hexapdf 0.19.2 → 0.19.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: cb1a2ebe9b89b1c97bcfe6a9227bbdc9b8627458900347ac0965784f4d95f83f
-  data.tar.gz: 294eeaa8e9d9926debf2d28d83c8cd419f5685a88b9d23c378fa783fce2d62cb
+  metadata.gz: 85c063a63af9729acc10a54ef53fba69d73b75f1c06ff0e6de246df920e17dd9
+  data.tar.gz: dcea10d0ccfe66282c92e6bb41c1d57927d525e1618753d598a910546c8a8e82
 SHA512:
-  metadata.gz: e2284d3397a856d895b8ec2c83b3ead95ba11a571df90f9fdaed7cd83e4880004ea04150df9205a534d203ae5a1958977a7755ad9f06d622c4f38ae43c0823ab
-  data.tar.gz: 2d6b454f78f7b319240fc062429c32f87ed350ab64a596e7b5050abe65b4e8b4bd9c3551cb87a699219fbd484b5c7670c9662674d9aa5dc8ef800dbe70a273c2
+  metadata.gz: 4f1d468375c4ce336e55a09d897de9a8c95babcfa1b4441cc04a9dc712435c64906016647baa030f5695f9434d18214c30bca746f245b53a69321139f63256b9
+  data.tar.gz: 1f1895ca7ad46bae1bf33790d4c8daa7f72adfcc00cc26e1611b7ece64d7a5a76e7f1e06be4022bb58f5dbd84154e9846cb4feb94f727c9abe99f5b7c8b87fcf

data/CHANGELOG.md

@@ -1,3 +1,11 @@
+## 0.19.3 - 2021-12-14
+
+### Fixed
+
+* Handling of invalid files where the "startxref" keyword and its value are on
+  the same line
+
+
 ## 0.19.2 - 2021-12-14
 
 ### Fixed

data/lib/hexapdf/parser.rb

@@ -342,7 +342,8 @@ module HexaPDF
       step_size = 1024
       pos = @io.pos
       eof_not_found = pos == 0
-      startxref_missing = false
+      startxref_missing = startxref_mangled = false
+      startxref_offset = nil
 
       while pos != 0
         @io.pos = [pos - step_size, 0].max
@@ -350,27 +351,31 @@ module HexaPDF
         lines = @io.read(step_size + 40).split(/[\r\n]+/)
 
         eof_index = lines.rindex {|l| l.strip == '%%EOF' }
-        unless eof_index
+        if !eof_index
           eof_not_found = true
-          next
-        end
-        unless eof_index >= 2 && lines[eof_index - 2].strip == "startxref"
+        elsif lines[eof_index - 1].strip =~ /\Astartxref\s(\d+)\z/
+          startxref_offset = $1.to_i
+          startxref_mangled = true
+          break # we found it even if it the syntax is not entirely correct
+        elsif eof_index < 2 || lines[eof_index - 2].strip != "startxref"
           startxref_missing = true
-          next
+        else
+          startxref_offset = lines[eof_index - 1].to_i
+          break # we found it
         end
-
-        break # we found the startxref offset
       end
 
       if eof_not_found
         maybe_raise("PDF file trailer with end-of-file marker not found", pos: pos,
                     force: !eof_index)
+      elsif startxref_mangled
+        maybe_raise("PDF file trailer keyword startxref on same line as value", pos: pos)
       elsif startxref_missing
         maybe_raise("PDF file trailer is missing startxref keyword", pos: pos,
                     force: eof_index < 2 || lines[eof_index - 2].strip != "startxref")
       end
 
-      @startxref_offset = lines[eof_index - 1].to_i
+      @startxref_offset = startxref_offset
     end
 
     # Returns the reconstructed revision.

data/lib/hexapdf/version.rb

@@ -37,6 +37,6 @@
 module HexaPDF
 
   # The version of HexaPDF.
-  VERSION = '0.19.2'
+  VERSION = '0.19.3'
 
 end

data/test/hexapdf/test_parser.rb

@@ -338,6 +338,11 @@ describe HexaPDF::Parser do
       assert_equal(5, @parser.startxref_offset)
     end
 
+    it "handles the case of startxref and its value being on the same line" do
+      create_parser("startxref 5\n%%EOF")
+      assert_equal(5, @parser.startxref_offset)
+    end
+
     it "fails even in big files when nothing is found" do
       create_parser("\nhallo" * 5000)
       exp = assert_raises(HexaPDF::MalformedPDFError) { @parser.startxref_offset }
@@ -366,6 +371,13 @@ describe HexaPDF::Parser do
       exp = assert_raises(HexaPDF::MalformedPDFError) { @parser.startxref_offset }
       assert_match(/end-of-file marker not found/, exp.message)
     end
+
+    it "fails on strict parsing if the startxref is on the same line as its value" do
+      @document.config['parser.on_correctable_error'] = proc { true }
+      create_parser("startxref 5\n%%EOF")
+      exp = assert_raises(HexaPDF::MalformedPDFError) { @parser.startxref_offset }
+      assert_match(/startxref on same line/, exp.message)
+    end
   end
 
   describe "file_header_version" do

data/test/hexapdf/test_writer.rb

@@ -40,7 +40,7 @@ describe HexaPDF::Writer do
       219
       %%EOF
       3 0 obj
-      <</Producer(HexaPDF version 0.19.2)>>
+      <</Producer(HexaPDF version 0.19.3)>>
       endobj
       xref
       3 1
@@ -72,7 +72,7 @@ describe HexaPDF::Writer do
       141
       %%EOF
       6 0 obj
-      <</Producer(HexaPDF version 0.19.2)>>
+      <</Producer(HexaPDF version 0.19.3)>>
       endobj
       2 0 obj
       <</Length 10>>stream

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: hexapdf
 version: !ruby/object:Gem::Version
-  version: 0.19.2
+  version: 0.19.3
 platform: ruby
 authors:
 - Thomas Leitner
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2021-12-13 00:00:00.000000000 Z
+date: 2021-12-14 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: cmdparse