hexapdf 1.11.0 → 1.11.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: da17a2a9dd8459e6f4624f59621c84f1b0be73f4d23347bffcdd6c534b58680e
-  data.tar.gz: ecdcfedc4f89722f799d04fda467acbed24bf256a1fa626623fce1d6399d6d98
+  metadata.gz: 3167ea077cd1778bb88e07ea026c1dfc58585624b79130f38ae995ca7829077e
+  data.tar.gz: 534b414e43ce0a459a75d1b89b05793480cfc4ceb1537855a28b0bac32a46d51
 SHA512:
-  metadata.gz: 5a39a7635005c7ef8be1dc0a1efc21b1aca00fba3bedb80886e5d909d9fc0ebfed0f967fff6f10f96ed6b05a883fee0337fea3802bfcde3a93a644e512d0ee3b
-  data.tar.gz: 4e9ccdbe854d0e5fb00e7042d61c758da93c5b354982b48a20f4a6cec3a07d209dac5711d1eb9b3b6274d8662acac8851124a838bed5795cc6e40c924f4bba26
+  metadata.gz: 5d70d58918d199a7a48090e1d57ec9a8ce88ad69f53ab5935c157bb526c034f597716e566bcf2d1b969f0edaaf0c54a4baa2883e4e9953ca79dcfa4ac68c7b59
+  data.tar.gz: 425b4a95c457e909e6bb31eaee868f23fc329472e1e068222f9dc5c73c9745a023b46c741644b93fd4433bdd740c9259f5614dd481298a8aa4f18e244799020f

data/CHANGELOG.md

@@ -1,3 +1,33 @@
+## 1.11.2 - 2026-10-20
+
+### Added
+
+* [HexaPDF::Type::OptionalContentMembership#ocgs] for retrieving the list of
+  optional content groups referenced by the membership dictionary
+
+### Fixed
+
+* Regression in [HexaPDF::Task::ImportPages] concerning files with shared
+  resources
+* [HexaPDF::Type::Catalog#acro_form] to treat a malformed /AcroForm value
+  like a missing AcroForm
+* [HexaPDF::Task::MergeAcroForm] for PDFs with no default resource dictionary
+* [HexaPDF::Layout::Frame] to work when degenerate polygons are in play
+
+
+## 1.11.1 - 2026-10-06
+
+### Changed
+
+* [HexaPDF::Type::OptionalContentGroup] methods `#apply_intent`, `#on!`,
+  `#off!`, `#add_to_ui` to return self
+
+### Fixed
+
+* [HexaPDF::Type::AcroForm::Form] validation to also remove objects with an
+  invalid /Type field
+
+
 ## 1.11.0 - 2026-09-22
 
 ### Added

data/examples/027-composer_optional_content.rb

@@ -21,13 +21,14 @@ HexaPDF::Composer.create('composer_optional_content.pdf') do |composer|
     answer: {font: 'ZapfDingbats', fill_color: "green"},
   )
 
-  all = composer.document.optional_content.ocg('All answers')
-  a1 = composer.document.optional_content.ocg('Answer 1')
-  a1m = composer.document.optional_content.create_ocmd([a1, all], policy: :any_on)
-  a2 = composer.document.optional_content.ocg('Answer 2')
-  a2m = composer.document.optional_content.create_ocmd([a2, all], policy: :any_on)
-  a3 = composer.document.optional_content.ocg('Answer 3')
-  a3m = composer.document.optional_content.create_ocmd([a3, all], policy: :any_on)
+  oc = composer.document.optional_content
+  all = oc.ocg('All answers').off!.add_to_ui
+  a1 = oc.ocg('Answer 1').off!.add_to_ui
+  a2 = oc.ocg('Answer 2').off!.add_to_ui
+  a3 = oc.ocg('Answer 3').off!.add_to_ui
+  a1m = oc.create_ocmd([a1, all], policy: :any_on)
+  a2m = oc.create_ocmd([a2, all], policy: :any_on)
+  a3m = oc.create_ocmd([a3, all], policy: :any_on)
 
   composer.text('The Great Ruby Quiz', text_align: :center, margin: [0, 0, 24],
                 font: 'Helvetica bold', font_size: 24)
@@ -41,7 +42,7 @@ HexaPDF::Composer.create('composer_optional_content.pdf') do |composer|
           answers.multiple do |answer|
             answer.text('Yukihiro “Matz” Matsumoto', position: :float)
             answer.text("\u{a0}\u{a0}✔", style: :answer,
-                        properties: {'optional_content' => a1})
+                        properties: {'optional_content' => a1m})
           end
           answers.text('Rob Pike')
         end
@@ -57,7 +58,7 @@ HexaPDF::Composer.create('composer_optional_content.pdf') do |composer|
           answers.multiple do |answer|
             answer.text('1993', position: :float)
             answer.text("\u{a0}\u{a0}✔", style: :answer,
-                        properties: {'optional_content' => a2})
+                        properties: {'optional_content' => a2m})
           end
         end
       end
@@ -68,7 +69,7 @@ HexaPDF::Composer.create('composer_optional_content.pdf') do |composer|
       answer = composer.document.layout.text('There are several PDF libraries for ' \
                                              'Ruby but the best is HexaPDF! :)',
                                              width: 400,
-                                             properties: {'optional_content' => a3})
+                                             properties: {'optional_content' => a3m})
       item.formatted_text([{box: answer}], border: {width: [0, 0, 1]})
     end
   end
@@ -77,9 +78,4 @@ HexaPDF::Composer.create('composer_optional_content.pdf') do |composer|
   action.add_state_change(:toggle, [a1, a2, a3])
   composer.text("Click to toggle answers", border: {width: 1, color: "red"},
                 align: :right, padding: 2, overlays: [[:link, action: action]])
-
-  composer.document.optional_content.default_configuration(
-    BaseState: :OFF,
-    Order: [all, a1, a2, a3],
-  )
 end

data/lib/hexapdf/image_loader/png.rb

@@ -346,33 +346,49 @@ module HexaPDF
       def separate_alpha_channel(offset, decode_parms)
         bytes_per_colors = (decode_parms[:BitsPerComponent] * decode_parms[:Colors] + 7) / 8
         bytes_per_alpha = (decode_parms[:BitsPerComponent] + 7) / 8
+        bytes_per_pixel = bytes_per_colors + bytes_per_alpha
         bytes_per_row = (decode_parms[:Columns] * decode_parms[:BitsPerComponent] *
           (decode_parms[:Colors] + 1) + 7) / 8 + 1
-        image_data = ''.b
-        mask_data = ''.b
 
         flate_decode = @document.config.constantize('filter.map', :FlateDecode)
         source = flate_decode.decoder(Fiber.new(&image_data_proc(offset)))
 
+        image_deflate = Zlib::Deflate.new(HexaPDF::GlobalConfiguration['filter.flate_compression'],
+                                          Zlib::MAX_WBITS,
+                                          HexaPDF::GlobalConfiguration['filter.flate_memory'])
+        mask_deflate = Zlib::Deflate.new(HexaPDF::GlobalConfiguration['filter.flate_compression'],
+                                         Zlib::MAX_WBITS,
+                                         HexaPDF::GlobalConfiguration['filter.flate_memory'])
+
+        image_data = ''.b
+        mask_data = ''.b
+        image_temp = ''.b
+        mask_temp = ''.b
         data = ''.b
         while source.alive? && (new_data = source.resume)
           data << new_data
-          while data.length >= bytes_per_row
-            i = 1
-            image_data << data.getbyte(0)
-            mask_data << data.getbyte(0)
-            while i < bytes_per_row
-              bytes_per_colors.times {|j| image_data << data.getbyte(i + j) }
-              i += bytes_per_colors
-              bytes_per_alpha.times {|j| mask_data << data.getbyte(i + j) }
-              i += bytes_per_alpha
+          i = 0
+          image_temp.clear
+          mask_temp.clear
+          while i + bytes_per_row <= data.length
+            image_temp << data.getbyte(i)
+            mask_temp << data.getbyte(i)
+            i += 1
+            row_loop_end_index = i + bytes_per_row - 1
+            while i < row_loop_end_index
+              image_temp << data.byteslice(i, bytes_per_colors)
+              mask_temp << data.byteslice(i + bytes_per_colors, bytes_per_alpha)
+              i += bytes_per_pixel
             end
-            data = data[bytes_per_row..-1]
           end
+          image_data << image_deflate.deflate(image_temp)
+          mask_data << mask_deflate.deflate(mask_temp)
+          data = data[i..-1]
         end
-
-        image_data = Filter.string_from_source(flate_decode.encoder(Fiber.new { image_data }))
-        mask_data = Filter.string_from_source(flate_decode.encoder(Fiber.new { mask_data }))
+        image_data << image_deflate.finish
+        mask_data << mask_deflate.finish
+        image_deflate.close
+        mask_deflate.close
 
         [image_data, mask_data]
       end

data/lib/hexapdf/layout/frame.rb

@@ -438,6 +438,11 @@ module HexaPDF
         # Just use the second top-most segment
         # TODO: not the optimal solution!
         index = segments.rindex {|s| s.start_point.y < @y }
+        unless index # degenerate case where all segments are on the same horizontal line
+          @x = @y = @available_width = @available_height = 0
+          return
+        end
+
         segment = segments[index]
         y = segment.start_point.y
         polygon = if segment.min.x == @x
@@ -477,7 +482,7 @@ module HexaPDF
       # Returns the horizontal segments of the frame's shape, sorted by maximum y-, then minimum
       # x-coordinate.
       def sorted_horizontal_segments
-        @shape.each_segment.select(&:horizontal?).sort! do |a, b|
+        @shape.each_segment.select {|s| s.horizontal? && !s.degenerate? }.sort! do |a, b|
           if a.start_point.y == b.start_point.y
             b.start_point.x <=> a.start_point.x
           else

data/lib/hexapdf/task/import_pages.rb

@@ -138,14 +138,16 @@ module HexaPDF
       def self.preserve_ocgs(doc, source, pages)
         # Find all OCGs used on all pages
         ocgs = Set.new
+        seen_resources = Set.new
         process_ocg_or_ocmd = lambda do |obj|
           if obj.type == :OCG
             ocgs << obj
           elsif obj.type == :OCMD
-            ocgs.merge(obj[:OCGs].to_ary)
+            ocgs.merge(obj.ocgs)
           end
         end
         process_page_or_form = lambda do |page_or_form|
+          next unless seen_resources.add?(page_or_form.resources.data) # handle shared resources dict
           page_or_form.resources[:Properties]&.each do |name, obj|
             process_ocg_or_ocmd.call(obj) if obj
           end
@@ -154,17 +156,11 @@ module HexaPDF
             process_page_or_form.call(obj) if obj[:Subtype] == :Form
           end
         end
-        seen_resources = {}
         pages.each do |page|
-          unless seen_resources[page.resources] # handle case when pages share the resources dict
-            process_page_or_form.call(page)
-          end
-
+          process_page_or_form.call(page)
           page.each_annotation do |annot|
             process_ocg_or_ocmd.call(annot[:OC]) if annot.key?(:OC)
           end
-
-          seen_resources[page.resources] = true
         end
 
         return if ocgs.empty?

data/lib/hexapdf/task/merge_acro_form.rb

@@ -128,7 +128,7 @@ module HexaPDF
         font_name_mapping = {}
         serializer = HexaPDF::Serializer.new
 
-        source_form.default_resources[:Font].each do |font_name, value|
+        source_form.default_resources[:Font]&.each do |font_name, value|
           new_name = target_resources.add_font(target_form.document.import(value))
           font_name_mapping[serializer.serialize(font_name)] = serializer.serialize(new_name)
         end

data/lib/hexapdf/type/acro_form/form.rb

@@ -573,7 +573,8 @@ module HexaPDF
 
           validate_array = lambda do |parent, container|
             container.map! do |field|
-              if !field.kind_of?(HexaPDF::Object) || !field.kind_of?(HexaPDF::Dictionary) || field.null?
+              if !field.kind_of?(HexaPDF::Object) || !field.kind_of?(HexaPDF::Dictionary) ||
+                 field.null? || (field.key?(:Type) && field[:Type] != :Annot)
                 yield("Invalid object in AcroForm field hierarchy", true)
                 next nil
               end

data/lib/hexapdf/type/catalog.rb

@@ -133,16 +133,16 @@ module HexaPDF
 
       # Returns the main AcroForm object.
       #
-      # * If an AcroForm object exists, the +create+ argument is not used.
+      # * If a valid AcroForm object exists, it is returned regardless of the +create+ argument.
       #
-      # * If no AcroForm object exists and +create+ is +true+, a new AcroForm object with default
-      #   settings will be created and returned.
+      # * If no valid AcroForm object exists and +create+ is +true+, a new AcroForm object with
+      #   default settings will be created and returned.
       #
       # * If no AcroForm object exists and +create+ is +false+, +nil+ is returned.
       #
       # See: AcroForm::Form
       def acro_form(create: false)
-        if (form = self[:AcroForm])
+        if (form = self[:AcroForm]).kind_of?(Dictionary)
           form
         elsif create
           form = self[:AcroForm] = document.add({}, type: :XXAcroForm)

data/lib/hexapdf/type/optional_content_group.rb

@@ -187,10 +187,14 @@ module HexaPDF
         end
       end
 
-      # Applies the given intent (:View, :Design or a custom intent) to the OCG.
+      # :call-seq:
+      #   ocg.apply_intent(intent)   -> ocg
+      #
+      # Applies the given intent (:View, :Design or a custom intent) to the OCG and returns +self+.
       def apply_intent(intent)
         self[:Intent] = key?(:Intent) ? Array(self[:Intent]) : []
         self[:Intent] << intent
+        self
       end
 
       # Returns +true+ if this OCG has an intent of :View.
@@ -209,23 +213,35 @@ module HexaPDF
         document.optional_content.default_configuration.ocg_on?(self)
       end
 
+      # :call-seq:
+      #   ocg.on!   -> ocg
+      #
       # Sets the state of the OCG to on in the default configuration (see
-      # OptionalContentProperties#default_configuration).
+      # OptionalContentProperties#default_configuration) and returns +self+.
       def on!
         document.optional_content.default_configuration.ocg_state(self, :on)
+        self
       end
 
+      # :call-seq:
+      #   ocg.off!   -> ocg
+      #
       # Sets the state of the OCG to off in the default configuration (see
-      # OptionalContentProperties#default_configuration).
+      # OptionalContentProperties#default_configuration) and returns +self+.
       def off!
         document.optional_content.default_configuration.ocg_state(self, :off)
+        self
       end
 
+      # :call-seq:
+      #   ocg.add_to_ui(path: nil)    -> ocg
+      #
       # Adds the OCG to the PDF processor's user interface in the default configuration (see
       # OptionalContentProperties#default_configuration), either at the top-level or under the given
-      # hierarchical +path+ but always as the last item.
+      # hierarchical +path+ but always as the last item, and returns +self+.
       def add_to_ui(path: nil)
         document.optional_content.default_configuration.add_ocg_to_ui(self, path: path)
+        self
       end
 
       # :call-seq:

data/lib/hexapdf/type/optional_content_membership.rb

@@ -57,6 +57,19 @@ module HexaPDF
                    allowed_values: [:AllOn, :AnyOn, :AnyOff, :AllOff]
       define_field :VE,   type: PDFArray
 
+      # Returns an array of OCGs whose states determine the visibility of content controlled by this
+      # membership dictionary.
+      def ocgs
+        if (ve = self[:VE])
+          ocgs = ve.to_ary.flatten.map! {|entry| document.deref(entry) }
+          ocgs.reject! {|entry| entry.kind_of?(Symbol) || entry.nil? || entry.null? }
+          ocgs.uniq!
+          ocgs
+        else
+          Array(self[:OCGs]).compact.uniq
+        end
+      end
+
     end
 
   end

data/lib/hexapdf/type/struct_tree_root.rb

@@ -54,7 +54,7 @@ module HexaPDF
       define_field :ParentTreeNextKey,     type: Integer
       define_field :RoleMap,               type: Dictionary
       define_field :ClassMap,              type: Dictionary
-      define_field :Namespaces,            type: PDFArray
+      define_field :Namespaces,            type: PDFArray, version: '2.0'
       define_field :PronounciationLexicon, type: PDFArray, version: '2.0'
       define_field :AF,                    type: PDFArray, version: '2.0'
 

data/lib/hexapdf/version.rb

@@ -37,6 +37,6 @@
 module HexaPDF
 
   # The version of HexaPDF.
-  VERSION = '1.11.0'
+  VERSION = '1.11.2'
 
 end

data/test/hexapdf/layout/test_frame.rb

@@ -502,6 +502,22 @@ describe HexaPDF::Layout::Frame do
       check_regions(frame, [[0, 100, 30, 100], [0, 60, 100, 60]])
     end
 
+    # o--------+
+    #
+    # Such a degenerate case can occur under special circumstances with e.g. a ListBox:
+    #
+    #   composer.list(height: 10, style: {padding: 5}) do |list|
+    #     list.text('test')
+    #   end
+    #
+    # Padding exactly fills the box, leaving zero height, and the algorithm in ListBox creates a
+    # degenerate Polygon.
+    it "works for a flat region with no height" do
+      shape = Geom2D::Polygon([0, 0], [100, 0], [100, 0], [0, 0])
+      frame = HexaPDF::Layout::Frame.new(0, 0, 100, 0, shape: shape)
+      check_regions(frame, [[0, 0, 100, 0]])
+    end
+
     # o---+     +--+
     # |   |  +--+  |
     # |   +--+     |

data/test/hexapdf/task/test_import_pages.rb

@@ -100,6 +100,7 @@ describe HexaPDF::Task::ImportPages do
       form = canvas.form
       form[:OC] = @ocg1
       form.canvas.optional_content(@ocmd)
+      form.resources.add_xobject(form)
       canvas.xobject(form, at: [0, 0])
       @target.task(:import_pages, source: @doc)
       assert_equal(['OCG', 'OCMD'], @target.optional_content.ocgs.map(&:name))

data/test/hexapdf/task/test_merge_acro_form.rb

@@ -47,6 +47,12 @@ describe HexaPDF::Task::MergeAcroForm do
     assert(@doc.acro_form.default_resources[:Font].key?(:F2))
   end
 
+  it "handles AcroForm dictionaries without /DR entry" do
+    @doc1.acro_form.delete(:DR)
+    @doc.task(:merge_acro_form, source: @doc1, pages: @pages)
+    refute(@doc.acro_form.default_resources[:Font].key?(:F2))
+  end
+
   it "updates the /SigFlags if necessary" do
     @doc.task(:merge_acro_form, source: @doc1, pages: [@pages[0]])
     refute(@doc.acro_form.signature_flag?(:signatures_exist))

data/test/hexapdf/test_document.rb

@@ -498,8 +498,14 @@ describe HexaPDF::Document do
   describe "acro_form" do
     it "returns the main AcroForm object" do
       assert_nil(@doc.acro_form)
+
+      @doc.acro_form(create: true)
+      refute_nil(@doc.acro_form)
+    end
+
+    it "treats a malformed /AcroForm value like a missing AcroForm" do
       @doc.catalog[:AcroForm] = 5
-      assert_equal(5, @doc.acro_form)
+      assert_nil(@doc.acro_form)
     end
 
     it "create the AcroForm object if instructed" do

data/test/hexapdf/type/acro_form/test_form.rb

@@ -537,6 +537,7 @@ describe HexaPDF::Type::AcroForm::Form do
           @doc.add({T: :Tx3, FT: :Tx, Kids: [@doc.add({T: :Tx4}),
                                              [:nothing],
                                              @doc.add({T: :Tx5, Kids: [@doc.add({T: :Tx6})]})]}),
+          @doc.wrap({Type: :Page}),
         ]
         @acro_form[:Fields][6][:Kids][0][:Parent] = @acro_form[:Fields][6]
         @acro_form[:Fields][6][:Kids][2][:Parent] = @acro_form[:Fields][6]

data/test/hexapdf/type/test_catalog.rb

@@ -62,19 +62,32 @@ describe HexaPDF::Type::Catalog do
 
   describe "acro_form" do
     it "returns an existing form object" do
-      @catalog[:AcroForm] = :test
-      assert_equal(:test, @catalog.acro_form)
+      @catalog[:AcroForm] = {}
+      form = @catalog[:AcroForm]
+      assert_same(form, @catalog.acro_form)
+    end
+
+    it "creates a new AcroForm object with defaults if create: true" do
+      form = @catalog.acro_form(create: true)
+      assert_kind_of(HexaPDF::Type::AcroForm::Form, form)
+      assert(form[:DA])
     end
 
     it "returns an existing form object even if create: true" do
+      form = @catalog.acro_form(create: true)
+      assert_same(form, @catalog.acro_form(create: true))
+    end
+
+    it "treats an invalid /AcroForm value like a missing AcroForm" do
       @catalog[:AcroForm] = :test
-      assert_equal(:test, @catalog.acro_form(create: true))
+      assert_nil(@catalog.acro_form)
     end
 
-    it "creates a new AcroForm object with defaults if create: true" do
+    it "replaces an invalid /AcroForm value if create: true" do
+      @catalog[:AcroForm] = :test
       form = @catalog.acro_form(create: true)
       assert_kind_of(HexaPDF::Type::AcroForm::Form, form)
-      assert(form[:DA])
+      assert_same(form, @catalog.acro_form)
     end
   end
 

data/test/hexapdf/type/test_optional_content_group.rb

@@ -69,7 +69,8 @@ describe HexaPDF::Type::OptionalContentGroup do
     end
 
     it "can apply one or more intents" do
-      @ocg.apply_intent(:View)
+      ocg = @ocg.apply_intent(:View)
+      assert_same(@ocg, ocg)
       @ocg.apply_intent(:Design)
       assert(@ocg.intent_view?)
       assert(@ocg.intent_design?)
@@ -85,17 +86,20 @@ describe HexaPDF::Type::OptionalContentGroup do
 
     it "can set its default state to on" do
       @doc.optional_content.default_configuration[:OFF] = [@ocg]
-      @ocg.on!
+      ocg = @ocg.on!
+      assert_same(@ocg, ocg)
       assert(@ocg.on?)
     end
 
     it "can set its default state to off" do
-      @ocg.off!
+      ocg = @ocg.off!
+      assert_same(@ocg, ocg)
       refute(@ocg.on?)
     end
 
     it "can add itself to the UI" do
-      @ocg.add_to_ui(path: 'Test')
+      ocg = @ocg.add_to_ui(path: 'Test')
+      assert_same(@ocg, ocg)
       assert_equal([['Test', @ocg]], @doc.optional_content.default_configuration[:Order].value)
     end
   end

data/test/hexapdf/type/test_optional_content_membership.rb

@@ -0,0 +1,40 @@
+# -*- encoding: utf-8 -*-
+
+require 'test_helper'
+require 'hexapdf/type/optional_content_membership'
+require 'hexapdf/document'
+
+describe HexaPDF::Type::OptionalContentMembership do
+  before do
+    @doc = HexaPDF::Document.new
+    @ocmd = @doc.add({Type: :OCMD})
+  end
+
+  describe "ocgs" do
+    before do
+      @ocg = @doc.add({Type: :OCG, Name: "test"})
+    end
+
+    it "uses the /OCGs entry if /VE is not available" do
+      assert_equal([], @ocmd.ocgs)
+
+      @ocmd[:OCGs] = []
+      assert_equal([], @ocmd.ocgs)
+
+      @ocmd[:OCGs] = @ocg
+      assert_equal([@ocg], @ocmd.ocgs)
+
+      null = @doc.add({})
+      null.data.value = nil
+      @ocmd[:OCGS] = [@ocg, nil, null, @ocg]
+      assert_equal([@ocg], @ocmd.ocgs)
+    end
+
+    it "uses the /VE entry if available" do
+      null = @doc.add({})
+      null.data.value = nil
+      @ocmd[:VE] = [:Or, @ocg, nil, [:Not, @ocg], [:And, @ocg, [:Or, null, @ocg]]]
+      assert_equal([@ocg], @ocmd.ocgs)
+    end
+  end
+end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: hexapdf
 version: !ruby/object:Gem::Version
-  version: 1.11.0
+  version: 1.11.2
 platform: ruby
 authors:
   - Thomas Leitner
@@ -922,6 +922,7 @@ files:
   - test/hexapdf/type/test_object_stream.rb
   - test/hexapdf/type/test_optional_content_configuration.rb
   - test/hexapdf/type/test_optional_content_group.rb
+  - test/hexapdf/type/test_optional_content_membership.rb
   - test/hexapdf/type/test_optional_content_properties.rb
   - test/hexapdf/type/test_outline.rb
   - test/hexapdf/type/test_outline_item.rb