Class: FiasParser::Parser

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

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(options = {}) ⇒ Parser

Returns a new instance of Parser.



53
54
55
 
# File 'lib/fias_parser.rb', line 53

def initialize( options = {} )
  @base_dir = options[:base_dir] || Dir.tmpdir
end

Instance Attribute Details

#date ⇒ Object

Returns the value of attribute date.



51
52
53
 
# File 'lib/fias_parser.rb', line 51

def date
  @date
end

Instance Method Details

#archive_file_path ⇒ Object 



156
157
158
 
# File 'lib/fias_parser.rb', line 156

def archive_file_path
  @archive_file_path ||= get_archive_file_path
end

#archive_path ⇒ Object 



148
149
150
 
# File 'lib/fias_parser.rb', line 148

def archive_path
  @archive_path ||= get_archive_path
end

#check_executables ⇒ Object 



78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
 
# File 'lib/fias_parser.rb', line 78

def check_executables
  line = Cocaine::CommandLine.new( "wget", "-h" )

  begin
    line.run
  rescue Cocaine::CommandNotFoundError => e
    puts 'wget is not installed. Run "sudo apt-get install wget"'

    return false
  end 

  line = Cocaine::CommandLine.new( "unar", "-h" )

  begin
    line.run
  rescue Cocaine::CommandNotFoundError => e
    puts 'unar is not installed. Run "sudo apt-get install unar"'

    return false
  end 

  true
end

#download_latest ⇒ Object 



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
127
128
129
130
131
132
133
 
# File 'lib/fias_parser.rb', line 102

def download_latest
  agent = Mechanize.new
  agent.user_agent = 'Mozilla/5.0 (Windows NT 6.1; WOW64; rv:10.0.1) Gecko/20100101 Firefox/10.0.1'

  url = 'http://fias.nalog.ru/Public/DownloadPage.aspx'

  doc = agent.get( url ).root

  post_data = {}

  doc.css( 'form input[type="hidden"]' ).each do |input|
    post_data[ input['name'] ] = input['value']
  end

  @date = doc.css( '#ctl00_contentPlaceHolder_downloadRadGrid_ctl00__0 tr td:first-child' ).text.gsub( /[^\d]/, '' )

  post_data['__EVENTTARGET'] = 'ctl00$contentPlaceHolder$downloadRadGrid$ctl00$ctl04$fullSZLinkButton'

  if File.exists?( self.archive_file_path )
    puts "File '#{self.archive_file_path}' already exists."        
    
    return
  end

  line = Cocaine::CommandLine.new( "wget", "--output-document=:out --post-data=:post_data :url" )

  line.run( {
    out: self.archive_file_path,
    post_data: Mechanize::Util.build_query_string( post_data ),
    url: url,
  } )
end

#get_archive_file_path ⇒ Object 



160
161
162
 
# File 'lib/fias_parser.rb', line 160

def get_archive_file_path
  File.join( @base_dir,  "#{@date}.rar" )
end

#get_archive_path ⇒ Object 



152
153
154
 
# File 'lib/fias_parser.rb', line 152

def get_archive_path
  File.join( @base_dir, @date )
end

#get_latest ⇒ Object 



57
58
59
60
61
62
 
# File 'lib/fias_parser.rb', line 57

def get_latest
  return unless self.check_executables

  self.download_latest
  self.unpack
end

#process(term, options = {}, &block) ⇒ Object 



64
65
66
67
68
69
70
71
72
73
74
75
76
 
# File 'lib/fias_parser.rb', line 64

def process( term, options = {}, &block ) 
  file_name = Dir.entries( self.archive_path ).find { |f| f =~ /#{term}/i }

  if file_name.nil?
    puts "File /#{term}/ not found in '#{self.archive_path}'"

    return
  end

  File.open( File.join( self.archive_path, file_name ) ) do |file|
    ::Ox.sax_parse( Document.new( options, block ), file )
  end
end

#unpack ⇒ Object 



135
136
137
138
139
140
141
142
143
144
145
146
 
# File 'lib/fias_parser.rb', line 135

def unpack
  dir = File.join( @base_dir )

  FileUtils.mkdir_p ( dir ) unless File.exists?( dir )

  line = Cocaine::CommandLine.new( "unar", "-o :dir :archive" )

  line.run( {
    dir: dir,
    archive: self.archive_file_path,
  } )
end