hexapdf 0.17.2 → 0.17.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 71e5d38f8a3d1a6b4371e5f8e2f991d36c0ea8c01ca5dc4ec1e132043fad00aa
-  data.tar.gz: cf78ef5b6fa915bac30fe75157fc414148c8bd2b8c5c7ee10fe4bb3cbf70076a
+  metadata.gz: 15efadb34e0a5dc93ec7659d02cb95a19400a38fe7ccffcbc2438fa6e6f51b8f
+  data.tar.gz: 29ddd43180432e648c107ae40c2618415f0ead8bf6e6fcbe46bc1f0557eae271
 SHA512:
-  metadata.gz: 2153a1c0eadc0ca940480502b42780717e2777416f1e85e1c9506b879e01df0b2bc4e6d838ecd140af4a62e93bd16aa08ab42bbf522f6bbd234c0d9da82e4115
-  data.tar.gz: e91b860ece2751633350739852b392719d9d09cb1bdd9e1c1896ef76e5828db9d2d0f9204b93521efbd6e5dcefc84b0451984392c2a4de3703a6060def1f5c75
+  metadata.gz: 0a3668dd377838665e75b6c8325efbe237e379823deb060acdf5c7be26d799ab4eee65e94a75b50e9ffc27c81693cb0158765402d9d6f4ddbacc3c1b3be531d4
+  data.tar.gz: 759a4a692997d50022f26fd6e5fbf06859a93a00b5d371286a084ffbcb98d3571d72eeb1d3c8d733b4b8917d195c236f7dac3654e4fddae8cb9c2cf7b2f9ab42

data/CHANGELOG.md

@@ -1,3 +1,12 @@
+## 0.17.3 - 2021-10-31
+
+### Fixed
+
+* Reconstruction of invalid PDF files where the PDF header is not at the start
+  of the file
+* Reconstruction of invalid PDF files where the first object is invalid
+
+
 ## 0.17.2 - 2021-10-26
 
 ### Fixed

data/lib/hexapdf/parser.rb

@@ -404,6 +404,7 @@ module HexaPDF
     def reconstruct_revision
       return if @in_reconstruct_revision
       @in_reconstruct_revision = true
+      @header_offset = 0
 
       raise unless @document.config['parser.try_xref_reconstruction']
       msg = "#{$!} - trying cross-reference table reconstruction"
@@ -428,8 +429,10 @@ module HexaPDF
           elsif gen.kind_of?(Integer) && tok.kind_of?(Tokenizer::Token) && tok == 'obj'
             xref.add_in_use_entry(token, gen, pos)
             if linearized.nil?
+              pos = @tokenizer.pos
               obj = @tokenizer.next_object rescue nil
               linearized = obj.kind_of?(Hash) && obj.key?(:Linearized)
+              @tokenizer.pos = pos
             end
             @tokenizer.scan_until(/(?:\n|\r\n?)endobj\b/)
           end

data/lib/hexapdf/version.rb

@@ -37,6 +37,6 @@
 module HexaPDF
 
   # The version of HexaPDF.
-  VERSION = '0.17.2'
+  VERSION = '0.17.3'
 
 end

data/test/hexapdf/test_dictionary_fields.rb

@@ -135,12 +135,12 @@ describe HexaPDF::DictionaryFields do
     end
 
     def configuration
-      {'document.on_invalid_string' => proc {|str| str }}
+      HexaPDF::Configuration.with_defaults
     end
 
     it "calls document.on_invalid_string if the provided string is invalid" do
       str = "\xfe\xff\xD8\x00\x00s\x00t".b
-      assert_equal("\xD8\x00\x00s\x00t".force_encoding("UTF-16BE"), @field.convert(str, self))
+      assert_equal("st", @field.convert(str, self))
     end
   end
 

data/test/hexapdf/test_parser.rb

@@ -607,11 +607,22 @@ describe HexaPDF::Parser do
       assert_equal(4, @parser.load_object(@xref).value)
     end
 
+    it "handles an invalid object as first object" do
+      create_parser("2 0 obj\n(a(b\nendobj\n1 0 obj\n6\nendobj #)(\ntrailer\n<</Size 1>>")
+      assert_equal(6, @parser.load_object(@xref).value)
+    end
+
     it "ignores invalid lines" do
       create_parser("1 0 obj\n5\nendobj\nhello there\n1 0 obj\n6\nendobj\ntrailer\n<</Size 1>>")
       assert_equal(6, @parser.load_object(@xref).value)
     end
 
+    it "resets the header offset" do
+      create_parser("1 0 obj\n5\nendobj\ntrailer\n<</Size 1>>")
+      @parser.instance_variable_set(:@header_offset, 5)
+      assert_equal(5, @parser.load_object(@xref).value)
+    end
+
     it "uses the last trailer" do
       create_parser("trailer <</Size 1>>\ntrailer <</Size 2/Prev 342>>")
       assert_equal({Size: 2}, @parser.reconstructed_revision.trailer.value)

data/test/hexapdf/test_writer.rb

@@ -40,7 +40,7 @@ describe HexaPDF::Writer do
       219
       %%EOF
       3 0 obj
-      <</Producer(HexaPDF version 0.17.2)>>
+      <</Producer(HexaPDF version 0.17.3)>>
       endobj
       xref
       3 1
@@ -72,7 +72,7 @@ describe HexaPDF::Writer do
       141
       %%EOF
       6 0 obj
-      <</Producer(HexaPDF version 0.17.2)>>
+      <</Producer(HexaPDF version 0.17.3)>>
       endobj
       2 0 obj
       <</Length 10>>stream

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: hexapdf
 version: !ruby/object:Gem::Version
-  version: 0.17.2
+  version: 0.17.3
 platform: ruby
 authors:
 - Thomas Leitner
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2021-10-26 00:00:00.000000000 Z
+date: 2021-10-31 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: cmdparse