pdfbeads 1.0.1 → 1.0.2

data/bin/pdfbeads

@@ -108,13 +108,13 @@ OptionParser.new() do |opts|
                 "Generate one shared JBIG2 dictionary per NUM pages.",
                 "This option is only applied when JBIG2 compression",
                 "is used. Default value is #{pageargs[:pages_per_dict]}") do |p|
-    pageargs[:pages_per_dict] = p
+    pageargs[:pages_per_dict] = p.to_i
   end
   opts.on("-r", "--force-resolution DPI",
                 "Set resolution for foreground mask images to the",
                 "specified value (in pixels per inch). Note that the",
                 "image is not actually resampled.") do |r|
-    pageargs[:st_resolution] = r
+    pageargs[:st_resolution] = r.to_f
   end
   opts.on("-x", "--max-colors NUM",
                 "If pdfbeads finds an indexed file with NUM or",
@@ -123,7 +123,7 @@ OptionParser.new() do |opts|
                 "PDF page mask. Otherwise the file is treated just",
                 "like a normal greyscale or color image. Default",
                 "value is #{pageargs[:maxcolors]}") do |num|
-    pageargs[:maxcolors] = num
+    pageargs[:maxcolors] = num.to_i
   end
 
   opts.separator "\n"

data/lib/imageinspector.rb

@@ -118,6 +118,22 @@ class ImageInspector::Image
     end
   end
 
+  def nextImage()
+    if @format.eql? :TIFF and @next_off > 0
+      begin
+        if @input.kind_of? IO or @input.kind_of? StringIO
+          tiffNext( @input )
+        else
+          File.open( @input, 'rb' ) { |io| tiffNext( io ) }
+        end
+        return true
+      rescue Exception => e
+        $stderr.puts( "Could not read data from #{@fname}: " << e.message )
+      end
+    end
+    false
+  end
+
   private
 
   def clearData
@@ -221,6 +237,11 @@ class ImageInspector::Image
     return ret
   end
 
+  def tiffNext( io )
+    sign = io.read( 4 )
+    tiffExamine( io,sign,@next_off )
+  end
+
   def tiffParseIFD( io,offset,intgr )
     packspec = [
       nil,              # nothing (shouldn't happen)
@@ -229,7 +250,7 @@ class ImageInspector::Image
       intgr,            # SHORT (16-bit unsigned integer)
       intgr.upcase,     # LONG (32-bit unsigned integer)
       intgr.upcase * 2, # RATIONAL (numerator + denominator)
-      intgr.upcase,     # SBYTE (8-bit signed integer)
+      'c',              # SBYTE (8-bit signed integer)
       'A*',             # undefined, but used for EXIF version
       intgr,            # SSHORT (16-bit signed integer)
       intgr.upcase,     # SLONG (32-bit signed integer)
@@ -271,7 +292,7 @@ class ImageInspector::Image
       @tags.has_key? 0x0100 and @tags.has_key? 0x0101 and 
       @tags.has_key? 0x0106 and @tags.has_key? 0x0111 and @tags.has_key? 0x0117 )
 
-    unless @tags.has_key? 0x8769
+    unless @format.eql? :JPEG
       @width = @tags[0x0100][0]; @height = @tags[0x0101][0]
 
       @tags[0x0111].each_index do |i|
@@ -334,8 +355,8 @@ class ImageInspector::Image
       buf.push( b )
       tag = buf[4..7].pack('c*')
 
-      # Currently we can detect only width/height for JP2 images.
-      # That's OK, as other parameters aren't needed for PDF generation.
+      # Currently no support for resolution, as I have never seen JP2 images
+      # with 'res '/'resc'/'resd' boxes, and not sure if they are ever used.
       if ['ftyp','jp2h','ihdr','colr','res ','resc',
           'resd','prfl','bpcc','pclr','cdef','jp2i'].include? tag
         length = buf[0..4].pack( 'c*' ).unpack( 'N' )[0]

data/lib/pdfbeads.rb

@@ -40,8 +40,8 @@ begin
   require 'hpricot'
   $has_hpricot = true
 rescue LoadError
-  $stderr.puts( "Warning: the hpricot extension is not available." )
-  $stderr.puts( "  pdfbeads will not be able to read OCR data from hOCR files." )
+  $stderr.puts( "Warning: the hpricot extension is not available. I'll not be able" )
+  $stderr.puts( "\tto create hidden text layer from hOCR files." )
   $has_hpricot = false
 end
 

data/lib/pdfbeads/pdfpage.rb

@@ -70,27 +70,34 @@ class PDFBeads::PageDataProvider < Array
       else
         @x_res = @y_res = fres
       end
+
       if insp.depth == 1
         @stencils << map
         $stderr.puts( "Prepared data for processing #{@name}\n" )
-        return 1
-      end
+        ret = 1
+
+      else
+        img = ImageList.new( @name )
+        # ImageMagick incorrectly identifies indexed PNG images as DirectClass.
+        # It also assigns a strange color value to fully opaque areas. So
+        # we have to use an independent approach to recognize indexed images.
+        unless insp.palette.nil?
+          img.class_type = PseudoClass
+          ret = processIndexed( img,@pageargs[:maxcolors],force )
+        end
+        ret = processMixed( img,force,map ) if ret == 0
+        img.destroy!
 
-      img = ImageList.new( @name )
-      # ImageMagick incorrectly identifies indexed PNG images as DirectClass.
-      # It also assigns a strange color value to fully opaque areas. So
-      # we have to use an independent approach to recognize indexed images.
-      unless insp.palette.nil?
-        img.class_type = PseudoClass
-        ret = processIndexed( img,@pageargs[:maxcolors],force )
+        # Make sure there are no more RMagick objects
+        GC.start
       end
-      ret = processMixed( img,force,map ) if ret == 0
-      img.destroy!
 
-      # Make sure there are no more RMagick objects
-      GC.start
       $stderr.puts( "Prepared data for processing #{@name}\n" )
-      return ret
+      if insp.nextImage
+        $stderr.puts( "Warning: #{@name} contains multiple images, but only the first one") 
+        $stderr.puts( "\tis going to be used\n" )
+      end
+      ret
     end
 
     def addSupplementaryFiles()

metadata

@@ -1,13 +1,13 @@
 --- !ruby/object:Gem::Specification 
 name: pdfbeads
 version: !ruby/object:Gem::Version 
-  hash: 21
+  hash: 19
   prerelease: false
   segments: 
   - 1
   - 0
-  - 1
-  version: 1.0.1
+  - 2
+  version: 1.0.2
 platform: ruby
 authors: 
 - Alexey Kryukov
@@ -15,7 +15,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2010-11-12 00:00:00 +03:00
+date: 2010-11-13 00:00:00 +03:00
 default_executable: pdfbeads
 dependencies: []