File 'lib/doko.rb', line 39

def initialize(str)
  if str.kind_of? URI 
    str = open(str.to_s).read
  elsif str.match( /^#{URI.regexp}$/ )
    str = open(str).read
  end
  if str.match(/<html/i)
    @text =  (Nokogiri::HTML(str)/"body").text
  else
    @text = str
  end
end

Class Method Details

.deep(str, base_uri = nil) ⇒ Object 



12
13
14
15
16
17
18
19
20
 
# File 'lib/doko.rb', line 12

def self.deep(str,base_uri=nil)
  addrs = parse(str)
  if addrs.empty?
    addrs = links(str,base_uri).map{ |u|
      parse(u)
    }.flatten
  end
  addrs
end

.links(str, base_uri = nil) ⇒ Object 



22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
 
# File 'lib/doko.rb', line 22

def self.links(str,base_uri=nil)
  out = []
  if str.match( /^#{URI.regexp}$/ )
    uri = URI.parse(str)
    doc = Nokogiri::HTML(open(uri).read)
  elsif str.kind_of? String
    uri = URI.parse(base_uri)
    doc = Nokogiri::HTML(str)
  end
  doc.search("a").each do |a|
    if a[:href] && a[:href].match(/access/) && !a[:href].match(/http/)
      out << uri + a[:href]
    end
  end
  out.uniq
end

.parse(str) ⇒ Object 



8
9
10
 
# File 'lib/doko.rb', line 8

def self.parse(str)
  self.new(str).parse
end

Instance Method Details

#parse ⇒ Object 



52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
 
# File 'lib/doko.rb', line 52

def parse
  body = @text
  body.tr!("０-９","0-9")
  body.tr!("（）","()")
  body.tr!("、",",")
  body.tr!("　"," ")
  body.tr!("．",".")
  body.tr!("：",":")
  blackchars =  ",()\n"

  addrs = body.scan(/\b([^\s,()]{2,3}(都|道|府|県)[^\s,()]{1,8}(市|区|町|村)[^#{blackchars}]+)/).map{ |m|
    clean(m[0])
  }
  if addrs.empty?
    addrs = body.scan(/([^\s]{1,6}(市|区).{1,8}(区|町|村)[^\s,()]{2,10}\d)/).map{ |m|
      clean(m[0])
    }
  end
  addrs.select{ |a|
    !a.match(/を/)
  }
end