mini_tesseract 0.0.7
Sign up to get free protection for your applications and to get access to all the features.
- data/.document +5 -0
- data/LICENSE +20 -0
- data/README.rdoc +61 -0
- data/Rakefile +58 -0
- data/lib/processors/mini_magick.rb +12 -0
- data/lib/processors/rmagick.rb +12 -0
- data/lib/rtesseract.rb +157 -0
- data/lib/rtesseract/errors.rb +6 -0
- data/lib/rtesseract/mixed.rb +53 -0
- data/mini_tesseract.gemspec +67 -0
- data/rtesseract.gemspec +66 -0
- data/test/helper.rb +11 -0
- data/test/images/mixed.tif +0 -0
- data/test/images/test.bmp +0 -0
- data/test/images/test.jpg +0 -0
- data/test/images/test.png +0 -0
- data/test/images/test.tif +0 -0
- data/test/images/test1.tif +0 -0
- data/test/test_mixed.rb +36 -0
- data/test/test_rtesseract.rb +82 -0
- metadata +120 -0
data/.document
ADDED
data/LICENSE
ADDED
@@ -0,0 +1,20 @@
|
|
1
|
+
Copyright (c) 2010 Danilo Jeremias da Silva
|
2
|
+
|
3
|
+
Permission is hereby granted, free of charge, to any person obtaining
|
4
|
+
a copy of this software and associated documentation files (the
|
5
|
+
"Software"), to deal in the Software without restriction, including
|
6
|
+
without limitation the rights to use, copy, modify, merge, publish,
|
7
|
+
distribute, sublicense, and/or sell copies of the Software, and to
|
8
|
+
permit persons to whom the Software is furnished to do so, subject to
|
9
|
+
the following conditions:
|
10
|
+
|
11
|
+
The above copyright notice and this permission notice shall be
|
12
|
+
included in all copies or substantial portions of the Software.
|
13
|
+
|
14
|
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
|
15
|
+
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
|
16
|
+
MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
|
17
|
+
NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
|
18
|
+
LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
|
19
|
+
OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
|
20
|
+
WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
|
data/README.rdoc
ADDED
@@ -0,0 +1,61 @@
|
|
1
|
+
= rtesseract
|
2
|
+
|
3
|
+
Ruby library for working with the Tesseract OCR.
|
4
|
+
|
5
|
+
== REQUIREMENTS:
|
6
|
+
|
7
|
+
To work properly rtesseract are needed:
|
8
|
+
* Tesseract - Program
|
9
|
+
* ImageMagic - Program
|
10
|
+
* RMagick - Gem
|
11
|
+
|
12
|
+
== EXAMPLE USAGE
|
13
|
+
|
14
|
+
It's very simple to use rtesseract:
|
15
|
+
|
16
|
+
=== CONVERT IMAGE TO STRING
|
17
|
+
|
18
|
+
image = RTesseract.new("my_image.jpg")
|
19
|
+
image.to_s #Getting the value
|
20
|
+
|
21
|
+
=== CHANGE THE IMAGE
|
22
|
+
|
23
|
+
image = RTesseract.new("my_image.jpg")
|
24
|
+
image.source = "new_image.png"
|
25
|
+
image.to_s
|
26
|
+
|
27
|
+
=== CONVERT PARTS OF IMAGE TO STRING
|
28
|
+
|
29
|
+
mix_block = RTesseract::Mixed.new("test.jpg") do |image|
|
30
|
+
image.area(28, 19, 25, 25)
|
31
|
+
image.area(180, 22, 20, 28)
|
32
|
+
image.area(218, 22, 24, 28)
|
33
|
+
image.area(248, 24, 22, 22)
|
34
|
+
end
|
35
|
+
mix_block.to_s
|
36
|
+
|
37
|
+
OR
|
38
|
+
|
39
|
+
mix_block = RTesseract::Mixed.new("test.jpg",{:areas => [
|
40
|
+
{:x => 28, :y=>19, :width=>25, :height=>25 },
|
41
|
+
{:x => 180, :y=>22, :width=>20, :height=>28},
|
42
|
+
{:x => 218, :y=>22, :width=>24, :height=>28},
|
43
|
+
{:x => 248, :y=>24, :width=>22, :height=>22}
|
44
|
+
]})
|
45
|
+
mix_block.to_s
|
46
|
+
|
47
|
+
== Note on Patches/Pull Requests
|
48
|
+
|
49
|
+
* Fork the project.
|
50
|
+
* Make your feature addition or bug fix.
|
51
|
+
* Send me a pull request. Bonus points for topic branches.
|
52
|
+
|
53
|
+
== Links
|
54
|
+
|
55
|
+
* Github - http://github.com/dannnylo/rtesseract
|
56
|
+
* Rubygems - http://rubygems.org/gems/rtesseract
|
57
|
+
|
58
|
+
== Copyright
|
59
|
+
|
60
|
+
Copyright (c) 2010 Danilo Jeremias da Silva. See LICENSE for details.
|
61
|
+
|
data/Rakefile
ADDED
@@ -0,0 +1,58 @@
|
|
1
|
+
require 'rubygems'
|
2
|
+
require 'rake'
|
3
|
+
|
4
|
+
begin
|
5
|
+
require 'jeweler'
|
6
|
+
Jeweler::Tasks.new do |gem|
|
7
|
+
gem.name = "mini_tesseract"
|
8
|
+
gem.version = '0.0.7'
|
9
|
+
gem.summary = "Ruby library for working with the Tesseract OCR."
|
10
|
+
gem.description = "Ruby library for working with the Tesseract OCR."
|
11
|
+
gem.email = "hoooopo@gmail.com"
|
12
|
+
gem.homepage = "http://github.com/hooopo/rtesseract"
|
13
|
+
gem.authors = ["Danilo Jeremias da Silva", "Hooopo"]
|
14
|
+
gem.add_development_dependency "jeweler", ">=1.4.0"
|
15
|
+
gem.add_development_dependency "thoughtbot-shoulda", ">= 0"
|
16
|
+
|
17
|
+
gem.add_runtime_dependency "rmagick", '>= 2.10.1'
|
18
|
+
gem.add_runtime_dependency "mini_magick", '~> 3.0'
|
19
|
+
end
|
20
|
+
Jeweler::GemcutterTasks.new
|
21
|
+
rescue LoadError
|
22
|
+
puts "Jeweler (or a dependency) not available. Install it with: gem install jeweler"
|
23
|
+
end
|
24
|
+
|
25
|
+
require 'rake/testtask'
|
26
|
+
Rake::TestTask.new(:test) do |test|
|
27
|
+
test.libs << 'lib' << 'test'
|
28
|
+
test.pattern = 'test/**/test_*.rb'
|
29
|
+
test.verbose = true
|
30
|
+
end
|
31
|
+
|
32
|
+
begin
|
33
|
+
require 'rcov/rcovtask'
|
34
|
+
Rcov::RcovTask.new do |test|
|
35
|
+
test.libs << 'test'
|
36
|
+
test.pattern = 'test/**/test_*.rb'
|
37
|
+
test.verbose = true
|
38
|
+
end
|
39
|
+
rescue LoadError
|
40
|
+
task :rcov do
|
41
|
+
abort "RCov is not available. In order to run rcov, you must: sudo gem install spicycode-rcov"
|
42
|
+
end
|
43
|
+
end
|
44
|
+
|
45
|
+
task :test => :check_dependencies
|
46
|
+
|
47
|
+
task :default => :test
|
48
|
+
|
49
|
+
require 'rake/rdoctask'
|
50
|
+
Rake::RDocTask.new do |rdoc|
|
51
|
+
version = File.exist?('VERSION') ? File.read('VERSION') : ""
|
52
|
+
|
53
|
+
rdoc.rdoc_dir = 'rdoc'
|
54
|
+
rdoc.title = "rtesseract #{version}"
|
55
|
+
rdoc.rdoc_files.include('README*')
|
56
|
+
rdoc.rdoc_files.include('lib/**/*.rb')
|
57
|
+
end
|
58
|
+
|
@@ -0,0 +1,12 @@
|
|
1
|
+
require 'mini_magick'
|
2
|
+
module MiniMagickProcessor
|
3
|
+
def image_to_tiff
|
4
|
+
generate_uid
|
5
|
+
tmp_file = Pathname.new(Dir::tmpdir).join("#{@uid}_#{@source.basename}.tif").to_s
|
6
|
+
cat = MiniMagick::Image.open(@source.to_s)
|
7
|
+
cat.format("tif")
|
8
|
+
cat.crop("#{@w}x#{@h}+#{@x}+#{@y}") unless [@x, @y, @w, @h].compact == []
|
9
|
+
cat.write tmp_file.to_s
|
10
|
+
return tmp_file
|
11
|
+
end
|
12
|
+
end
|
@@ -0,0 +1,12 @@
|
|
1
|
+
require 'rmagick'
|
2
|
+
module RMagickProcessor
|
3
|
+
def image_to_tiff
|
4
|
+
puts "processing with rmagick"
|
5
|
+
generate_uid
|
6
|
+
tmp_file = Pathname.new(Dir::tmpdir).join("#{@uid}_#{@source.basename}.tif").to_s
|
7
|
+
cat = Magick::Image.read(@source.to_s).first
|
8
|
+
cat.crop!(@x, @y, @w, @h) unless [@x, @y, @w, @h].compact == []
|
9
|
+
cat.write tmp_file.to_s
|
10
|
+
return tmp_file
|
11
|
+
end
|
12
|
+
end
|
data/lib/rtesseract.rb
ADDED
@@ -0,0 +1,157 @@
|
|
1
|
+
require "pathname"
|
2
|
+
require "tempfile"
|
3
|
+
|
4
|
+
require "rtesseract/errors"
|
5
|
+
require "rtesseract/mixed"
|
6
|
+
|
7
|
+
class RTesseract
|
8
|
+
VERSION = '0.0.7'
|
9
|
+
attr_accessor :options
|
10
|
+
attr_writer :lang
|
11
|
+
attr_writer :psm
|
12
|
+
attr_reader :processor
|
13
|
+
|
14
|
+
def initialize(src = "", options = {})
|
15
|
+
@uid = options.delete(:uid) || nil
|
16
|
+
@source = Pathname.new src
|
17
|
+
@command = options.delete(:command) || "tesseract"
|
18
|
+
@lang = options.delete(:lang) || options.delete("lang") || ""
|
19
|
+
@psm = options.delete(:psm) || options.delete("psm") || nil
|
20
|
+
@clear_console_output = options.delete(:clear_console_output)
|
21
|
+
@clear_console_output = true if @clear_console_output.nil?
|
22
|
+
@options = options
|
23
|
+
@value = ""
|
24
|
+
@x, @y, @w, @h = []
|
25
|
+
@processor = options.delete(:processor) || options.delete("processor")
|
26
|
+
choose_processor!
|
27
|
+
end
|
28
|
+
|
29
|
+
def source= src
|
30
|
+
@value = ""
|
31
|
+
@source = Pathname.new src
|
32
|
+
end
|
33
|
+
|
34
|
+
def image_name
|
35
|
+
@source.basename
|
36
|
+
end
|
37
|
+
|
38
|
+
|
39
|
+
#Crop image to convert
|
40
|
+
def crop!(x,y,width,height)
|
41
|
+
@x, @y, @w, @h = x, y, width, height
|
42
|
+
self
|
43
|
+
end
|
44
|
+
|
45
|
+
#Remove files
|
46
|
+
def remove_file(files=[])
|
47
|
+
files.each do |file|
|
48
|
+
begin
|
49
|
+
File.unlink(file) if File.exist?(file)
|
50
|
+
rescue
|
51
|
+
system "rm -f #{file}"
|
52
|
+
end
|
53
|
+
end
|
54
|
+
true
|
55
|
+
rescue
|
56
|
+
raise RTesseract::TempFilesNotRemovedError
|
57
|
+
end
|
58
|
+
|
59
|
+
def generate_uid
|
60
|
+
@uid = rand.to_s[2,10] if @uid.nil?
|
61
|
+
@uid
|
62
|
+
end
|
63
|
+
|
64
|
+
# Select the language
|
65
|
+
#===Languages
|
66
|
+
## * eng - English
|
67
|
+
## * deu - German
|
68
|
+
## * deu-f - German fraktur
|
69
|
+
## * fra - French
|
70
|
+
## * ita - Italian
|
71
|
+
## * nld - Dutch
|
72
|
+
## * por - Portuguese
|
73
|
+
## * spa - Spanish
|
74
|
+
## * vie - Vietnamese
|
75
|
+
## Note: Make sure you have installed the language to tesseract
|
76
|
+
def lang
|
77
|
+
language = "#{@lang}".strip.downcase
|
78
|
+
{ #Aliases to languages names
|
79
|
+
"eng" => ["en","en-us","english"],
|
80
|
+
"ita" => ["it"],
|
81
|
+
"por" => ["pt","pt-br","portuguese"],
|
82
|
+
"spa" => ["sp"]
|
83
|
+
}.each do |value,names|
|
84
|
+
return " -l #{value} " if names.include? language
|
85
|
+
end
|
86
|
+
return " -l #{language} " if language.size > 0
|
87
|
+
""
|
88
|
+
rescue
|
89
|
+
""
|
90
|
+
end
|
91
|
+
|
92
|
+
#Page Segment Mode
|
93
|
+
def psm
|
94
|
+
@psm.nil? ? "" : " -psm #{@psm} "
|
95
|
+
rescue
|
96
|
+
""
|
97
|
+
end
|
98
|
+
|
99
|
+
def config
|
100
|
+
@options ||= {}
|
101
|
+
@options.collect{|k,v| "#{k} #{v}" }.join("\n")
|
102
|
+
end
|
103
|
+
|
104
|
+
def config_file
|
105
|
+
return "" if @options == {}
|
106
|
+
conf = Tempfile.new("config")
|
107
|
+
conf.write(config)
|
108
|
+
conf.path
|
109
|
+
end
|
110
|
+
|
111
|
+
#TODO: Clear console for MacOS or Windows
|
112
|
+
def clear_console_output
|
113
|
+
return "" unless @clear_console_output
|
114
|
+
return "2>/dev/null" if File.exist?("/dev/null") #Linux console clear
|
115
|
+
end
|
116
|
+
|
117
|
+
#Convert image to string
|
118
|
+
def convert
|
119
|
+
generate_uid
|
120
|
+
tmp_file = Pathname.new(Dir::tmpdir).join("#{@uid}_#{@source.basename}")
|
121
|
+
tmp_image = image_to_tiff
|
122
|
+
`#{@command} #{tmp_image} #{tmp_file.to_s} #{lang} #{psm} #{config_file} #{clear_console_output}`
|
123
|
+
@value = File.read("#{tmp_file.to_s}.txt").to_s
|
124
|
+
@uid = nil
|
125
|
+
remove_file([tmp_image,"#{tmp_file.to_s}.txt"])
|
126
|
+
rescue
|
127
|
+
raise RTesseract::ConversionError
|
128
|
+
end
|
129
|
+
|
130
|
+
#Output value
|
131
|
+
def to_s
|
132
|
+
return @value if @value != ""
|
133
|
+
if @source.file?
|
134
|
+
convert
|
135
|
+
@value
|
136
|
+
else
|
137
|
+
raise RTesseract::ImageNotSelectedError
|
138
|
+
end
|
139
|
+
end
|
140
|
+
|
141
|
+
#Remove spaces and break-lines
|
142
|
+
def to_s_without_spaces
|
143
|
+
to_s.gsub(" ","").gsub("\n","").gsub("\r","")
|
144
|
+
end
|
145
|
+
|
146
|
+
private
|
147
|
+
def choose_processor!
|
148
|
+
if @processor.to_s == "mini_magick"
|
149
|
+
require File.expand_path(File.dirname(__FILE__) + "/processors/mini_magick.rb")
|
150
|
+
self.class.send(:include, MiniMagickProcessor)
|
151
|
+
else
|
152
|
+
require File.expand_path(File.dirname(__FILE__) + "/processors/rmagick.rb")
|
153
|
+
self.class.send(:include, RMagickProcessor)
|
154
|
+
end
|
155
|
+
end
|
156
|
+
end
|
157
|
+
|
@@ -0,0 +1,53 @@
|
|
1
|
+
class RTesseract
|
2
|
+
class Mixed
|
3
|
+
def initialize(src="", options={})
|
4
|
+
@source = Pathname.new src
|
5
|
+
@options = options
|
6
|
+
@value = ""
|
7
|
+
@areas = options.delete(:areas) || []
|
8
|
+
yield self if block_given?
|
9
|
+
end
|
10
|
+
|
11
|
+
def area(x, y, width, height)
|
12
|
+
@value = ""
|
13
|
+
@areas << {:x => x, :y => y, :width => width, :height => height}
|
14
|
+
end
|
15
|
+
|
16
|
+
def areas
|
17
|
+
@areas
|
18
|
+
end
|
19
|
+
|
20
|
+
def clear_areas
|
21
|
+
@areas = []
|
22
|
+
end
|
23
|
+
|
24
|
+
#Convert parts of image to string
|
25
|
+
def convert
|
26
|
+
@value = ""
|
27
|
+
@areas.each do |area|
|
28
|
+
image = RTesseract.new(@source.to_s,@options)
|
29
|
+
image.crop!(area[:x].to_i, area[:y].to_i, area[:width].to_i, area[:height].to_i)
|
30
|
+
@value << image.to_s
|
31
|
+
end
|
32
|
+
rescue
|
33
|
+
raise RTesseract::ConversionError
|
34
|
+
end
|
35
|
+
|
36
|
+
#Output value
|
37
|
+
def to_s
|
38
|
+
return @value if @value != ""
|
39
|
+
if @source.file?
|
40
|
+
convert
|
41
|
+
@value
|
42
|
+
else
|
43
|
+
raise RTesseract::ImageNotSelectedError
|
44
|
+
end
|
45
|
+
end
|
46
|
+
|
47
|
+
#Remove spaces and break-lines
|
48
|
+
def to_s_without_spaces
|
49
|
+
to_s.gsub(" ","").gsub("\n","").gsub("\r","")
|
50
|
+
end
|
51
|
+
end
|
52
|
+
end
|
53
|
+
|
@@ -0,0 +1,67 @@
|
|
1
|
+
# Generated by jeweler
|
2
|
+
# DO NOT EDIT THIS FILE DIRECTLY
|
3
|
+
# Instead, edit Jeweler::Tasks in Rakefile, and run 'rake gemspec'
|
4
|
+
# -*- encoding: utf-8 -*-
|
5
|
+
|
6
|
+
Gem::Specification.new do |s|
|
7
|
+
s.name = %q{mini_tesseract}
|
8
|
+
s.version = "0.0.7"
|
9
|
+
|
10
|
+
s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
|
11
|
+
s.authors = ["Danilo Jeremias da Silva", "Hooopo"]
|
12
|
+
s.date = %q{2011-08-23}
|
13
|
+
s.description = %q{Ruby library for working with the Tesseract OCR.}
|
14
|
+
s.email = %q{hoooopo@gmail.com}
|
15
|
+
s.extra_rdoc_files = [
|
16
|
+
"LICENSE",
|
17
|
+
"README.rdoc"
|
18
|
+
]
|
19
|
+
s.files = [
|
20
|
+
".document",
|
21
|
+
"LICENSE",
|
22
|
+
"README.rdoc",
|
23
|
+
"Rakefile",
|
24
|
+
"lib/processors/mini_magick.rb",
|
25
|
+
"lib/processors/rmagick.rb",
|
26
|
+
"lib/rtesseract.rb",
|
27
|
+
"lib/rtesseract/errors.rb",
|
28
|
+
"lib/rtesseract/mixed.rb",
|
29
|
+
"mini_tesseract.gemspec",
|
30
|
+
"rtesseract.gemspec",
|
31
|
+
"test/helper.rb",
|
32
|
+
"test/images/mixed.tif",
|
33
|
+
"test/images/test.bmp",
|
34
|
+
"test/images/test.jpg",
|
35
|
+
"test/images/test.png",
|
36
|
+
"test/images/test.tif",
|
37
|
+
"test/images/test1.tif",
|
38
|
+
"test/test_mixed.rb",
|
39
|
+
"test/test_rtesseract.rb"
|
40
|
+
]
|
41
|
+
s.homepage = %q{http://github.com/hooopo/rtesseract}
|
42
|
+
s.require_paths = ["lib"]
|
43
|
+
s.rubygems_version = %q{1.3.9.2}
|
44
|
+
s.summary = %q{Ruby library for working with the Tesseract OCR.}
|
45
|
+
|
46
|
+
if s.respond_to? :specification_version then
|
47
|
+
s.specification_version = 3
|
48
|
+
|
49
|
+
if Gem::Version.new(Gem::VERSION) >= Gem::Version.new('1.2.0') then
|
50
|
+
s.add_development_dependency(%q<jeweler>, [">= 1.4.0"])
|
51
|
+
s.add_development_dependency(%q<thoughtbot-shoulda>, [">= 0"])
|
52
|
+
s.add_runtime_dependency(%q<rmagick>, [">= 2.10.1"])
|
53
|
+
s.add_runtime_dependency(%q<mini_magick>, ["~> 3.0"])
|
54
|
+
else
|
55
|
+
s.add_dependency(%q<jeweler>, [">= 1.4.0"])
|
56
|
+
s.add_dependency(%q<thoughtbot-shoulda>, [">= 0"])
|
57
|
+
s.add_dependency(%q<rmagick>, [">= 2.10.1"])
|
58
|
+
s.add_dependency(%q<mini_magick>, ["~> 3.0"])
|
59
|
+
end
|
60
|
+
else
|
61
|
+
s.add_dependency(%q<jeweler>, [">= 1.4.0"])
|
62
|
+
s.add_dependency(%q<thoughtbot-shoulda>, [">= 0"])
|
63
|
+
s.add_dependency(%q<rmagick>, [">= 2.10.1"])
|
64
|
+
s.add_dependency(%q<mini_magick>, ["~> 3.0"])
|
65
|
+
end
|
66
|
+
end
|
67
|
+
|
data/rtesseract.gemspec
ADDED
@@ -0,0 +1,66 @@
|
|
1
|
+
# Generated by jeweler
|
2
|
+
# DO NOT EDIT THIS FILE DIRECTLY
|
3
|
+
# Instead, edit Jeweler::Tasks in Rakefile, and run 'rake gemspec'
|
4
|
+
# -*- encoding: utf-8 -*-
|
5
|
+
|
6
|
+
Gem::Specification.new do |s|
|
7
|
+
s.name = %q{rtesseract}
|
8
|
+
s.version = "0.0.7"
|
9
|
+
|
10
|
+
s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
|
11
|
+
s.authors = ["Danilo Jeremias da Silva"]
|
12
|
+
s.date = %q{2011-03-16}
|
13
|
+
s.description = %q{Ruby library for working with the Tesseract OCR.}
|
14
|
+
s.email = %q{dannnylo@gmail.com}
|
15
|
+
s.extra_rdoc_files = [
|
16
|
+
"LICENSE",
|
17
|
+
"README.rdoc"
|
18
|
+
]
|
19
|
+
s.files = [
|
20
|
+
".document",
|
21
|
+
"LICENSE",
|
22
|
+
"README.rdoc",
|
23
|
+
"Rakefile",
|
24
|
+
"lib/rtesseract.rb",
|
25
|
+
"lib/rtesseract/errors.rb",
|
26
|
+
"lib/rtesseract/mixed.rb",
|
27
|
+
"rtesseract.gemspec",
|
28
|
+
"test/helper.rb",
|
29
|
+
"test/images/mixed.tif",
|
30
|
+
"test/images/test.bmp",
|
31
|
+
"test/images/test.jpg",
|
32
|
+
"test/images/test.png",
|
33
|
+
"test/images/test.tif",
|
34
|
+
"test/images/test1.tif",
|
35
|
+
"test/test_mixed.rb",
|
36
|
+
"test/test_rtesseract.rb"
|
37
|
+
]
|
38
|
+
s.homepage = %q{http://github.com/dannnylo/rtesseract}
|
39
|
+
s.require_paths = ["lib"]
|
40
|
+
s.rubygems_version = %q{1.4.2}
|
41
|
+
s.summary = %q{Ruby library for working with the Tesseract OCR.}
|
42
|
+
s.test_files = [
|
43
|
+
"test/helper.rb",
|
44
|
+
"test/test_mixed.rb",
|
45
|
+
"test/test_rtesseract.rb"
|
46
|
+
]
|
47
|
+
|
48
|
+
if s.respond_to? :specification_version then
|
49
|
+
s.specification_version = 3
|
50
|
+
|
51
|
+
if Gem::Version.new(Gem::VERSION) >= Gem::Version.new('1.2.0') then
|
52
|
+
s.add_development_dependency(%q<jeweler>, [">= 1.4.0"])
|
53
|
+
s.add_development_dependency(%q<thoughtbot-shoulda>, [">= 0"])
|
54
|
+
s.add_runtime_dependency(%q<rmagick>, [">= 2.10.1"])
|
55
|
+
else
|
56
|
+
s.add_dependency(%q<jeweler>, [">= 1.4.0"])
|
57
|
+
s.add_dependency(%q<thoughtbot-shoulda>, [">= 0"])
|
58
|
+
s.add_dependency(%q<rmagick>, [">= 2.10.1"])
|
59
|
+
end
|
60
|
+
else
|
61
|
+
s.add_dependency(%q<jeweler>, [">= 1.4.0"])
|
62
|
+
s.add_dependency(%q<thoughtbot-shoulda>, [">= 0"])
|
63
|
+
s.add_dependency(%q<rmagick>, [">= 2.10.1"])
|
64
|
+
end
|
65
|
+
end
|
66
|
+
|
data/test/helper.rb
ADDED
Binary file
|
Binary file
|
Binary file
|
Binary file
|
Binary file
|
Binary file
|
data/test/test_mixed.rb
ADDED
@@ -0,0 +1,36 @@
|
|
1
|
+
require 'helper'
|
2
|
+
require 'pathname'
|
3
|
+
|
4
|
+
class TestMixed < Test::Unit::TestCase
|
5
|
+
context "Path" do
|
6
|
+
setup do
|
7
|
+
@path = Pathname.new(__FILE__.gsub("test_mixed.rb","")).expand_path
|
8
|
+
@image_tiff = @path.join("images","mixed.tif").to_s
|
9
|
+
@image2_tiff = @path.join("images","mixed2.tif").to_s
|
10
|
+
end
|
11
|
+
|
12
|
+
should "be instantiable" do
|
13
|
+
assert_equal RTesseract::Mixed.new.class , RTesseract::Mixed
|
14
|
+
assert_equal RTesseract::Mixed.new(@image_tiff).class , RTesseract::Mixed
|
15
|
+
end
|
16
|
+
|
17
|
+
should "translate parts of the image to text" do
|
18
|
+
mix_block = RTesseract::Mixed.new(@image_tiff) do |image|
|
19
|
+
image.area(28, 19, 25, 25) #position of 4
|
20
|
+
image.area(180, 22, 20, 28) # position of 3
|
21
|
+
image.area(218, 22, 24, 28) # position of z
|
22
|
+
image.area(248, 24, 22, 22) # position of z
|
23
|
+
end
|
24
|
+
assert_equal mix_block.to_s_without_spaces , "43ZZ"
|
25
|
+
|
26
|
+
mix_block = RTesseract::Mixed.new(@image_tiff,{:areas => [
|
27
|
+
{:x => 28, :y=>19, :width=>25, :height=>25 }, #position of 4
|
28
|
+
{:x => 180, :y=>22, :width=>20, :height=>28}, # position of 3
|
29
|
+
{:x => 218, :y=>22, :width=>24, :height=>28}, # position of z
|
30
|
+
{:x => 248, :y=>24, :width=>22, :height=>22} # position of z
|
31
|
+
]})
|
32
|
+
assert_equal mix_block.to_s_without_spaces , "43ZZ"
|
33
|
+
end
|
34
|
+
end
|
35
|
+
end
|
36
|
+
|
@@ -0,0 +1,82 @@
|
|
1
|
+
require 'helper'
|
2
|
+
require 'pathname'
|
3
|
+
|
4
|
+
class TestRtesseract < Test::Unit::TestCase
|
5
|
+
context "Path" do
|
6
|
+
setup do
|
7
|
+
@path = Pathname.new(__FILE__.gsub("test_rtesseract.rb","")).expand_path
|
8
|
+
@image_tiff = @path.join("images","test.tif").to_s
|
9
|
+
end
|
10
|
+
|
11
|
+
should "be instantiable" do
|
12
|
+
assert_equal RTesseract.new.class , RTesseract
|
13
|
+
assert_equal RTesseract.new("").class , RTesseract
|
14
|
+
assert_equal RTesseract.new(@image_tiff).class , RTesseract
|
15
|
+
end
|
16
|
+
|
17
|
+
should "translate image to text" do
|
18
|
+
assert_equal RTesseract.new(@image_tiff).to_s_without_spaces , "43ZZ"
|
19
|
+
assert_equal RTesseract.new(@path.join("images","test1.tif").to_s).to_s_without_spaces , "V2V4"
|
20
|
+
end
|
21
|
+
|
22
|
+
should "translate images .png, .jpg, .bmp" do
|
23
|
+
assert_equal RTesseract.new(@path.join("images","test.png").to_s).to_s_without_spaces , "HW9W"
|
24
|
+
assert_equal RTesseract.new(@path.join("images","test.jpg").to_s).to_s_without_spaces , "3R8Z"
|
25
|
+
assert_equal RTesseract.new(@path.join("images","test.bmp").to_s).to_s_without_spaces , "ZLA6"
|
26
|
+
end
|
27
|
+
|
28
|
+
should "change the image" do
|
29
|
+
image = RTesseract.new(@image_tiff)
|
30
|
+
assert_equal image.to_s_without_spaces,"43ZZ"
|
31
|
+
image.source = @path.join("images","test1.tif").to_s
|
32
|
+
assert_equal image.to_s_without_spaces,"V2V4"
|
33
|
+
end
|
34
|
+
|
35
|
+
should "select the language" do
|
36
|
+
#English
|
37
|
+
assert_equal RTesseract.new(@image_tiff,{:lang=>"eng"}).lang , " -l eng "
|
38
|
+
assert_equal RTesseract.new(@image_tiff,{:lang=>"en"}).lang , " -l eng "
|
39
|
+
assert_equal RTesseract.new(@image_tiff,{:lang=>"en-US"}).lang , " -l eng "
|
40
|
+
assert_equal RTesseract.new(@image_tiff,{:lang=>"english"}).lang , " -l eng "
|
41
|
+
|
42
|
+
#Portuguese
|
43
|
+
assert_equal RTesseract.new(@image_tiff,{:lang=>"por"}).lang , " -l por "
|
44
|
+
assert_equal RTesseract.new(@image_tiff,{:lang=>"pt-BR"}).lang , " -l por "
|
45
|
+
assert_equal RTesseract.new(@image_tiff,{:lang=>"pt-br"}).lang , " -l por "
|
46
|
+
assert_equal RTesseract.new(@image_tiff,{:lang=>"pt"}).lang , " -l por "
|
47
|
+
assert_equal RTesseract.new(@image_tiff,{:lang=>"portuguese"}).lang , " -l por "
|
48
|
+
|
49
|
+
assert_equal RTesseract.new(@image_tiff,{:lang=>"eng"}).to_s_without_spaces , "43ZZ"
|
50
|
+
assert_equal RTesseract.new(@image_tiff,{:lang=>"por"}).to_s_without_spaces , "43ZZ"
|
51
|
+
|
52
|
+
assert_equal RTesseract.new(@image_tiff,{:lang=>"eng"}).lang , " -l eng "
|
53
|
+
end
|
54
|
+
|
55
|
+
should "be configurable" do
|
56
|
+
assert_equal RTesseract.new(@image_tiff,{:chop_enable=>0,:enable_assoc=>0,:display_text=>0}).config , "chop_enable 0\nenable_assoc 0\ndisplay_text 0"
|
57
|
+
assert_equal RTesseract.new(@image_tiff,{:chop_enable=>0}).config , "chop_enable 0"
|
58
|
+
assert_equal RTesseract.new(@image_tiff,{:chop_enable=>0,:enable_assoc=>0}).config , "chop_enable 0\nenable_assoc 0"
|
59
|
+
assert_equal RTesseract.new(@image_tiff,{:chop_enable=>0}).to_s_without_spaces , "43ZZ"
|
60
|
+
end
|
61
|
+
|
62
|
+
should "crop image" do
|
63
|
+
assert_equal RTesseract.new(@image_tiff).crop!(140,10,36,40).to_s_without_spaces, "4"
|
64
|
+
assert_equal RTesseract.new(@image_tiff).crop!(180,10,36,40).to_s_without_spaces, "3"
|
65
|
+
assert_equal RTesseract.new(@image_tiff).crop!(200,10,36,40).to_s_without_spaces, "Z"
|
66
|
+
assert_equal RTesseract.new(@image_tiff).crop!(220,10,30,40).to_s_without_spaces, "Z"
|
67
|
+
end
|
68
|
+
|
69
|
+
should "unique uid" do
|
70
|
+
assert_not_equal RTesseract.new(@image_tiff).generate_uid , RTesseract.new(@image_tiff).generate_uid
|
71
|
+
end
|
72
|
+
|
73
|
+
should "generate a unique id" do
|
74
|
+
reg = RTesseract.new(@image_tiff)
|
75
|
+
assert_equal reg.generate_uid , reg.generate_uid
|
76
|
+
value = reg.generate_uid
|
77
|
+
reg.convert
|
78
|
+
assert_not_equal value , reg.generate_uid
|
79
|
+
end
|
80
|
+
end
|
81
|
+
end
|
82
|
+
|
metadata
ADDED
@@ -0,0 +1,120 @@
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
2
|
+
name: mini_tesseract
|
3
|
+
version: !ruby/object:Gem::Version
|
4
|
+
prerelease:
|
5
|
+
version: 0.0.7
|
6
|
+
platform: ruby
|
7
|
+
authors:
|
8
|
+
- Danilo Jeremias da Silva
|
9
|
+
- Hooopo
|
10
|
+
autorequire:
|
11
|
+
bindir: bin
|
12
|
+
cert_chain: []
|
13
|
+
|
14
|
+
date: 2011-08-23 00:00:00 +08:00
|
15
|
+
default_executable:
|
16
|
+
dependencies:
|
17
|
+
- !ruby/object:Gem::Dependency
|
18
|
+
name: jeweler
|
19
|
+
prerelease: false
|
20
|
+
requirement: &id001 !ruby/object:Gem::Requirement
|
21
|
+
none: false
|
22
|
+
requirements:
|
23
|
+
- - ">="
|
24
|
+
- !ruby/object:Gem::Version
|
25
|
+
version: 1.4.0
|
26
|
+
type: :development
|
27
|
+
version_requirements: *id001
|
28
|
+
- !ruby/object:Gem::Dependency
|
29
|
+
name: thoughtbot-shoulda
|
30
|
+
prerelease: false
|
31
|
+
requirement: &id002 !ruby/object:Gem::Requirement
|
32
|
+
none: false
|
33
|
+
requirements:
|
34
|
+
- - ">="
|
35
|
+
- !ruby/object:Gem::Version
|
36
|
+
version: "0"
|
37
|
+
type: :development
|
38
|
+
version_requirements: *id002
|
39
|
+
- !ruby/object:Gem::Dependency
|
40
|
+
name: rmagick
|
41
|
+
prerelease: false
|
42
|
+
requirement: &id003 !ruby/object:Gem::Requirement
|
43
|
+
none: false
|
44
|
+
requirements:
|
45
|
+
- - ">="
|
46
|
+
- !ruby/object:Gem::Version
|
47
|
+
version: 2.10.1
|
48
|
+
type: :runtime
|
49
|
+
version_requirements: *id003
|
50
|
+
- !ruby/object:Gem::Dependency
|
51
|
+
name: mini_magick
|
52
|
+
prerelease: false
|
53
|
+
requirement: &id004 !ruby/object:Gem::Requirement
|
54
|
+
none: false
|
55
|
+
requirements:
|
56
|
+
- - ~>
|
57
|
+
- !ruby/object:Gem::Version
|
58
|
+
version: "3.0"
|
59
|
+
type: :runtime
|
60
|
+
version_requirements: *id004
|
61
|
+
description: Ruby library for working with the Tesseract OCR.
|
62
|
+
email: hoooopo@gmail.com
|
63
|
+
executables: []
|
64
|
+
|
65
|
+
extensions: []
|
66
|
+
|
67
|
+
extra_rdoc_files:
|
68
|
+
- LICENSE
|
69
|
+
- README.rdoc
|
70
|
+
files:
|
71
|
+
- .document
|
72
|
+
- LICENSE
|
73
|
+
- README.rdoc
|
74
|
+
- Rakefile
|
75
|
+
- lib/processors/mini_magick.rb
|
76
|
+
- lib/processors/rmagick.rb
|
77
|
+
- lib/rtesseract.rb
|
78
|
+
- lib/rtesseract/errors.rb
|
79
|
+
- lib/rtesseract/mixed.rb
|
80
|
+
- mini_tesseract.gemspec
|
81
|
+
- rtesseract.gemspec
|
82
|
+
- test/helper.rb
|
83
|
+
- test/images/mixed.tif
|
84
|
+
- test/images/test.bmp
|
85
|
+
- test/images/test.jpg
|
86
|
+
- test/images/test.png
|
87
|
+
- test/images/test.tif
|
88
|
+
- test/images/test1.tif
|
89
|
+
- test/test_mixed.rb
|
90
|
+
- test/test_rtesseract.rb
|
91
|
+
has_rdoc: true
|
92
|
+
homepage: http://github.com/hooopo/rtesseract
|
93
|
+
licenses: []
|
94
|
+
|
95
|
+
post_install_message:
|
96
|
+
rdoc_options: []
|
97
|
+
|
98
|
+
require_paths:
|
99
|
+
- lib
|
100
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
101
|
+
none: false
|
102
|
+
requirements:
|
103
|
+
- - ">="
|
104
|
+
- !ruby/object:Gem::Version
|
105
|
+
version: "0"
|
106
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
107
|
+
none: false
|
108
|
+
requirements:
|
109
|
+
- - ">="
|
110
|
+
- !ruby/object:Gem::Version
|
111
|
+
version: "0"
|
112
|
+
requirements: []
|
113
|
+
|
114
|
+
rubyforge_project:
|
115
|
+
rubygems_version: 1.3.9.2
|
116
|
+
signing_key:
|
117
|
+
specification_version: 3
|
118
|
+
summary: Ruby library for working with the Tesseract OCR.
|
119
|
+
test_files: []
|
120
|
+
|