Class: RTesseract

Inherits:
Object
  • Object
show all
Defined in:
lib/rtesseract.rb,
lib/rtesseract/mixed.rb,
lib/rtesseract/errors.rb

Defined Under Namespace

Classes: ConversionError, ImageNotSelectedError, Mixed, TempFilesNotRemovedError

Constant Summary collapse

VERSION = 
'0.0.7'

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(src = "", options = {}) ⇒ RTesseract

Returns a new instance of RTesseract.



14
15
16
17
18
19
20
21
22
23
24
25
26
27
 
# File 'lib/rtesseract.rb', line 14

def initialize(src = "", options = {})
  @uid     = options.delete(:uid) || nil
  @source  = Pathname.new src
  @command = options.delete(:command) || "tesseract"
  @lang    = options.delete(:lang)    || options.delete("lang") || ""
  @psm     = options.delete(:psm)     || options.delete("psm")  || nil
  @clear_console_output = options.delete(:clear_console_output)
  @clear_console_output = true if @clear_console_output.nil?
  @options = options
  @value   = ""
  @x, @y, @w, @h = []
  @processor = options.delete(:processor) || options.delete("processor")
  choose_processor!
end

Instance Attribute Details

#lang ⇒ Object

Select the language

Languages

  • eng - English

  • deu - German

  • deu-f - German fraktur

  • fra - French

  • ita - Italian

  • nld - Dutch

  • por - Portuguese

  • spa - Spanish

  • vie - Vietnamese

Note: Make sure you have installed the language to tesseract



76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
 
# File 'lib/rtesseract.rb', line 76

def lang
  language = "#{@lang}".strip.downcase
  { #Aliases to languages names
    "eng" => ["en","en-us","english"],
    "ita" => ["it"],
    "por" => ["pt","pt-br","portuguese"],
    "spa" => ["sp"]
  }.each do |value,names|
    return " -l #{value} " if names.include? language
  end
  return " -l #{language} " if language.size > 0
  ""
rescue
  ""
end

#options ⇒ Object

Returns the value of attribute options.



9
10
11
 
# File 'lib/rtesseract.rb', line 9

def options
  @options
end

#processor ⇒ Object (readonly)

Returns the value of attribute processor.



12
13
14
 
# File 'lib/rtesseract.rb', line 12

def processor
  @processor
end

#psm ⇒ Object

Page Segment Mode



93
94
95
96
97
 
# File 'lib/rtesseract.rb', line 93

def psm
  @psm.nil? ? "" : " -psm #{@psm} "
rescue
  ""
end

Instance Method Details

#clear_console_output ⇒ Object

TODO: Clear console for MacOS or Windows



112
113
114
115
 
# File 'lib/rtesseract.rb', line 112

def clear_console_output
  return "" unless @clear_console_output
  return "2>/dev/null" if File.exist?("/dev/null") #Linux console clear
end

#config ⇒ Object 



99
100
101
102
 
# File 'lib/rtesseract.rb', line 99

def config
  @options ||= {}
  @options.collect{|k,v| "#{k} #{v}" }.join("\n")
end

#config_file ⇒ Object 



104
105
106
107
108
109
 
# File 'lib/rtesseract.rb', line 104

def config_file
  return "" if @options == {}
  conf = Tempfile.new("config")
  conf.write(config)
  conf.path
end

#convert ⇒ Object

Convert image to string



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

def convert
  generate_uid
  tmp_file  = Pathname.new(Dir::tmpdir).join("#{@uid}_#{@source.basename}")
  tmp_image = image_to_tiff
  `#{@command} #{tmp_image} #{tmp_file.to_s} #{lang} #{psm} #{config_file} #{clear_console_output}`
  @value = File.read("#{tmp_file.to_s}.txt").to_s
  @uid = nil
  remove_file([tmp_image,"#{tmp_file.to_s}.txt"])
rescue
  raise RTesseract::ConversionError
end

#crop!(x, y, width, height) ⇒ Object

Crop image to convert



40
41
42
43
 
# File 'lib/rtesseract.rb', line 40

def crop!(x,y,width,height)
  @x, @y, @w, @h = x, y, width, height
  self
end

#generate_uid ⇒ Object 



59
60
61
62
 
# File 'lib/rtesseract.rb', line 59

def generate_uid
  @uid = rand.to_s[2,10] if @uid.nil?
  @uid
end

#image_name ⇒ Object 



34
35
36
 
# File 'lib/rtesseract.rb', line 34

def image_name
  @source.basename
end

#remove_file(files = []) ⇒ Object

Remove files



46
47
48
49
50
51
52
53
54
55
56
57
 
# File 'lib/rtesseract.rb', line 46

def remove_file(files=[])
  files.each do |file|
    begin
      File.unlink(file) if File.exist?(file)
    rescue
      system "rm -f #{file}"
    end
  end
  true
rescue
  raise RTesseract::TempFilesNotRemovedError
end

#source=(src) ⇒ Object 



29
30
31
32
 
# File 'lib/rtesseract.rb', line 29

def source= src
  @value = ""
  @source = Pathname.new src
end

#to_s ⇒ Object

Output value



131
132
133
134
135
136
137
138
139
 
# File 'lib/rtesseract.rb', line 131

def to_s
  return @value if @value != ""
  if @source.file?
    convert
    @value
  else
    raise RTesseract::ImageNotSelectedError
  end
end

#to_s_without_spaces ⇒ Object

Remove spaces and break-lines



142
143
144
 
# File 'lib/rtesseract.rb', line 142

def to_s_without_spaces
  to_s.gsub(" ","").gsub("\n","").gsub("\r","")
end