hexapdf 1.0.0 → 1.0.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 2889ba1d03e2c351efd694b1583063023fff97c0da636ff5103f88538255735c
-  data.tar.gz: 6fb4727db05900e8fccba2ad4e093d1092e17305e5b5616ded97a76cf835673c
+  metadata.gz: 61b0fb56c6522f2af82eb8ffb10570c45bb11460cf4c048c1bdfe8d9daf71afe
+  data.tar.gz: 91cb053019c367825ac0799a84e4ddad837fe283a6ab2bc6df16ee9ed9f2456d
 SHA512:
-  metadata.gz: 00be8ed2c306a88e5bfc0eada97a7e6bf802ec269e832bb21b3521d4077b18ecad11946ddc6f8a6d575820e66339059e59ba2c4cdd2b74d6c7d6defd0f2f5256
-  data.tar.gz: 94c6a8178ead2a986921b72b07ef5dc388a5fa6a67945573eec921db30c9940d241d3f47591c8fbbf9bdcf313df0dda536f0fc78e0e946e27dfa3bc13dad9a28
+  metadata.gz: 9a71ee1e9307f0ef67c9dec108c7f68db45166a62f9b6ec60915ce2c089cf0e9ec5bfcd8d74e8b31b63238a09c820a0798689a84e5ea0b1577e2492e5a1d425e
+  data.tar.gz: b20043cead03f7fc7fe527fdbcb3674ab2d1da06b546bac9c1549b6eb6d143232453132709d93ae008d78a83bff36cf85fd0dbc0938da848e7847a1830e6011e

data/CHANGELOG.md

@@ -1,3 +1,18 @@
+## 1.0.1 - 2024-11-04
+
+### Changed
+
+* Informational output on errors when running CLI commands to provide more
+  details
+
+### Fixed
+
+* Parsing of indirect objects the value of which is an indirect reference
+* Writing of the initial cross-reference section to ensure a single subsection
+* [HexaPDF::Utils::SortedTreeNode] to wrap all /Kids entries with the correct
+  type class
+
+
 ## 1.0.0 - 2024-10-26
 
 ### Added

data/lib/hexapdf/cli.rb

@@ -64,8 +64,21 @@ module HexaPDF
     rescue StandardError => e
       $stderr.puts "Problem encountered: #{e.message}"
       unless e.kind_of?(HexaPDF::Error)
+        $stderr.puts "Backtrace (last 10 lines):"
+        $stderr.puts e.backtrace[0, 10]
+        $stderr.puts
         $stderr.puts "--> The problem might indicate a faulty PDF or a bug in HexaPDF."
-        $stderr.puts "--> Please report this at https://github.com/gettalong/hexapdf/issues - thanks!"
+        $stderr.puts "--> Please report this at"
+        $stderr.puts "-->"
+        $stderr.puts "-->     https://github.com/gettalong/hexapdf/issues"
+        $stderr.puts "-->"
+        $stderr.puts "--> and include the information above as well as the output of running"
+        $stderr.puts "--> the following command on the input PDF:"
+        $stderr.puts "-->"
+        $stderr.puts "-->     hexapdf info --check INPUT.PDF"
+        $stderr.puts "-->"
+        $stderr.puts "--> If possible, please also provide the input PDF."
+        $stderr.puts "--> Thanks!"
       end
       exit(1)
     end

data/lib/hexapdf/parser.rb

@@ -116,7 +116,11 @@ module HexaPDF
                         "the values (#{xref_entry.oid},#{xref_entry.gen}) from the xref")
       end
 
-      @document.wrap(obj, oid: oid, gen: gen, stream: stream)
+      if obj.kind_of?(Reference)
+        @document.deref(obj)
+      else
+        @document.wrap(obj, oid: oid, gen: gen, stream: stream)
+      end
     rescue HexaPDF::MalformedPDFError
       reconstructed_revision.object(xref_entry) ||
         @document.wrap(nil, oid: xref_entry.oid, gen: xref_entry.gen)

data/lib/hexapdf/utils/sorted_tree_node.rb

@@ -174,6 +174,7 @@ module HexaPDF
           elsif node.key?(:Kids)
             index = find_in_intermediate_node(node[:Kids], key)
             node = node[:Kids][index]
+            node = document.wrap(node, type: self.class) if node
             break unless node && key >= node[:Limits][0] && key <= node[:Limits][1]
           else
             break
@@ -194,7 +195,7 @@ module HexaPDF
         container_name = leaf_node_container_name
         stack = [self]
         until stack.empty?
-          node = stack.pop
+          node = document.wrap(stack.pop, type: self.class)
           if node.key?(container_name)
             data = node[container_name]
             index = 0
@@ -217,7 +218,7 @@ module HexaPDF
       def path_to_key(node, key, stack)
         return unless node.key?(:Kids)
         index = find_in_intermediate_node(node[:Kids], key)
-        stack << node[:Kids][index]
+        stack << document.wrap(node[:Kids][index], type: self.class)
         path_to_key(stack.last, key, stack)
       end
 
@@ -307,6 +308,15 @@ module HexaPDF
         super
         container_name = leaf_node_container_name
 
+        if key?(:Kids)
+          self[:Kids].each do |kid|
+            unless kid.indirect?
+              yield("Children of sorted tree nodes must be indirect", true)
+              document.add(kid)
+            end
+          end
+        end
+
         # All keys of the container must be lexically ordered strings and the container must be
         # correctly formatted
         if key?(container_name)

data/lib/hexapdf/version.rb

@@ -37,6 +37,6 @@
 module HexaPDF
 
   # The version of HexaPDF.
-  VERSION = '1.0.0'
+  VERSION = '1.0.1'
 
 end

data/lib/hexapdf/writer.rb

@@ -149,6 +149,7 @@ module HexaPDF
       obj_to_stm = object_streams.each_with_object({}) {|stm, m| m.update(stm.write_objects(rev)) }
 
       xref_section = XRefSection.new
+      xref_section.mark_as_initial_section! unless previous_xref_pos
       xref_section.add_free_entry(0, 65535) if previous_xref_pos.nil?
       rev.each do |obj|
         if obj.null?

data/lib/hexapdf/xref_section.rb

@@ -111,6 +111,13 @@ module HexaPDF
     # used.
     private :'[]='
 
+    # Marks this XRefSection object as being the first cross-reference section in a PDF file.
+    #
+    # This has the consequence that only a single sub-section is created.
+    def mark_as_initial_section!
+      @initial_section = true
+    end
+
     # Adds an in-use entry to the cross-reference section.
     #
     # See: ::in_use_entry
@@ -147,15 +154,24 @@ module HexaPDF
     # If this section contains no objects, a single empty array is yielded (corresponding to a
     # subsection with zero elements).
     #
-    # The subsections are dynamically generated based on the object numbers in this section.
+    # The subsections are dynamically generated based on the object numbers in this section. In case
+    # the section was marked as the initial section (see #mark_as_initial_section!) only a single
+    # subsection is yielded.
     def each_subsection
       return to_enum(__method__) unless block_given?
 
       temp = []
       oids.sort.each do |oid|
-        if !temp.empty? && temp[-1].oid + 1 != oid
-          yield(temp)
-          temp = []
+        expected_next_oid = !temp.empty? && temp[-1].oid + 1
+        if expected_next_oid && expected_next_oid != oid
+          if @initial_section
+            expected_next_oid.upto(oid - 1) do |free_oid|
+              temp << self.class.free_entry(free_oid, 0)
+            end
+          else
+            yield(temp)
+            temp = []
+          end
         end
         temp << self[oid]
       end

data/test/hexapdf/test_parser.rb

@@ -33,18 +33,23 @@ describe HexaPDF::Parser do
       endstream
       endobj
 
+      5 0 obj
+      1 0 R
+      endobj
+
       xref
       0 4
       0000000000 65535 f 
       0000000010 00000 n 
       0000000029 00000 n 
       0000000000 65535 f 
-      3 1
+      3 2
       0000000556 00000 n 
+      0000000308 00000 n
       trailer
       << /Test (now) >>
       startxref
-      308
+      330
       %%EOF
     EOF
   end
@@ -305,6 +310,11 @@ describe HexaPDF::Parser do
       assert_equal(0, obj.gen)
     end
 
+    it "handles the case of the value of an indirect object being an indirect reference" do
+      obj = @parser.load_object(HexaPDF::XRefSection.in_use_entry(5, 0, 308))
+      assert_equal(1, obj.oid)
+    end
+
     describe "with strict parsing" do
       it "raises an error if an indirect object has an offset of 0" do
         @document.config['parser.on_correctable_error'] = proc { true }
@@ -343,13 +353,13 @@ describe HexaPDF::Parser do
 
   describe "startxref_offset" do
     it "caches the offset value" do
-      assert_equal(308, @parser.startxref_offset)
-      @parser.instance_eval { @io }.string.sub!(/308\n/, "309\n")
-      assert_equal(308, @parser.startxref_offset)
+      assert_equal(330, @parser.startxref_offset)
+      @parser.instance_eval { @io }.string.sub!(/330\n/, "309\n")
+      assert_equal(330, @parser.startxref_offset)
     end
 
     it "returns the correct offset" do
-      assert_equal(308, @parser.startxref_offset)
+      assert_equal(330, @parser.startxref_offset)
     end
 
     it "ignores garbage at the end of the file" do

data/test/hexapdf/test_writer.rb

@@ -53,8 +53,8 @@ describe HexaPDF::Writer do
     EOF
 
     xref_stream = case HexaPDF::VERSION.length
-                  when 5 then "x\xDAcbdlc``b`\xB0\x04\x93\x93\x19\x18\x00\f\x0F\x01["
-                  when 6 then "x\xDAcbdlg``b`\xB0\x04\x93\x93\x18\x18\x00\f\e\x01["
+                  when 5 then "x\xDAcbdlg``b`\xB0\x04\x93\x93\x19\x18\x00\f\x1E\x01\\"
+                  when 6 then "x\xDAcbd\xEC```b`\xB0\x04\x93\x93\x18\x18\x00\f*\x01\\"
                   else fail
                   end
     @compressed_input_io = StringIO.new(<<~EOF.force_encoding(Encoding::BINARY))
@@ -69,8 +69,8 @@ describe HexaPDF::Writer do
       20
       endobj
       3 0 obj
-      <</Size 6/Type/XRef/W[1 1 2]/Index[0 4 5 1]/Filter/FlateDecode/DecodeParms<</Columns 4/Predictor 12>>/Length 31>>stream
-      x\xDAcb`\xF8\xFF\x9F\x89\x89\x95\x91\x91\xE9\x7F\x19\x03\x03\x13\x83\x10\x88he`\x00\x00B4\x04\x1E
+      <</Size 6/Type/XRef/W[1 1 2]/Index[0 6]/Filter/FlateDecode/DecodeParms<</Columns 4/Predictor 12>>/Length 36>>stream
+      x\xDAcb`\xF8\xFF\x9F\x89\x89\x95\x91\x91\xE9\x7F\x19\x03\x03\x13\x83\x10\x90\xF8_\f\x14c\x14bd\x04\x00lk\a 
       endstream
       endobj
       startxref
@@ -90,7 +90,7 @@ describe HexaPDF::Writer do
       endstream
       endobj
       startxref
-      #{442 + HexaPDF::VERSION.length}
+      #{443 + HexaPDF::VERSION.length}
       %%EOF
     EOF
   end

data/test/hexapdf/test_xref_section.rb

@@ -57,5 +57,20 @@ describe HexaPDF::XRefSection do
       @xref_section.add_in_use_entry(20, 0, 0)
       assert_subsections([[1, 2], [10, 11], [20]])
     end
+
+    it "yields a single subsection if the section was marked as the initial one" do
+      @xref_section.mark_as_initial_section!
+      @xref_section.add_in_use_entry(6, 0, 0)
+      @xref_section.add_in_use_entry(7, 0, 0)
+      @xref_section.add_in_use_entry(9, 0, 0)
+      @xref_section.add_in_use_entry(1, 0, 0)
+      @xref_section.add_in_use_entry(2, 0, 0)
+      result = @xref_section.each_subsection.map {|s| s.map {|e| [e.oid, e.type] }}
+      assert_equal([[[1, :in_use], [2, :in_use],
+                     [3, :free], [4, :free], [5, :free],
+                     [6, :in_use], [7, :in_use],
+                     [8, :free],
+                     [9, :in_use]]], result)
+    end
   end
 end

data/test/hexapdf/utils/test_sorted_tree_node.rb

@@ -12,10 +12,12 @@ describe HexaPDF::Utils::SortedTreeNode do
   end
 
   def add_multilevel_entries
-    @kid11 = @doc.add({Limits: ['c', 'f'], Names: ['c', 1, 'f', 1]}, type: HexaPDF::NameTreeNode)
+    item = @doc.add(1)
+    @item_ref = HexaPDF::Reference.new(item.oid, item.gen)
+    @kid11 = @doc.add({Limits: ['c', 'f'], Names: ['c', @item_ref, 'f', 1]}, type: HexaPDF::NameTreeNode)
     @kid12 = @doc.add({Limits: ['i', 'm'], Names: ['i', 1, 'm', 1]}, type: HexaPDF::NameTreeNode)
     ref = HexaPDF::Reference.new(@kid11.oid, @kid11.gen)
-    @kid1 = @doc.add({Limits: ['c', 'm'], Kids: [ref, @kid12]}, type: HexaPDF::NameTreeNode)
+    @kid1 = @doc.add({Limits: ['c', 'm'], Kids: [ref, @kid12]})
     @kid21 = @doc.add({Limits: ['o', 'q'], Names: ['o', 1, 'q', 1]}, type: HexaPDF::NameTreeNode)
     @kid221 = @doc.add({Limits: ['s', 'u'], Names: ['s', 1, 'u', 1]}, type: HexaPDF::NameTreeNode)
     @kid22 = @doc.add({Limits: ['s', 'u'], Kids: [@kid221]}, type: HexaPDF::NameTreeNode)
@@ -75,7 +77,7 @@ describe HexaPDF::Utils::SortedTreeNode do
       @root.add_entry('v', 1)
       assert_equal(['a', 'm'], @kid1[:Limits].value)
       assert_equal(['a', 'f'], @kid11[:Limits].value)
-      assert_equal(['a', 1, 'c', 1, 'e', 1, 'f', 1], @kid11[:Names].value)
+      assert_equal(['a', 1, 'c', @item_ref, 'e', 1, 'f', 1], @kid11[:Names].value)
       assert_equal(['g', 'm'], @kid12[:Limits].value)
       assert_equal(['g', 1, 'i', 1, 'j', 1, 'm', 1], @kid12[:Names].value)
       assert_equal(['n', 'v'], @kid2[:Limits].value)
@@ -203,13 +205,12 @@ describe HexaPDF::Utils::SortedTreeNode do
     end
 
     it "checks that all kid objects are indirect objects" do
-      @root[:Kids][0] = ref = HexaPDF::Reference.new(@kid1.oid, @kid1.gen)
       assert(@root.validate)
 
-      @root[:Kids][0] = ref
+      @root[:Kids][0] = @kid1
       @kid1.oid = 0
       assert(@root.validate do |message, c|
-               assert_match(/must be an indirect object/, message)
+               assert_match(/children.*must be indirect/i, message)
                assert(c)
              end)
       assert(@kid1.indirect?)

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: hexapdf
 version: !ruby/object:Gem::Version
-  version: 1.0.0
+  version: 1.0.1
 platform: ruby
 authors:
 - Thomas Leitner
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2024-10-26 00:00:00.000000000 Z
+date: 2024-11-04 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: cmdparse