pdfbeads 1.0.8 → 1.0.9

data/ChangeLog

@@ -42,6 +42,16 @@
 
 2012 March 5 (Alexey Kryukov) Version 1.0.8
 
-    Palette colors in PNG images were treated as signed chars and this could cause
-    indexed images to be incorrecty displayed in the resulting PDF.
+    * Palette colors in PNG images were treated as signed chars and this could cause
+      indexed images to be incorrecty displayed in the resulting PDF.
 
+2012 April 22 (Alexey Kryukov) Version 1.0.9
+
+    + Add an option allowing to delete image files produced as an intermediate stage
+      during the PDF creation process.
+
+    * Processing indexed images with a small number of colors was broken.
+
+    * Don't attempt to use 'ocrx_word' elements which contain no bounding box
+      data (this should fix the problem with the hOCR output produced by some
+      tesseract versions).

data/bin/pdfbeads

@@ -42,7 +42,8 @@ pdfargs  = Hash[
   :labels         => nil,
   :toc            => nil,
   :pagelayout     => 'TwoPageRight',
-  :meta           => nil
+  :meta           => nil,
+  :delfiles       => false
 ]
 pageargs = Hash[
   :threshold       => 1,
@@ -169,6 +170,10 @@ OptionParser.new() do |opts|
   opts.separator "\n"
   opts.separator "General options:\n"
 
+  opts.on("-d", "--delete",
+                "Delete intermediate image files used to create PDF") do |d|
+    pdfargs[:delfiles] = d
+  end
   opts.on("-o", "--output FILE",
                 "Print output to a file instead of STDERR") do |f|
     outpath = f

data/lib/imageinspector.rb

@@ -444,14 +444,14 @@ class ImageInspector::Image
             @x_dpi = (x_dpm/100 * 2.54).round
             @y_dpi = (y_dpm/100 * 2.54).round
           when 'tRNS'
-            trans = Hash.new[]
+            trans = Hash.new()
             case @cspace
               when :Indexed
                 # Indexed colour, RGB. Each byte in this chunk is an alpha for
                 # the palette index in the PLTE ("palette") chunk up until the
                 # last non-opaque entry. Set up an array, stretching over all
                 # palette entries which will be 0 (opaque) or 1 (transparent).
-                @trans = io.read( size ).unpack( 'C*' )
+                @trans = io.read( length ).unpack( 'C*' )
               when :DeviceGray
                 # Greyscale. Corresponding to entries in the PLTE chunk.
                 # Grey is two bytes, range 0 .. (2 ^ bit-depth) - 1

data/lib/pdfbeads/pdfbuilder.rb

@@ -326,6 +326,17 @@ class PDFBeads::PDFBuilder
       cat.addToDict('Outlines', ref(toc[0][:pdfobj].getID))
       cat.addToDict('PageMode', "/UseOutlines")
     end
+
+    if @pdfargs[:delfiles]
+      pagefiles.each do |p|
+        $stderr.puts( "Cleaning up temporary files for #{p.name}" )
+        safe_delete( p.fg_layer ) if p.fg_created
+        safe_delete( p.bg_layer ) if p.bg_created
+        p.stencils.each do |s|
+          safe_delete( s[:path] ) if s[:created]
+        end
+      end
+    end
   end
 
   # Output the created PDF file to the disk.
@@ -347,6 +358,15 @@ class PDFBeads::PDFBuilder
 
   private
 
+  def safe_delete( path )
+    begin
+      File.delete( path )
+      $stderr.puts( " Deleted #{path}" )
+    rescue Exception => e
+        $stderr.puts( "Could not delete #{path}: #{e.message}" )
+    end
+  end
+
   def parseMeta( path )
     ret = Hash.new()
     return ret if path.nil? or path.eql? ''
@@ -390,8 +410,9 @@ class PDFBeads::PDFBuilder
       end
 
       item_text = item[:title].to_binary
-      item_text.sub!( /\x28/,"\x5C\x28" )
-      item_text.sub!( /\x29/,"\x5C\x29" )
+      item_text.gsub!( /\x5C/,"\x5C\x5C" )
+      item_text.gsub!( /\x28/,"\x5C\x28" )
+      item_text.gsub!( /\x29/,"\x5C\x29" )
       item[:pdfobj] = XObj.new(Hash[
         'Title'  => "(\xFE\xFF#{item_text.to_text})",
         'Parent' => ref(item[:parent][:pdfobj].getID),
@@ -465,6 +486,7 @@ class PDFBeads::PDFBuilder
     if ocr_words.length > 0
       ocr_words.each do |word|
         bbox = elementCoordinates( word,xscale,yscale )
+        next if bbox == [0,0,0,0]
         txt = elementText( word,charset )
         units << [txt,bbox]
       end

data/lib/pdfbeads/pdfpage.rb

@@ -37,7 +37,7 @@ class PDFBeads::PageDataProvider < Array
   # Allows to collect data needed for building an individual page
   # of a PDF document and gives access to those data.
   class PageData
-    attr_reader :name, :basename, :s_type, :stencils, :hocr_path
+    attr_reader :name, :basename, :s_type, :stencils, :hocr_path, :fg_created, :bg_created
     attr_accessor :width, :height, :x_res, :y_res, :fg_layer, :bg_layer
 
     def initialize( path,basename,args,exts,pref )
@@ -49,6 +49,7 @@ class PDFBeads::PageDataProvider < Array
       @exts = exts
       @pref = pref
       @bg_layer = @fg_layer = nil
+      @bg_created = @fg_created = false
     end
 
     def fillStencilArray()
@@ -60,6 +61,7 @@ class PDFBeads::PageDataProvider < Array
       map = Hash[
         :path => @name,
         :rgb  => [0.0, 0.0, 0.0],
+        :created => false
       ]
 
       insp = ImageInspector.new( @name )
@@ -72,7 +74,7 @@ class PDFBeads::PageDataProvider < Array
         @x_res = @y_res = fres
       end
 
-      if insp.depth == 1
+      if insp.depth == 1 and insp.trans.nil?
         @stencils << map
         ret = 1
 
@@ -190,6 +192,7 @@ class PDFBeads::PageDataProvider < Array
           px = Pixel.from_color( color )
           unless color.eql? exc
             cpath = "#{@basename}.#{color}.tiff"
+            created = false
             if not File.exists? cpath or force
               bitonal = img.copy
               # Caution: replacing colors in the colormap currently only works
@@ -208,10 +211,12 @@ class PDFBeads::PageDataProvider < Array
                 self.compression = Group4Compression
               end
               bitonal.destroy!
+              created = true
             end
             cmap = Hash[
               :path => cpath,
-              :rgb  => [px.red.to_f/QuantumRange, px.green.to_f/QuantumRange, px.blue.to_f/QuantumRange]
+              :rgb  => [px.red.to_f/QuantumRange, px.green.to_f/QuantumRange, px.blue.to_f/QuantumRange],
+              :created => created
             ]
             @stencils << cmap
             ret += 1
@@ -231,6 +236,7 @@ class PDFBeads::PageDataProvider < Array
           self.compression = Group4Compression
         }
         bitonal.destroy!
+        map[:created] = true
       end
 
       bgf = @pageargs[:bg_format]
@@ -263,6 +269,7 @@ class PDFBeads::PageDataProvider < Array
         end
 
         writeImage( img,bgpath,bgf )
+        @bg_created = true
       end
 
       map[:path] = binpath
@@ -327,7 +334,10 @@ class PDFBeads::PageDataProvider < Array
       end
 
       bgpath = "#{@basename}.bg." << fmt.downcase
-      @bg_layer = bgpath if writeImage( bg,bgpath,fmt )
+      if writeImage( bg,bgpath,fmt )
+        @bg_layer = bgpath
+        @bg_created = true
+      end
 
       bg.destroy!
       no_fg.destroy!
@@ -352,7 +362,10 @@ class PDFBeads::PageDataProvider < Array
         fg.alpha( DeactivateAlphaChannel )
 
         fgpath = "#{@basename}.fg." << fmt.downcase
-        @fg_layer = fgpath if writeImage( fg,fgpath,fmt )
+        if writeImage( fg,fgpath,fmt )
+          @fg_layer = fgpath
+          @fg_created = true
+        end
 
         fg.destroy!
         no_bg.destroy!

metadata

@@ -1,13 +1,13 @@
 --- !ruby/object:Gem::Specification 
 name: pdfbeads
 version: !ruby/object:Gem::Version 
-  hash: 7
+  hash: 5
   prerelease: 
   segments: 
   - 1
   - 0
-  - 8
-  version: 1.0.8
+  - 9
+  version: 1.0.9
 platform: ruby
 authors: 
 - Alexey Kryukov
@@ -15,7 +15,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2012-03-05 00:00:00 +04:00
+date: 2012-04-21 00:00:00 +04:00
 default_executable: pdfbeads
 dependencies: []