rtesseract 2.1.0 → 3.0.0

data/lib/rtesseract/command.rb

@@ -0,0 +1,41 @@
+require 'tmpdir'
+
+class RTesseract
+  class Command
+    FIXED = [:command, :psm, :oem, :lang, :tessdata_dir, :user_words, :user_patterns, :config_file]
+
+    attr_reader :options
+
+    def initialize(source, output, options)
+      @source = source
+      @output = output
+      @options = options
+    end
+
+    def configs
+      @options.to_h.map { |key, value| ['-c', "#{key}=#{value}"] unless FIXED.include?(key) }.compact
+    end
+
+    def full_command
+      command = [options.command, @source, @output]
+
+      command << ['--psm', options.psm.to_s] if options.psm
+      command << ['--oem', options.oem.to_s] if options.oem
+      command << ['-l', options.lang] if options.lang
+
+      command << ['--tessdata_dir', options.tessdata_dir] if options.tessdata_dir
+      command << ['--user_words', options.user_words] if options.user_words
+      command << ['--user_patterns', options.user_patterns] if options.user_patterns
+
+      command << configs
+
+      command << options.config_file.to_s if options.config_file
+
+      command.flatten
+    end
+
+    def run
+      Open3.capture2e(*full_command)
+    end
+  end
+end

data/lib/rtesseract/configuration.rb

@@ -1,75 +1,26 @@
-# RTesseract
-class RTesseract
-  # Aliases to languages names
-  LANGUAGES = {
-    'en' => 'eng',
-    'en-us' => 'eng',
-    'english' => 'eng',
-    'pt' => 'por',
-    'pt-br' => 'por',
-    'portuguese' => 'por',
-    'it' => 'ita',
-    'sp' => 'spa'
-  }.freeze
-
-  # Configuration class
-  class Configuration
-    attr_accessor :processor, :lang, :psm, :tessdata_dir, :user_words, :user_patterns, :command, :debug, :options_cmd
-
-    def initialize
-      @processor = 'rmagick'
-    end
-
-    # Global configuration
-    def parent
-      @parent ||= RTesseract.configuration || RTesseract::Configuration.new
-    end
-
-    # Set value of option
-    def option(options, name, default = nil)
-      self.instance_variable_set("@#{name}", options.option(name, parent.send(name)) || default)
-    end
+require 'ostruct'
 
-    # Return the values of options
-    def load_options(options, names = [])
-      names.each { |name| option(options, name, nil) }
+class RTesseract
+  class Configuration < OpenStruct
+    def merge(options)
+      RTesseract::Configuration.new(self.to_h.merge(options))
     end
   end
 
   class << self
-    attr_accessor :configuration
-  end
-
-  def self.configure
-    self.configuration ||= Configuration.new
-    yield(configuration)
-    self.clear_pdf_option
-  end
-
-  # Clear pdf option
-  def self.clear_pdf_option
-    if self.configuration.options_cmd
-      self.configuration.options_cmd.delete('pdf')
-      self.configuration.options_cmd.delete(:pdf)
+    def config
+      @config ||= RTesseract::Configuration.new(
+                    command: 'tesseract',
+                    debug_file: '/dev/null'
+                  )
     end
-  end
 
-  # Default command
-  def self.default_command
-    TesseractBin::Executables[:tesseract] || 'tesseract'
-  rescue
-    'tesseract'
-  end
+    def configure
+      yield(config) if block_given?
+    end
 
-  # Local config to instance
-  def self.local_config(options = {})
-    RTesseract::Configuration.new.tap do |config|
-      config.command = config.option(options, :command, RTesseract.default_command)
-      config.processor = config.option(options, :processor, 'rmagick')
-      config.load_options(options, [:lang, :psm, :tessdata_dir, :user_words, :user_patterns])
-      config.debug = config.option(options, :debug, false)
-      pdf_opts = lambda { |o| o == 'pdf' || o == :pdf }
-      config.options_cmd = [options.option(:options, nil)].delete_if(&pdf_opts).flatten.compact
+    def reset_config!
+      @config = nil
     end
   end
 end

data/lib/rtesseract/pdf.rb

@@ -0,0 +1,18 @@
+require 'tmpdir'
+
+class RTesseract
+  module Pdf
+    def self.temp_dir
+      @file_path = Pathname.new(Dir.tmpdir)
+    end
+
+    def self.run(source, options)
+      name = "rtesseract_#{SecureRandom.uuid}"
+      options.tessedit_create_pdf = 1
+
+      RTesseract::Command.new(source, temp_dir.join(name).to_s, options).run
+
+      File.open(temp_dir.join("#{name}.pdf").to_s, 'r')
+    end
+  end
+end

data/lib/rtesseract/text.rb

@@ -0,0 +1,9 @@
+require 'open3'
+
+class RTesseract
+  module Text
+    def self.run(source, options)
+      RTesseract::Command.new(source, 'stdout', options).run.first
+    end
+  end
+end

data/lib/rtesseract/tsv.rb

@@ -0,0 +1,18 @@
+require 'tmpdir'
+
+class RTesseract
+  module Tsv
+    def self.temp_dir
+      @file_path = Pathname.new(Dir.tmpdir)
+    end
+
+    def self.run(source, options)
+      name = "rtesseract_#{SecureRandom.uuid}"
+      options.tessedit_create_tsv = 1
+
+      RTesseract::Command.new(source, temp_dir.join(name).to_s, options).run
+
+      File.open(temp_dir.join("#{name}.tsv").to_s, 'r')
+    end
+  end
+end

data/lib/rtesseract/version.rb

@@ -0,0 +1,3 @@
+class RTesseract
+  VERSION = '3.0.0'.freeze
+end

data/lib/rtesseract.rb

@@ -1,238 +1,46 @@
-# encoding: UTF-8
-require 'pathname'
-require 'tempfile'
+require "rtesseract/check"
+require "rtesseract/configuration"
+require "rtesseract/command"
+require "rtesseract/text"
+require "rtesseract/pdf"
+require "rtesseract/box"
+require "rtesseract/tsv"
 
-require 'rtesseract/utils'
-require 'rtesseract/configuration'
-require 'rtesseract/errors'
-
-# Ruby wrapper for Tesseract OCR
 class RTesseract
-  attr_accessor :configuration
-  attr_reader :processor
-  attr_reader :source
-
-  def initialize(src = '', options = {})
-    self.configuration = RTesseract.local_config(options)
-    @options = options || {}
-    @points = {}
-    @processor = RTesseract::Processor.choose_processor!(configuration.processor)
-    self.source = src
-    initialize_hook
-  end
-
-  # Hook to end of initialize method
-  def initialize_hook
-  end
-
-  # Define the source
-  def source=(src)
-    @value = nil
-    @pdf_path = nil
-    @source = @processor.image?(src) ? src : Pathname.new(src)
-  end
-
-  # Crop image to convert
-  def crop!(points = {})
-    @value = nil
-    @points = points
-    self
-  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 = (configuration.lang || 'eng').to_s.strip.downcase
-    " -l #{LANGUAGES[language] || language} "
-  rescue
-    ''
-  end
-
-  # Convert option to command
-  def option_to_string(prefix, value = nil)
-    (value.nil? ? '' : " #{prefix} #{value} ")
-  rescue
-    ''
-  end
-
-  # Page Segment Mode
-  def psm
-    option_to_string('-psm', configuration.psm)
-  end
-
-  # Tessdata Dir
-  def tessdata_dir
-    option_to_string('--tessdata-dir', configuration.tessdata_dir)
-  end
-
-  # User Words
-  def user_words
-    option_to_string('--user-words', configuration.user_words)
-  end
-
-  # User Patterns
-  def user_patterns
-    option_to_string('--user-patterns', configuration.user_patterns)
-  end
-
-  # Options on line
-  def options_cmd
-    configuration.options_cmd
-  end
-
-  # Hook to before config
-  def config_hook
-  end
-
-  # Convert configurations
-  def config
-    @options ||= {}
-    config_hook
-    @options.map { |k, v| "#{k} #{v}" }.join("\n")
-  end
-
-  # Write config to file
-  def config_file
-    config_hook
-    return '' if @options == {}
-    conf = Tempfile.new('config')
-    conf.write(config)
-    conf.flush
-    conf.path
-  end
-
-  # TODO: Clear console for MacOS or Windows
-  def clear_console_output
-    return '' if configuration.debug
-    return '2>/dev/null' if File.exist?('/dev/null') # Linux console clear
-  end
-
-  # Get image
-  def image
-    (@image = @processor.image_to_tif(@source, @points)).path
-  end
-
-  # Extension of file
-  def file_ext
-    '.txt'
-  end
+  class Error < StandardError; end
 
-  # Detect version number
-  def tesseract_version
-    RTesseract::Utils.version_number
-  end
+  check_version!
 
+  attr_reader :config, :source
 
-  # Rand file path
-  def file_dest
-    @file_dest = Pathname.new(Dir.tmpdir).join("#{Time.now.to_f}#{rand(1500)}").to_s
-  end
-
-  # Full path of file with txt extension
-  def file_with_ext(ext = nil)
-    [@file_dest, ext || file_ext].join('')
-  end
-
-  # Run command
-  def convert_command
-    `#{configuration.command} "#{image}" "#{file_dest}" #{lang} #{psm} #{tessdata_dir} #{user_words} #{user_patterns} #{config_file} #{clear_console_output} #{options_cmd.join(' ')}`
-  end
-
-  # Is pdf output?
-  def pdf?
-    options_cmd.include? 'pdf'
-  end
-
-  # Read result file
-  def convert_text
-    @value = File.read(file_with_ext).to_s
+  def initialize(src = '', options = {})
+    @source = src
+    @config = RTesseract.config.merge(options)
   end
 
-  # Store pdf result path
-  def convert_pdf
-    @pdf_path = file_with_ext('.pdf')
+  def to_box
+    Box.run(@source, config)
   end
 
-  # Convert result to proper type
-  def convert_result
-    if pdf?
-      convert_pdf
-    else
-      convert_text
-      RTesseract::Utils.remove_files([@image, file_with_ext])
-    end
+  def words
+    to_box.map { |word| word[:word] }
   end
 
-  # Hook to convert
-  def after_convert_hook
+  def to_pdf
+    Pdf.run(@source, config)
   end
 
-  # Convert image to string
-  def convert
-    convert_command
-    after_convert_hook
-    convert_result
-  rescue => error
-    raise RTesseract::ConversionError.new(error), error, caller
+  def to_tsv
+    Tsv.run(@source, config)
   end
 
   # Output value
   def to_s
-    return @value if @value
-
-    if @processor.image?(@source) || @source.file?
-      convert
-      @value
-    else
-      fail RTesseract::ImageNotSelectedError.new(@source)
-    end
+    Text.run(@source, config)
   end
 
   # Remove spaces and break-lines
   def to_s_without_spaces
     to_s.gsub(/\s/, '')
   end
-
-  # Output pdf path
-  def to_pdf
-    return @pdf_path if @pdf_path
-
-    fail TesseractVersionError.new if tesseract_version.nil? || tesseract_version < 3.03
-
-    if @processor.image?(@source) || @source.file?
-      options_cmd << 'pdf'
-      convert
-      options_cmd.delete('pdf')
-      @pdf_path
-    else
-      fail RTesseract::ImageNotSelectedError.new(@source)
-    end
-  end
-
-  # Destroy pdf file
-  def clean
-    RTesseract::Utils.remove_files([@pdf_path])
-  end
-
 end
-
-require 'rtesseract/mixed'
-require 'rtesseract/box'
-require 'rtesseract/box_char'
-require 'rtesseract/blob'
-require 'rtesseract/processor'
-
-# Processors
-require 'processors/rmagick.rb'
-require 'processors/mini_magick.rb'
-require 'processors/none.rb'

data/rtesseract.gemspec

@@ -1,99 +1,33 @@
-# Generated by jeweler
-# DO NOT EDIT THIS FILE DIRECTLY
-# Instead, edit Jeweler::Tasks in Rakefile, and run 'rake gemspec'
-# -*- encoding: utf-8 -*-
-# stub: rtesseract 2.1.0 ruby lib
 
-Gem::Specification.new do |s|
-  s.name = "rtesseract"
-  s.version = "2.1.0"
+lib = File.expand_path("../lib", __FILE__)
+$LOAD_PATH.unshift(lib) unless $LOAD_PATH.include?(lib)
+require "rtesseract/version"
 
-  s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
-  s.require_paths = ["lib"]
-  s.authors = ["Danilo Jeremias da Silva"]
-  s.date = "2016-09-08"
-  s.description = "Ruby library for working with the Tesseract OCR."
-  s.email = "dannnylo@gmail.com"
-  s.extra_rdoc_files = [
-    "LICENSE.txt",
-    "README.rdoc"
-  ]
-  s.files = [
-    ".document",
-    ".rspec",
-    ".travis.yml",
-    "CHANGELOG.md",
-    "Gemfile",
-    "Gemfile.lock",
-    "LICENSE.txt",
-    "README.rdoc",
-    "Rakefile",
-    "VERSION",
-    "lib/processors/mini_magick.rb",
-    "lib/processors/none.rb",
-    "lib/processors/rmagick.rb",
-    "lib/rtesseract.rb",
-    "lib/rtesseract/blob.rb",
-    "lib/rtesseract/box.rb",
-    "lib/rtesseract/box_char.rb",
-    "lib/rtesseract/configuration.rb",
-    "lib/rtesseract/errors.rb",
-    "lib/rtesseract/mixed.rb",
-    "lib/rtesseract/processor.rb",
-    "lib/rtesseract/utils.rb",
-    "rtesseract.gemspec",
-    "spec/configs/eng.user-words.txt",
-    "spec/images/README.pdf",
-    "spec/images/blank.tif",
-    "spec/images/mixed.tif",
-    "spec/images/orientation_reverse.png",
-    "spec/images/test with spaces.tif",
-    "spec/images/test-pdf.png",
-    "spec/images/test.bmp",
-    "spec/images/test.jpg",
-    "spec/images/test.png",
-    "spec/images/test.tif",
-    "spec/images/test1.tif",
-    "spec/images/test_words.png",
-    "spec/rtesseract_box_char_spec.rb",
-    "spec/rtesseract_box_spec.rb",
-    "spec/rtesseract_mixed_spec.rb",
-    "spec/rtesseract_spec.rb",
-    "spec/spec_helper.rb"
-  ]
-  s.homepage = "http://github.com/dannnylo/rtesseract"
-  s.licenses = ["MIT"]
-  s.rubygems_version = "2.4.8"
-  s.summary = "Ruby library for working with the Tesseract OCR."
+Gem::Specification.new do |spec|
+  spec.name          = "rtesseract"
+  spec.version       = RTesseract::VERSION
+  spec.authors       = ["Danilo Jeremias da Silva"]
+  spec.email         = ["dannnylo@gmail.com"]
 
-  if s.respond_to? :specification_version then
-    s.specification_version = 4
+  spec.summary       = "Ruby library for working with the Tesseract OCR.".freeze
+  spec.description   = "Ruby library for working with the Tesseract OCR.".freeze
+  spec.homepage      = "http://github.com/dannnylo/rtesseract".freeze
+  spec.license       = "MIT"
 
-    if Gem::Version.new(Gem::VERSION) >= Gem::Version.new('1.2.0') then
-      s.add_runtime_dependency(%q<nokogiri>, [">= 0"])
-      s.add_development_dependency(%q<rspec>, [">= 0"])
-      s.add_development_dependency(%q<rdoc>, [">= 0"])
-      s.add_development_dependency(%q<bundler>, [">= 0"])
-      s.add_development_dependency(%q<jeweler>, ["~> 2.1.1"])
-      s.add_development_dependency(%q<simplecov>, [">= 0"])
-      s.add_development_dependency(%q<coveralls>, [">= 0"])
-    else
-      s.add_dependency(%q<nokogiri>, [">= 0"])
-      s.add_dependency(%q<rspec>, [">= 0"])
-      s.add_dependency(%q<rdoc>, [">= 0"])
-      s.add_dependency(%q<bundler>, [">= 0"])
-      s.add_dependency(%q<jeweler>, ["~> 2.1.1"])
-      s.add_dependency(%q<simplecov>, [">= 0"])
-      s.add_dependency(%q<coveralls>, [">= 0"])
-    end
-  else
-    s.add_dependency(%q<nokogiri>, [">= 0"])
-    s.add_dependency(%q<rspec>, [">= 0"])
-    s.add_dependency(%q<rdoc>, [">= 0"])
-    s.add_dependency(%q<bundler>, [">= 0"])
-    s.add_dependency(%q<jeweler>, ["~> 2.1.1"])
-    s.add_dependency(%q<simplecov>, [">= 0"])
-    s.add_dependency(%q<coveralls>, [">= 0"])
+  # Specify which files should be added to the gem when it is released.
+  # The `git ls-files -z` loads the files in the RubyGem that have been added into git.
+  spec.files         = Dir.chdir(File.expand_path('..', __FILE__)) do
+    `git ls-files -z`.split("\x0").reject { |f| f.match(%r{^(test|spec|features)/}) }
   end
-end
+  spec.bindir        = "exe"
+  spec.executables   = spec.files.grep(%r{^exe/}) { |f| File.basename(f) }
+  spec.require_paths = ["lib"]
+
+  spec.add_development_dependency "bundler", "~> 1.17"
+  spec.add_development_dependency "rake", "~> 10.0"
+  spec.add_development_dependency "rspec", "~> 3.0"
+  spec.add_development_dependency "simplecov"
+  spec.add_development_dependency "coveralls"
 
+  spec.add_dependency "nokogiri"
+end