hexapdf 0.9.3 → 0.10.0

data/lib/hexapdf/font/true_type_wrapper.rb

@@ -237,9 +237,9 @@ module HexaPDF
         end
 
         fd.flag(:fixed_pitch) if @wrapped_font[:post].is_fixed_pitch? ||
-            @wrapped_font[:hhea].num_of_long_hor_metrics == 1
+          @wrapped_font[:hhea].num_of_long_hor_metrics == 1
         fd.flag(:italic) if @wrapped_font[:'OS/2'].selection_include?(:italic) ||
-            @wrapped_font[:'OS/2'].selection_include?(:oblique)
+          @wrapped_font[:'OS/2'].selection_include?(:oblique)
         fd.flag(:symbolic)
 
         cid_font = @document.add(Type: :Font, Subtype: :CIDFontType2,

data/lib/hexapdf/font/type1/afm_parser.rb

@@ -128,7 +128,7 @@ module HexaPDF
           when :CharacterSet then @metrics.character_set = parse_string
           when :EncodingScheme then @metrics.encoding_scheme = parse_string
           when :Weight then @metrics.weight = parse_string
-          when :FontBBox then
+          when :FontBBox
             @metrics.bounding_box = [parse_number, parse_number, parse_number, parse_number]
           when :CapHeight then @metrics.cap_height = parse_number
           when :XHeight then @metrics.x_height = parse_number

data/lib/hexapdf/image_loader/jpeg.rb

@@ -41,7 +41,7 @@ module HexaPDF
 
     # This module is used for loading images in the JPEG format from files or IO streams.
     #
-    # See: PDF1.7 s7.4.8, ITU T.81 Annex B
+    # See: PDF1.7 s7.4.8, ITU T.81 Annex B, ITU T.872
     module JPEG
 
       # The magic marker that tells us if the file/IO contains an image in JPEG format.
@@ -56,13 +56,13 @@ module HexaPDF
         0xC0, 0xC1, 0xC2, 0xC3, 0xC5, 0xC6, 0xC7, 0xC9, 0xCA, 0xCB, 0xCD, 0xCE, 0xCF
       ].freeze
 
-      # Adobe uses the marker 0xEE (APPE) for its purposes. We need to use it for determinig
-      # whether to invert the colors for CMYK/YCCK images or not (Adobe does this...).
-      #
-      # The marker also let's us distinguish between YCCK and CMYK images. However, we don't
-      # actually need this information (and we don't need to set the /ColorTransform value)
-      # because if the image has this information it is automically used.
-      ADOBE_MARKER = 0xEE
+      # Adobe uses the marker 0xEE (APPE or APP14) for its purposes. We need to use it for
+      # determinig whether we have a CMYK or YCCK image.
+      APP14_MARKER = 0xEE
+
+      # Value of the 12th byte in an APP14 marker specifying that the image uses CMYK color
+      # encoding, with all four colors complemented.
+      APP14_TRANSFORM_CMYK = 0
 
       # End-of-image marker
       EOI_MARKER = 0xD9
@@ -121,9 +121,11 @@ module HexaPDF
           # but those shouldn't appear here)
           length = io.read(2).unpack1('n')
 
-          if code1 == ADOBE_MARKER # Adobe apps invert the colors when using CMYK color space
-            invert_colors = true
-            io.seek(length - 2, IO::SEEK_CUR)
+          # According to T.872 6.1 and 6.5.3, if this marker is present, we need to use it for
+          # correctly determining whether complemented CMYK or YCCK is used
+          if code1 == APP14_MARKER
+            io.seek(length - 3, IO::SEEK_CUR)
+            invert_colors = true if io.getbyte == APP14_TRANSFORM_CMYK
             next
           elsif !SOF_MARKERS.include?(code1)
             io.seek(length - 2, IO::SEEK_CUR)

data/lib/hexapdf/reference.rb

@@ -90,6 +90,11 @@ module HexaPDF
       oid.hash ^ gen.hash
     end
 
+    # Returns the object identifier as "oid,gen".
+    def to_s
+      "#{oid} #{gen} R"
+    end
+
     def inspect #:nodoc:
       "#<#{self.class.name} [#{oid}, #{gen}]>"
     end

data/lib/hexapdf/revision.rb

@@ -89,6 +89,20 @@ module HexaPDF
       ((a = @xref_section.max_oid) < (b = @objects.max_oid) ? b : a) + 1
     end
 
+    # :call-seq:
+    #   revision.xref(ref)    -> xref_entry or nil
+    #   revision.xref(oid)    -> xref_entry or nil
+    #
+    # Returns an XRefSection::Entry structure for the given reference or object number if it is
+    # available, or +nil+ otherwise.
+    def xref(ref)
+      if ref.respond_to?(:oid)
+        @xref_section[ref.oid, ref.gen]
+      else
+        @xref_section[ref, nil]
+      end
+    end
+
     # :call-seq:
     #   revision.object(ref)    -> obj or nil
     #   revision.object(oid)    -> obj or nil

data/lib/hexapdf/serializer.rb

@@ -186,7 +186,7 @@ module HexaPDF
       NAME_CACHE[obj] ||=
         begin
           str = obj.to_s.force_encoding(Encoding::BINARY)
-          str.gsub!(NAME_REGEXP) {|m| NAME_SUBSTS[m] }
+          str.gsub!(NAME_REGEXP, NAME_SUBSTS)
           "/#{str}"
         end
     end
@@ -203,7 +203,7 @@ module HexaPDF
       while index < obj.size
         tmp = __serialize(obj[index])
         str << " " unless BYTE_IS_DELIMITER[tmp.getbyte(0)] ||
-            BYTE_IS_DELIMITER[str.getbyte(-1)]
+          BYTE_IS_DELIMITER[str.getbyte(-1)]
         str << tmp
         index += 1
       end
@@ -217,10 +217,10 @@ module HexaPDF
       str = +"<<"
       obj.each do |k, v|
         next if v.nil? || (v.respond_to?(:null?) && v.null?)
-        str << __serialize(k)
+        str << serialize_symbol(k)
         tmp = __serialize(v)
         str << " " unless BYTE_IS_DELIMITER[tmp.getbyte(0)] ||
-            BYTE_IS_DELIMITER[str.getbyte(-1)]
+          BYTE_IS_DELIMITER[str.getbyte(-1)]
         str << tmp
       end
       str << ">>"
@@ -311,7 +311,7 @@ module HexaPDF
 
       if @io && fiber.respond_to?(:length) && fiber.length >= 0
         obj.value[:Length] = fiber.length
-        @io << __serialize(obj.value)
+        @io << serialize_hash(obj.value)
         @io << "stream\n"
         while fiber.alive? && (data = fiber.resume)
           @io << data.freeze
@@ -323,7 +323,7 @@ module HexaPDF
         data = Filter.string_from_source(fiber)
         obj.value[:Length] = data.size
 
-        str = __serialize(obj.value)
+        str = serialize_hash(obj.value)
         str << "stream\n"
         str << data
         str << "\nendstream"

data/lib/hexapdf/type/annotations/markup_annotation.rb

@@ -47,7 +47,7 @@ module HexaPDF
       class MarkupAnnotation < Annotation
 
         define_field :T,            type: String, version: '1.1'
-        define_field :Popup,        type: :Annotation, version: '1.3'
+        define_field :Popup,        type: :Annot, version: '1.3'
         define_field :CA,           type: Numeric, default: 1.0, version: '1.4'
         define_field :RC,           type: [Stream, String], version: '1.5'
         define_field :CreationDate, type: PDFDate, version: '1.5'

data/lib/hexapdf/type/image.rb

@@ -253,8 +253,10 @@ module HexaPDF
         if filter == :FlateDecode && self[:DecodeParms] && self[:DecodeParms][:Predictor].to_i >= 10
           data = stream_source
         else
+          colors = (color_type == ImageLoader::PNG::INDEXED ? 1 : info.components)
           flate_decode = config.constantize('filter.map', :FlateDecode)
-          data = flate_decode.encoder(stream_decoder, Predictor: 15, Colors: 1, Columns: info.width,
+          data = flate_decode.encoder(stream_decoder, Predictor: 15,
+                                      Colors: colors, Columns: info.width,
                                       BitsPerComponent: info.bits_per_component)
         end
         io << png_chunk('IDAT', Filter.string_from_source(data))

data/lib/hexapdf/version.rb

@@ -37,6 +37,6 @@
 module HexaPDF
 
   # The version of HexaPDF.
-  VERSION = '0.9.3'
+  VERSION = '0.10.0'
 
 end

data/lib/hexapdf/xref_section.rb

@@ -82,6 +82,14 @@ module HexaPDF
       def compressed?
         type == :compressed
       end
+
+      def to_s
+        case type
+        when :free then "xref #{oid},#{gen} type=free"
+        when :in_use then "xref #{oid},#{gen} type=normal pos=#{pos}"
+        when :compressed then "xref #{oid},#{gen} type=compressed objstm=#{objstm},0 index=#{pos}"
+        end
+      end
     end
 
     # Creates an in-use cross-reference entry. See Entry for details on the arguments.

data/man/man1/hexapdf.1

@@ -189,7 +189,9 @@ This command prints the application help if no arguments are given\. If one or m
 .SS "images"
 Synopsis: \fBimages\fP [\fBOPTIONS\fP] \fIPDF\fP
 .P
-This command extracts images from the \fIPDF\fP\&\. If the \fB\-\-extract\fP option is not specified, the images are listed with their indices and additional information, sorted by page number\. The \fB\-\-extract\fP option can then be used to extract one or more images, saving them to files called \fBPREFIX\-N\.EXT\fP where the prefix can be set via \fB\-\-prefix\fP, \fIN\fP is the image index and \fIEXT\fP is either png, jpg or jpx\.
+This command extracts images from the \fIPDF\fP\&\. If the \fB\-\-extract\fP option is not specified, the images are listed with their indices and additional information, sorted by page number\. Note that if an image is used multiple times on a page, only the first occurence of it will be included\.
+.P
+The \fB\-\-extract\fP option can then be used to extract one or more images, saving them to files called \fBPREFIX\-N\.EXT\fP where the prefix can be set via \fB\-\-prefix\fP, \fIN\fP is the image index and \fIEXT\fP is either png, jpg or jpx\.
 .TP
 \fB\-e\fP [\fIA,B,C,\.\.\.\fP], \fB\-\-extract\fP [\fIA,B,C,\.\.\.\fP]
 The indices of the images that should be extracted\. Use \fI0\fP or no value to extract all images\.
@@ -230,6 +232,15 @@ The number of color components\.
 \fBbpc\fP
 The number of bits per color component\.
 .TP
+\fBx\-ppi\fP
+The pixels per inch (PPI) of the x\-direction of the image, as found on the page\.
+.TP
+\fBy\-ppi\fP
+The pixels per inch (PPI) of the y\-direction of the image, as found on the page\.
+.TP
+\fBsize\fP
+The file size of the image as stored in the PDF\.
+.TP
 \fBtype\fP
 The image type\. Either jpg (JPEG), jp2 (JPEG2000), ccitt (CCITT Group 3 or 4 Fax), jbig2 (JBIG2) or png (PNG)\.
 .TP
@@ -244,32 +255,87 @@ This command reads the \fIFILE\fP and shows general information about it, like a
 \fB\-p\fP \fIPASSWORD\fP, \fB\-\-password\fP \fIPASSWORD\fP
 The password to decrypt the PDF \fIFILE\fP\&\. Use \fB\-\fP for \fIPASSWORD\fP for reading it from standard input\.
 .SS "inspect"
-Synopsis: \fBinspect\fP [\fBOPTIONS\fP] \fIFILE\fP
+Synopsis: \fBinspect\fP [\fBOPTIONS\fP] \fIFILE\fP \fI[[CMD [ARGS]]\.\.\.]\fP
 .P
 This command is useful when one needs to inspect the internal object structure or a stream of a PDF file\.
 .P
-If no option is given, the PDF trailer is shown\. Otherwise the various, mutually exclusive display options define what is shown\. If multiple such options are specified only the last one is respected\. Note that PDF objects are always shown in the native PDF syntax\.
+If no arguments are given, the interactive mode is started\. This interactive mode allows you to execute inspection commands without re\-parsing the PDF file, leading to better performance for big PDF files\.
+.P
+Otherwise the arguments are interpreted as interactive mode commands and executed\. It is possible to specify more than one command in this way by separating them with semicolons, or whitespace in case the number of command arguments is fixed\.
 .TP
-\fB\-\-catalog\fP
-Show the PDF catalog dictionary\.
+\fB\-p\fP \fIPASSWORD\fP, \fB\-\-password\fP \fIPASSWORD\fP
+The password to decrypt the PDF \fIFILE\fP\&\. Use \fB\-\fP for \fIPASSWORD\fP for reading it from standard input\.
+.P
+If an interactive mode command or argument is \fBOID[,GEN]\fP, object and generation numbers are expected\. The generation number defaults to 0 if not given\. PDF objects are always shown in the native PDF syntax\.
+.P
+The available commands are:
 .TP
-\fB\-c\fP, \fB\-\-page\-count\fP
-Print the number of pages\.
+\fBOID[,GEN] | o[bject] OID[,GEN]\fP
+Print the given indirect object\.
 .TP
-\fB\-\-pages\fP [\fIPAGES\fP]
-Show the pages with their object and generation numbers and their associated content streams\. If a range is specified, only those pages are listed\. See the \fBPAGES SPECIFICATION\fP below for details on the allowed format of \fIPAGES\fP\&\.
+\fBr[ecursive] OID[,GEN]\fP
+Print the given indirect object recursively\. This means that all references found in the object are resolved and the resulting objects themselves recursively printed\.
+.RS
+.P
+To make it easier to compare such structures between PDF files, the entries of dictionaries are printed in sorted order and the original references are replaced by custom ones\. Once an indirect object is first encountered, it is preceeded by either \fB{obj INDEX}\fP or \fB{obj page PAGEINDEX}\fP where \fBINDEX\fP is an increasing number and \fBPAGEINDEX\fP is the index of the page\. Later references are replaced by \fB{ref INDEX}\fP and \fB{ref page PAGEINDEX}\fP respectively\.
+.P
+Here is a simplified example output:
+.sp
+.RS 4
+.EX
+<<
+  /Info {obj 1} <<
+    /Producer (HexaPDF version 0\.9\.3)
+  >>
+  /Root {obj 2} <<
+    /Pages {obj 3} <<
+      /Count 1
+      /Kids [{obj page 1} <<
+        /MediaBox [0 0 595 842 ]
+        /Parent {ref 3}
+        /Type /Page
+      >> ]
+      /Type /Pages
+    >>
+    /Type /Catalog
+  >>
+  /Size 4
+>>
+.EE
+.RE
+.P
+On line 2 the indirect object for the key \fB/Info\fP is shown, preceeded by the custom reference\. On line 8 is an example for a page object with the special reference key\. And on line 10 there is a back reference to the object with index 3 which is started on line 6\.
+.RE
 .TP
-\fB\-o\fP \fIOID\fP[,\fIGEN\fP], \fB\-\-object\fP \fIOID\fP[,\fIGEN\fP]
-Show the object with the given object and generation numbers\. The generation number defaults to 0 if not given\.
+\fBs[tream] OID[,GEN]\fP
+Print the filtered stream, i\.e\. the stream with all filters applied\. This is useful, for example, to view the contents of content streams\.
 .TP
-\fB\-s\fP \fIOID\fP[,\fIGEN\fP], \fB\-\-stream\fP \fIOID\fP[,\fIGEN\fP]
-Show the filtered stream data (add \fB\-\-raw\fP to get the raw stream data) of the object with the given object and generation numbers\. The generation number defaults to 0 if not given\.
+\fBraw[\-stream] OID[,GEN]\fP
+Print the raw stream, i\.e\. the stream as it appears in the file\. This is useful, for example, to extract streams into files\.
 .TP
-\fB\-\-raw\fP
-Modifies \fB\-\-stream\fP to show the raw stream data instead of the filtered one\.
+\fBx[ref] OID[,GEN]\fP
+Print the cross\-reference entry for the given indirect object\.
 .TP
-\fB\-p\fP \fIPASSWORD\fP, \fB\-\-password\fP \fIPASSWORD\fP
-The password to decrypt the PDF \fIFILE\fP\&\. Use \fB\-\fP for \fIPASSWORD\fP for reading it from standard input\.
+\fBc[atalog]\fP
+Print the catalog dictionary\.
+.TP
+\fBt[railer]\fP
+Print the trailer dictionary\.
+.TP
+\fBp[ages] [RANGE]\fP
+Print the pages with their object and generation numbers and their associated content streams\. If a range is specified, only those pages are listed\. See the \fBPAGES SPECIFICATION\fP below for details on the allowed format of \fIPAGES\fP\&\.
+.TP
+\fBpc | page\-count\fP
+Print the number of pages\.
+.TP
+\fBsearch REGEXP\fP
+Print all objects matching the pattern\. Each object is preceeded by \fBobj OID GEN\fP and followed by \fBendobj\fP to make it easier to further explore the data\.
+.TP
+\fBh[elp]\fP
+Print the available commands with a short description\.
+.TP
+\fBq[uit]Quit\fP
+Quit the interactive mode\.
 .SS "merge"
 Synopsis: \fBmerge\fP [\fBOPTIONS\fP] { \fIINPUT\fP | \fB\-\-empty\fP } [\fIINPUT\fP]\.\.\. \fIOUTPUT\fP
 .P
@@ -426,11 +492,13 @@ Image info and extraction: The first command lists the images of the \fBinput\.p
 .P
 File information: Show general information about the PDF file, like PDF version, number of pages, creator, creation date and encryption related information\.
 .SS "inspect"
-\fBhexapdf inspect input\.pdf\fP
-.br
 \fBhexapdf inspect input\.pdf \-o 3\fP
 .P
-Inspect a PDF: These commands can be used to inspect the internal object structure of a PDF file\. The first command shows the PDF trailer object\. The second one shows the object with the object number 3\.
+Show the object with the object number 3 of the given PDF file\.
+.P
+\fBhexapdf inspect input\.pdf\fP
+.P
+Start the interactive inspection mode\.
 .SS "batch"
 \fBhexapdf batch \'info {}\' input1\.pdf input2\.pdf input3\.pdf\fP
 .P

data/test/hexapdf/content/test_canvas.rb

@@ -777,9 +777,9 @@ describe HexaPDF::Content::Canvas do
 
   describe "xobject" do
     before do
-      @image = @doc.add(Subtype: :Image, Width: 10, Height: 5)
+      @image = @doc.add(Type: :XObject, Subtype: :Image, Width: 10, Height: 5)
       @image.source_path = File.join(TEST_DATA_DIR, 'images', 'gray.jpg')
-      @form = @doc.add(Subtype: :Form, BBox: [100, 50, 200, 100])
+      @form = @doc.add(Type: :XObject, Subtype: :Form, BBox: [100, 50, 200, 100])
     end
 
     it "can use any xobject specified via a filename" do

data/test/hexapdf/document/test_images.rb

@@ -16,7 +16,7 @@ describe HexaPDF::Document::Images do
         @loader.define_singleton_method(:handles?) {|*| true }
         @loader.define_singleton_method(:load) do |doc, s|
           s = HexaPDF::StreamData.new(s) if s.kind_of?(IO)
-          doc.add({Subtype: :Image}, stream: s)
+          doc.add({Type: :XObject, Subtype: :Image}, stream: s)
         end
         @doc.config['image_loader'].unshift(@loader)
       end
@@ -60,10 +60,12 @@ describe HexaPDF::Document::Images do
     it "iterates over all non-mask images" do
       @doc.add(5)
       images = []
-      images << @doc.add(Subtype: :Image)
-      images << @doc.add(Subtype: :Image, Mask: [5, 6])
-      images << @doc.add(Subtype: :Image, Mask: @doc.add(Subtype: :Image))
-      images << @doc.add(Subtype: :Image, SMask: @doc.add(Subtype: :Image))
+      images << @doc.add(Type: :XObject, Subtype: :Image)
+      images << @doc.add(Type: :XObject, Subtype: :Image, Mask: [5, 6])
+      images << @doc.add(Type: :XObject, Subtype: :Image,
+                         Mask: @doc.add(Type: :XObject, Subtype: :Image))
+      images << @doc.add(Type: :XObject, Subtype: :Image,
+                         SMask: @doc.add(Type: :XObject, Subtype: :Image))
       assert_equal(images.sort, @doc.images.to_a.sort)
     end
   end

data/test/hexapdf/encryption/test_security_handler.rb

@@ -298,6 +298,12 @@ describe HexaPDF::Encryption::SecurityHandler do
       assert_equal(@encrypted, @handler.decrypt(@obj)[:Key])
     end
 
+    it "doesn't decrypt the /Contents of a signature dictionary" do
+      @obj[:Type] = :Sig
+      @obj[:Contents] = "test"
+      assert_equal("test", @handler.decrypt(@obj)[:Contents])
+    end
+
     it "fails if V < 5 and the object number changes" do
       @obj.oid = 55
       @handler.decrypt(@obj)
@@ -332,6 +338,13 @@ describe HexaPDF::Encryption::SecurityHandler do
       @stream[:Type] = :XRef
       assert_equal('string', @handler.encrypt_stream(@stream).resume)
     end
+
+    it "doesn't encrypt the /Contents key of signature dictionaries" do
+      @obj[:Type] = :Sig
+      @obj[:Contents] = "test"
+      refute_equal('test', @handler.encrypt_string("test", @obj))
+      assert_equal('test', @handler.encrypt_string(@obj[:Contents], @obj))
+    end
   end
 
   it "works correctly with different decryption and encryption handlers" do

data/test/hexapdf/image_loader/test_jpeg.rb

@@ -66,6 +66,16 @@ describe HexaPDF::ImageLoader::JPEG do
       assert_equal(File.binread(jpeg), image.stream)
     end
 
+    it "works for a YCCK jpeg" do
+      jpeg = @images.grep(/ycck\.jpg/).first
+      image = @loader.load(@doc, jpeg)
+      assert_equal(5, image[:Width])
+      assert_equal(5, image[:Height])
+      assert_equal(:DeviceCMYK, image[:ColorSpace])
+      refute(image.key?(:Decode))
+      assert_equal(File.binread(jpeg), image.stream)
+    end
+
     it "fails if the JPEG is corrupt" do
       exp = assert_raises(HexaPDF::Error) do
         @loader.load(@doc, StringIO.new("some non JPEG data"))

data/test/hexapdf/image_loader/test_png.rb

@@ -206,8 +206,9 @@ describe HexaPDF::ImageLoader::PNG do
     end
 
     it "works for a true color 8-bit png with alpha" do
-      png = @images.grep(/truecolour-alpha-8bit\.png/).first
-      image = @loader.load(@doc, png)
+      png_data = File.binread(@images.grep(/truecolour-alpha-8bit\.png/).first)
+      png_data[33, 0] = [0, "tRNS", 0].pack('NA4N') # add invalid tRNS chunk
+      image = @loader.load(@doc, StringIO.new(png_data))
       data = [[12, 92, 146, 80, 136, 175, 167, 193, 213, 97, 175, 101, 38, 113, 50],
               [12, 92, 146, 81, 137, 176, 168, 194, 214, 97, 175, 101, 38, 113, 49],
               [12, 92, 146, 81, 137, 176, 169, 195, 214, 96, 175, 101, 37, 113, 49],

data/test/hexapdf/test_dictionary.rb

@@ -230,6 +230,15 @@ describe HexaPDF::Dictionary do
 
       @obj.value[:Inherited] = Class.new(Array).new([5])
       assert(@obj.validate(auto_correct: false))
+
+      @test_class.define_field(:StringField, type: String)
+      @test_class.define_field(:NameField, type: Symbol)
+      @obj.value[:StringField] = :symbol
+      refute(@obj.validate(auto_correct: false))
+      assert(@obj.validate(auto_correct: true))
+      @obj.value[:NameField] = "string"
+      assert(@obj.validate(auto_correct: true))
+      assert(@obj.validate(auto_correct: true))
     end
 
     it "checks whether a field needs to be indirect w/wo auto_correct" do