rtesseract 0.0.2 → 0.0.3

data/Rakefile

@@ -5,7 +5,7 @@ begin
   require 'jeweler'
   Jeweler::Tasks.new do |gem|
     gem.name = "rtesseract"
-    gem.version = '0.0.2'
+    gem.version = '0.0.3'
     gem.summary = "Ruby library for working with the Tesseract OCR."
     gem.description = "Ruby library for working with the Tesseract OCR."
     gem.email = "dannnylo@gmail.com"

data/lib/rtesseract.rb

@@ -3,18 +3,21 @@ require "pathname"
 require "tempfile"
 
 class RTesseract
-  VERSION = '0.0.2'
+  VERSION = '0.0.3'
+  attr_accessor :options
+  attr_writer   :lang
 
   def initialize(src="", options={})
-    @source = Pathname.new src
-    @command = options[:command] || "tesseract"
+    @source  = Pathname.new src
+    @command = options.delete(:command) || "tesseract"
+    @lang    = options.delete(:lang) || options.delete("lang") || ""
     @options = options
     @value   = ""
   end
 
   def source= src
-    @source = Pathname.new src
     @value = ""
+    @source = Pathname.new src
   end
 
   def image_name
@@ -30,7 +33,7 @@ class RTesseract
   end
 
   #Remove files
-  def remove_file(files)
+  def remove_file(files=[])
     files.each do |file|
       begin
         File.unlink(file) if File.exist?(file)
@@ -43,11 +46,42 @@ class RTesseract
       raise "Error on remove file."
   end
 
+  # 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
+  def lang
+    language = "#{@lang}".strip
+    {"eng" => ["eng","en","en-us","english"], "deu" => ["deu"], "deu-f" => ["deu-f"] , "fra" => ["fra"], "ita" => ["ita","it"] , "nld" => ["nld"] , "por" => ["por","pt","pt-br","portuguese"] , "spa" => ["spa"] , "vie" => ["vie"]}.each do |value,names|
+      return " -l #{value} " if names.include? language.downcase
+    end
+    ""
+  rescue
+    ""
+  end
+
+  def config
+    @options ||= {}
+    @options.collect{|k,v| "#{k} #{v}" }.join("\n")
+  end
+
+  def config_file
+    #TODO: create the config
+  end
+
   #Convert image to string
   def convert
     tmp_file  = Pathname.new(Dir::tmpdir).join("#{@source.basename}")
     tmp_image = image_to_tiff
-    `#{@command} #{tmp_image} #{tmp_file.to_s}`
+    `#{@command} #{tmp_image} #{tmp_file.to_s} #{lang}`
     @value = File.read("#{tmp_file.to_s}.txt").to_s
     remove_file([tmp_image,"#{tmp_file.to_s}.txt"])
   rescue

data/rtesseract.gemspec

@@ -5,11 +5,11 @@
 
 Gem::Specification.new do |s|
   s.name = %q{rtesseract}
-  s.version = "0.0.2"
+  s.version = "0.0.3"
 
   s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
   s.authors = ["Danilo Jeremias da Silva"]
-  s.date = %q{2010-09-10}
+  s.date = %q{2010-10-11}
   s.description = %q{Ruby library for working with the Tesseract OCR.}
   s.email = %q{dannnylo@gmail.com}
   s.extra_rdoc_files = [
@@ -25,6 +25,11 @@ Gem::Specification.new do |s|
      "lib/rtesseract.rb",
      "rtesseract.gemspec",
      "test/helper.rb",
+     "test/images/test.bmp",
+     "test/images/test.jpg",
+     "test/images/test.png",
+     "test/images/test.tif",
+     "test/images/test1.tif",
      "test/test_rtesseract.rb"
   ]
   s.homepage = %q{http://github.com/dannnylo/rtesseract}

data/test/test_rtesseract.rb

@@ -1,9 +1,62 @@
 require 'helper'
+require 'pathname'
 
 class TestRtesseract < Test::Unit::TestCase
-  should "probably rename this file and start testing for real" do
-    flunk "hey buddy, you should probably rename this file and start testing for real"
-  end
+  context "Path" do
+    setup do
+      @path = Pathname.new(__FILE__.gsub("test_rtesseract.rb","")).expand_path
+      @image_tiff = @path.join("images","test.tif").to_s
+    end
+
+    should "be instantiable" do
+      assert_equal RTesseract.new.class , RTesseract
+      assert_equal RTesseract.new("").class , RTesseract
+      assert_equal RTesseract.new(@image_tiff).class , RTesseract
+    end
+
+    should "translate image to text" do
+      assert_equal RTesseract.new(@image_tiff).to_s_without_spaces , "43ZZ"
+      assert_equal RTesseract.new(@path.join("images","test1.tif").to_s).to_s_without_spaces , "V2V4"
+    end
+
+    should "translate images .png, .jpg, .bmp" do
+      assert_equal RTesseract.new(@path.join("images","test.png").to_s).to_s_without_spaces , "HW9W"
+      assert_equal RTesseract.new(@path.join("images","test.jpg").to_s).to_s_without_spaces , "3R8Z"
+      assert_equal RTesseract.new(@path.join("images","test.bmp").to_s).to_s_without_spaces , "ZLA6"
+    end
+
+    should "change the image" do
+      image = RTesseract.new(@image_tiff)
+      assert_equal image.to_s_without_spaces,"43ZZ"
+      image.source = @path.join("images","test1.tif").to_s
+      assert_equal image.to_s_without_spaces,"V2V4"
+    end
 
+    should "select the language" do
+      #English
+      assert_equal RTesseract.new(@image_tiff,{:lang=>"eng"}).lang , " -l eng "
+      assert_equal RTesseract.new(@image_tiff,{:lang=>"en"}).lang , " -l eng "
+      assert_equal RTesseract.new(@image_tiff,{:lang=>"en-US"}).lang , " -l eng "
+      assert_equal RTesseract.new(@image_tiff,{:lang=>"english"}).lang , " -l eng "
+
+      #Portuguese
+      assert_equal RTesseract.new(@image_tiff,{:lang=>"por"}).lang , " -l por "
+      assert_equal RTesseract.new(@image_tiff,{:lang=>"pt-BR"}).lang , " -l por "
+      assert_equal RTesseract.new(@image_tiff,{:lang=>"pt-br"}).lang , " -l por "
+      assert_equal RTesseract.new(@image_tiff,{:lang=>"pt"}).lang , " -l por "
+      assert_equal RTesseract.new(@image_tiff,{:lang=>"portuguese"}).lang , " -l por "
+
+      assert_equal RTesseract.new(@image_tiff,{:lang=>"eng"}).to_s_without_spaces , "43ZZ"
+      assert_equal RTesseract.new(@image_tiff,{:lang=>"por"}).to_s_without_spaces , "43ZZ"
+
+      assert_equal RTesseract.new(@image_tiff,{:lang=>"eng"}).lang , " -l eng "
+    end
+
+    should "be configurable" do
+      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"
+      assert_equal RTesseract.new(@image_tiff,{:chop_enable=>0}).config , "chop_enable 0"
+      assert_equal RTesseract.new(@image_tiff,{:enable_assoc=>0,:chop_enable=>0}).config , "chop_enable 0\nenable_assoc 0"
+    end
+  end
 end
 

metadata

@@ -1,13 +1,13 @@
 --- !ruby/object:Gem::Specification 
 name: rtesseract
 version: !ruby/object:Gem::Version 
-  hash: 27
+  hash: 25
   prerelease: false
   segments: 
   - 0
   - 0
-  - 2
-  version: 0.0.2
+  - 3
+  version: 0.0.3
 platform: ruby
 authors: 
 - Danilo Jeremias da Silva
@@ -15,7 +15,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2010-09-10 00:00:00 -03:00
+date: 2010-10-11 00:00:00 -03:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency 
@@ -66,6 +66,11 @@ files:
 - lib/rtesseract.rb
 - rtesseract.gemspec
 - test/helper.rb
+- test/images/test.bmp
+- test/images/test.jpg
+- test/images/test.png
+- test/images/test.tif
+- test/images/test1.tif
 - test/test_rtesseract.rb
 has_rdoc: true
 homepage: http://github.com/dannnylo/rtesseract