Class: MHT::MhtmlGenerator

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

Overview

generate mhtml file

uri target uri

return mhtml file mhtml = MHT::MhtmlGenerator.generate(“rubygems.org/”) open(“output.mht”, “w+”){|f| f.write mhtml }

Instance Attribute Summary collapse

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize ⇒ MhtmlGenerator

Returns a new instance of MhtmlGenerator.



27
28
29
30
31
32
33
34
 
# File 'lib/mht.rb', line 27

def initialize
  @contents = {}
  @src = StringIO.new
  @boundary = "mimepart_#{Digest::MD5.hexdigest(Time.now.to_s)}"
  @threads  = []
  @queue    = Queue.new
  @conf     = { :base64_except=>["html"]}
end

Instance Attribute Details

#conf ⇒ Object

Returns the value of attribute conf.



22
23
24
 
# File 'lib/mht.rb', line 22

def conf
  @conf
end

Class Method Details

.generate(uri) ⇒ Object 



23
24
25
26
 
# File 'lib/mht.rb', line 23

def MhtmlGenerator.generate(uri)
  generateror = MhtmlGenerator.new
  return generateror.convert(uri)
end

Instance Method Details

#add_html_content(cid) ⇒ Object 



117
118
119
120
121
122
123
124
125
126
127
128
129
 
# File 'lib/mht.rb', line 117

def add_html_content(cid)
  filename = File.basename(URI(@contents[cid][:uri]).path)
  @src.puts "--#{@boundary}"
  @src.puts "Content-Disposition: inline; filename=" + filename 
  @src.puts "Content-Type: #{@contents[cid][:content_type]}"
  @src.puts "Content-Location: #{@contents[cid][:uri]}"
  @src.puts "Content-Transfer-Encoding: Base64"
  @src.puts "Content-Id: #{cid}"
  @src.puts ""
  @src.puts "#{Base64.encode64(@contents[cid][:body])}"
  @src.puts ""
   return
end

#attach_contents ⇒ Object 



108
109
110
111
112
113
114
115
116
 
# File 'lib/mht.rb', line 108

def attach_contents
  #prepeare_queue
  @contents.each{|k,v| @queue.push k}
  #start download threads
  self.start_download_thread
  # wait until download finished.
  @threads.each{|t|t.join}
  @contents.each{|k,v|self.add_html_content(k)}
end

#convert(filename_or_uri) ⇒ Object 



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
 
# File 'lib/mht.rb', line 35

def convert(filename_or_uri)
    f = open(filename_or_uri)
    html = f.read
    @parser = Nokogiri::HTML html
    @src.puts "Subject: " + @parser.search("title").text()
    @src.puts "Content-Type: multipart/related; boundary=#{@boundary}"
    @src.puts "Content-Location: #{filename_or_uri}"
    @src.puts "Date: #{Time.now.to_s}"
    @src.puts "MIME-Version: 1.0"
    @src.puts ""
    @src.puts "mime mhtml content"
    @src.puts ""
    #imgs
    @parser.search('img').each{|i| 
        uri = i.attr('src');
        uri = URI::join( filename_or_uri, uri).to_s
        uid = Digest::MD5.hexdigest(uri)
        @contents[uid] = uri
        i.set_attribute('src',"cid:#{uid}")
      }
    #styles
    @parser.search('link[rel=stylesheet]').each{|i|
        uri = i.attr('href');
        uri = URI::join( filename_or_uri, uri).to_s
        uid = Digest::MD5.hexdigest(uri)
        @contents[uid] = uri
        i.set_attribute('href',"cid:#{uid}")
      }
    #scripts
    @parser.search('script').map{ |i|
        next unless i.attr('src');
        uri = i.attr('src');
        uri = URI::join( filename_or_uri, uri).to_s
        uid = Digest::MD5.hexdigest(uri)
        @contents[uid] = uri
        i.set_attribute('src',"cid:#{uid}")
    }
    @src.puts "--#{@boundary}"
    @src.puts "Content-Disposition: inline; filename=default.htm"
    @src.puts "Content-Type: #{f.meta['content-type']}"
    @src.puts "Content-Id: #{Digest::MD5.hexdigest(filename_or_uri)}"
    @src.puts "Content-Location: #{filename_or_uri}"
    @src.puts "Content-Transfer-Encoding: 8bit" if @conf[:base64_except].find("html")
    @src.puts "Content-Transfer-Encoding: Base64" unless @conf[:base64_except].find("html")
    @src.puts ""
    #@src.puts html
    @src.puts "#{html}"                      if @conf[:base64_except].find("html")
    #@src.puts "#{Base64.encode64(html)}" unless @conf[:base64_except].find("html")
    @src.puts ""
    self.attach_contents
    @src.puts "--#{@boundary}--"
    @src.rewind
    return @src.read
end

#download_finished? ⇒ Boolean

Returns:

  • (Boolean) 


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

def download_finished?
  @contents.find{|k,v| v.class != Hash } == nil
end

#start_download_thread(num = 5) ⇒ Object 



89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
 
# File 'lib/mht.rb', line 89

def start_download_thread(num=5)
  num.times{
    t = Thread.start{
      while(@queue.empty? == false)
        k = @queue.pop
        v = @contents[k]
        next if v.class == Hash
        f = open(v)
        meta = f.meta
        @contents[k] = { :body=>f.read, :uri=> v, :content_type=> f.meta["content-type"] }
      end
    }
    @threads.push t
  }
  return @threads
end