ruby_pager 0.1.1

data/lib/ruby_pager/text_line.rb

@@ -0,0 +1,72 @@
+
+module RubyPager
+
+  class Text_Line
+    attr_reader :id, :index  ,:text ,:contour, :baseline
+    def initialize(ex_index, ex_data)
+      @data=ex_data
+      @index=ex_index
+      @id = @data["@id"]
+      @text = @data["TextEquiv"]["Unicode"]
+      load_coords()
+      load_baseline()
+    end
+
+    def self.blank
+      return Text_Line.new(0,Text_Line.blank_data)
+    end
+
+    def id= ex_id
+      raise(ArgumentError, "Got passed a non string object") if ex_id.class != String
+      @id=ex_id
+    end
+
+    def text= ex_text
+      raise(ArgumentError, "Got passed a non string object") if ex_text.class != String
+      @text=ex_text
+    end
+
+    def index=(ex_index)
+      raise(ArgumentError, "Got passed a negative value to update the index") if ex_index.to_i < 0
+      @index=ex_index.to_i
+    end
+
+    def get_consolidated_data
+      consolidate_data()
+      return @data
+    end
+
+    def baseline_vertical_noise(ex_std_dev)
+      @baseline.vertical_noise(ex_std_dev)
+    end
+
+    def self.blank_data
+      res=Hash.new
+      res["@id"]=""
+      res["TextEquiv"]=Hash.new
+      res["TextEquiv"]["Unicode"]=""
+      res["Baseline"]=Hash.new
+      res["Baseline"]["@points"]=Coords.blank_data
+      res["Coords"]=Hash.new
+      res["Coords"]["@points"]=Coords.blank_data
+      return res
+    end
+    private
+
+    def load_coords()
+      @contour = Coords.new(@data["Coords"]["@points"]);
+    end
+
+    def load_baseline()
+      @baseline = Coords.new(@data["Baseline"]["@points"]);
+    end
+
+    def consolidate_data()
+      @data["@id"]=@id
+      @data["TextEquiv"]["Unicode"]=@text
+      @data["Baseline"]["@points"]=@baseline.get_consolidated_data
+      @data["Coords"]["@points"]=@contour.get_consolidated_data
+    end
+  end
+
+end

data/lib/ruby_pager/text_region.rb

@@ -0,0 +1,130 @@
+
+module RubyPager
+
+  class Text_Region
+    attr_reader :id, :index, :custom, :contour
+    def initialize(ex_index, ex_data)
+      @logger = Utils::ApplicationLogger.instance
+      @data=ex_data
+      @index=ex_index
+      @id = @data["@id"]
+      @custom=@data["@custom"]
+      @text_lines=Hash.new
+      load_text_lines()
+      load_contour()
+    end
+
+    def self.blank
+      return Text_Region.new(0,Text_Region.blank_data)
+    end
+
+  def size
+      return @text_lines.size
+    end
+
+    def id= (ex_id)
+      raise(ArgumentError, "Got passed a non string object") if ex_id.class != String
+      @id=ex_id
+    end
+
+    def index= (ex_index)
+      raise(ArgumentError, "Got passed a negative value to update the index") if ex_index.to_i < 0
+      @index=ex_index.to_i
+    end
+
+    def [](ex_key)
+      raise(RangeError, "Index #{ex_key} is out of range") unless @text_lines.has_key? ex_key
+      return @text_lines[ex_key]
+    end
+
+    def has_line? line_id
+      return @text_lines.has_key? line_id
+    end
+
+    def delete(ex_line_id)
+      if has_line? ex_line_id
+        @logger.info("Deleting text region #{ex_line_id}")
+        @text_lines.delete(ex_line_id)
+        review_lines_index
+      else
+        raise(ArgumentError, "Line id #{ex_line_id} does not exist so it can not be deleted")
+      end
+    end
+
+    def clear_text_lines()
+      @text_lines.clear
+    end
+
+    def push(ex_line)
+      raise(ArgumentError, "Got passed a non text line object") if ex_line.class != RubyPager::Text_Line
+      raise(ArgumentError, "Text line id already in use") if @text_lines.has_key? ex_line.id
+      ex_line.index=@text_lines.size
+      @text_lines[ex_line.id]=ex_line
+
+    end
+
+    def get_consolidated_data
+      consolidate_data()
+      return @data
+    end
+
+    def self.blank_data
+      res=Hash.new
+      res["@id"]=""
+      res["@custom"]=""
+      res["TextLine"]=Array.new
+      res["Coords"]=Hash.new
+      res["Coords"]["@points"]=Coords.blank_data
+      return res
+    end
+
+    def baseline_vertical_noise(ex_std_dev)
+      @text_lines.values.each {|text_line| text_line.baseline_vertical_noise(ex_std_dev) }
+    end
+
+    private
+
+    def load_text_lines()
+      if @data["TextLine"]
+        if @data["TextLine"].class == Array
+          line_array=@data["TextLine"]
+          line_array.each_with_index {|text_line,index |
+            @text_lines[text_line["@id"]]=Text_Line.new(index,text_line)
+          }
+        end
+
+        if @data["TextLine"].class == Hash
+          text_line=@data["TextLine"]
+          @text_lines[text_line["@id"]]=Text_Line.new(0,text_line)
+        end
+      end
+    end
+
+    def load_contour()
+      @contour = Coords.new(@data["Coords"]["@points"]);
+    end
+
+    def consolidate_data
+      @data["@custom"]=@custom
+      @data["@id"]=@id
+      @data["Coords"]["@points"]=@contour.get_consolidated_data
+      @data["TextLine"].clear if @data["TextLine"] and @data["TextLine"].class == Array
+      if@text_lines.length>1
+        @text_lines.values.each {|text_line|
+         @data["TextLine"].push(text_line.get_consolidated_data)
+        }
+      end
+      @data["TextLine"]=@text_lines.values[0].get_consolidated_data if@text_lines.length==1
+    end
+
+    def review_lines_index
+      index =0
+      @text_lines.values.each {|line|
+        line.index=index
+        index+=1
+      }
+    end
+
+  end
+
+end

data/lib/ruby_pager/version.rb

@@ -0,0 +1,3 @@
+module RubyPager
+  VERSION = "0.1.1"
+end

data/lib/ruby_pager/xml.rb

@@ -0,0 +1,90 @@
+require 'nokogiri'
+require 'nori'
+require 'ap'
+module RubyPager
+
+
+  class XML
+    def self.exists?(file_name)
+      if file_name.end_with?(".xml")
+        return File.file?(file_name)
+      end
+      return false
+    end
+
+    def self.load(file_name)
+      if self.exists?(file_name)
+        xml  = Nokogiri::XML(File.open(file_name)) { |config| config.strict.noblanks }
+        hash = Nori.new(:parser => :nokogiri, :advanced_typecasting => false).parse(xml.to_s)
+        return hash
+      end
+      return nil
+    end
+
+    def self.save(file_name,hash)
+      File.open(file_name, 'w') { |file| file.write(self.generate_xml(hash)) } ;
+    end
+
+    def self.generate_xml(data, parent = false, opt = {})
+      return if data.to_s.empty?
+      return unless data.is_a?(Hash)
+
+      unless parent
+        # assume that if the hash has a single key that it should be the root
+        root, data = (data.length == 1) ? data.shift : ["root", data]
+        attr = Hash.new
+        data.each{|label,value|
+          if(label.start_with?("@")) then
+            new_label = label.dup
+            new_label[0]= ''
+            attr[new_label]=value
+          end
+        }
+        builder = Nokogiri::XML::Builder.new(opt) do |xml|
+          xml.send(root,attr) {
+            data.each{|label,value|
+              if(label.start_with?("@")) then
+                data.delete(label)
+              end
+            }
+            generate_xml(data, xml)
+          }
+        end
+
+        return builder.to_xml
+      end
+      data.each { |label, value|
+        if value.is_a?(Hash)
+          attr = Hash.new
+          value.each{|vlabel,vvalue|
+            if(vlabel.start_with?("@")) then
+              new_label = vlabel.dup
+              new_label[0]= ''
+              attr[new_label]=vvalue
+            end
+          }
+
+          parent.send(label,attr) {
+            value.each{|vlabel,vvalue|
+              if(vlabel.start_with?("@")) then
+                value.delete(vlabel)
+              end
+            }
+            generate_xml(value, parent)
+          }
+
+        elsif value.is_a?(Array)
+          value.each { |el|
+            # lets trick the above into firing so we do not need to rewrite the checks
+            el = {label => el}
+            generate_xml(el, parent)
+          }
+
+        else
+          parent.send(label, value)
+        end
+      }
+    end
+  end
+
+end

data/lib/ruby_pager.rb

@@ -0,0 +1,18 @@
+require "ruby_pager/version"
+require "ruby_pager/xml"
+require "ruby_pager/page"
+require "ruby_pager/metadata"
+require "ruby_pager/reading_order"
+require "ruby_pager/text_region"
+require "ruby_pager/text_line"
+require "ruby_pager/coords"
+require "ruby_pager/coord"
+require "ruby_pager/image_data"
+require "ruby_pager/image"
+require "ruby_pager/extendmatrix2"
+require "ruby_pager/histogram"
+require "ruby_pager/gaussian_noise"
+require "ruby_pager/application_logger"
+module RubyPager
+  # Your code goes here...
+end

data/no_lines.xml

@@ -0,0 +1,14 @@
+<?xml version="1.0"?>
+<PcGts xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns="http://schema.primaresearch.org/PAGE/gts/pagecontent/2013-07-15" xsi:schemaLocation="http://schema.primaresearch.org/PAGE/gts/pagecontent/2013-07-15 http://schema.primaresearch.org/PAGE/gts/pagecontent/2013-07-15/pagecontent.xsd" pcGtsId="">
+  <Metadata>
+    <Creator>Ruby Page</Creator>
+    <Created>2018-02-08T16:46:58</Created>
+    <LastChange>2018-02-08T16:46:58</LastChange>
+  </Metadata>
+  <Page imageFilename="test.jpg" imageWidth="4715" imageHeight="3483">
+    <TextRegion id="empty_region" custom="">
+      <Coords points="0,0 0,4715 3483,4715 3483,0"/>
+    </TextRegion>
+    <ReadingOrder/>
+  </Page>
+</PcGts>

data/no_regions.xml

@@ -0,0 +1,11 @@
+<?xml version="1.0"?>
+<PcGts xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns="http://schema.primaresearch.org/PAGE/gts/pagecontent/2013-07-15" xsi:schemaLocation="http://schema.primaresearch.org/PAGE/gts/pagecontent/2013-07-15 http://schema.primaresearch.org/PAGE/gts/pagecontent/2013-07-15/pagecontent.xsd" pcGtsId="">
+  <Metadata>
+    <Creator>Ruby Page</Creator>
+    <Created>2018-02-08T16:46:58</Created>
+    <LastChange>2018-02-08T16:46:58</LastChange>
+  </Metadata>
+  <Page imageFilename="test.jpg" imageWidth="4715" imageHeight="3483">
+    <ReadingOrder/>
+  </Page>
+</PcGts>

data/one_line.xml

@@ -0,0 +1,21 @@
+<?xml version="1.0"?>
+<PcGts xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns="http://schema.primaresearch.org/PAGE/gts/pagecontent/2013-07-15" xsi:schemaLocation="http://schema.primaresearch.org/PAGE/gts/pagecontent/2013-07-15 http://schema.primaresearch.org/PAGE/gts/pagecontent/2013-07-15/pagecontent.xsd" pcGtsId="">
+  <Metadata>
+    <Creator>Ruby Page</Creator>
+    <Created>2018-02-08T16:46:58</Created>
+    <LastChange>2018-02-08T16:46:58</LastChange>
+  </Metadata>
+  <Page imageFilename="test.jpg" imageWidth="4715" imageHeight="3483">
+    <TextRegion id="line_region" custom="">
+      <Coords points="0,0 0,4715 3483,4715 3483,0"/>
+			<TextLine id="line_0">
+				<Coords points="16,0 305,0 305,112 16,112" />
+				<Baseline points="" />
+				<TextEquiv>
+					<Unicode>QUE 0.0260531 SE 0.03597 CE 0.042705 A 0.00188892 LES 0.214666 TE 0.0261389 LE 0.0465788 DE 0.201388 PER 0.0331285 DES 0.0896861 ME 0.0215401 EN 0.0273062 ES 0.0502558 ET 0.0750531 PRE 0.0306363 NOB 2.82646e-05 NOI 9.50498e-05 RE 0.0464671 NE 0.026413 NOM 0.000249426 NON 0.000673057 NOS 0.00177215 NOT 5.88758e-05 NOU 0.00122318 NOV 1.05446e-05 NOZ 1.33179e-05</Unicode>
+				</TextEquiv>
+			</TextLine>
+    </TextRegion>
+    <ReadingOrder/>
+  </Page>
+</PcGts>

data/ruby_pager.gemspec

@@ -0,0 +1,45 @@
+lib = File.expand_path("../lib", __FILE__)
+$LOAD_PATH.unshift(lib) unless $LOAD_PATH.include?(lib)
+	require "ruby_pager/version"
+
+	Gem::Specification.new do |spec|
+	  spec.name          = "ruby_pager"
+	    spec.version       = RubyPager::VERSION
+	      spec.authors       = ["Vicente Bosch"]
+	        spec.email         = ["vbosch@gmail.com"]
+
+	          spec.summary       = "command line tools to read and modify PAGE xmls"
+	            spec.description   = "command line tools to read and modify PAGE xmls"
+	              spec.homepage      = "https://github.com/vbosch/ruby_pager"
+	                spec.license       = "MIT"
+
+	                  # Prevent pushing this gem to RubyGems.org. To allow pushes either set the 'allowed_push_host'
+	                  # to allow pushing to a single host or delete this section to allow pushing to any host.
+	                  if spec.respond_to?(:metadata)
+	    spec.metadata["allowed_push_host"] = "https://rubygems.org"
+	      else
+	          raise "RubyGems 2.0 or newer is required to protect against " \
+	          	        "public gem pushes."
+	          	          end
+
+	          	            spec.files         = `git ls-files -z`.split("\x0").reject do |f|
+	          	                f.match(%r{^(test|spec|features)/})
+	  end
+	    spec.bindir        = "exe"
+	      spec.executables   = spec.files.grep(%r{^exe/}) { |f| File.basename(f) }
+	        spec.require_paths = ["lib"]
+	          spec.add_development_dependency "bundler"
+	            spec.add_development_dependency "rake"
+	              spec.add_development_dependency "rspec"
+	                spec.add_development_dependency "aruba"
+	                  spec.add_development_dependency "simplecov"
+	                    spec.add_dependency "pry"
+	                      spec.add_dependency "rgeo"
+	                        spec.add_dependency "gosu"
+	                          spec.add_dependency "rmagick"
+	                            spec.add_dependency "trollop"
+	                              spec.add_dependency "nori"
+	                                spec.add_dependency "awesome_print"
+	                                  spec.add_dependency "nokogiri"
+	                                    spec.add_dependency "extendmatrix"
+	                                    end