parser.rb in combine_pdf-0.2.16

- old
+ new

@@ -97,14 +97,14 @@
 
 			## search for objects streams
 			object_streams = @parsed.select {|obj| obj.is_a?(Hash) && obj[:Type] == :ObjStm}
 			unless object_streams.empty?
 				warn "PDF 1.5 Object streams found - they are not fully supported! attempting to extract objects."
-				
+
 				object_streams.each do |o|
 					## un-encode (using the correct filter) the object streams
-					PDFFilter.inflate_object o 
+					PDFFilter.inflate_object o
 					## extract objects from stream to top level arry @parsed
 					@scanner = StringScanner.new o[:raw_stream_content]
 					stream_data = _parse_
 					id_array = []
 					while stream_data[0].is_a? Fixnum
@@ -121,11 +121,10 @@
 			end
 
 			# Strings were unified, we can let them go..
 			@strings_dictionary.clear
 
-			
 			# serialize_objects_and_references.catalog_pages
 
 			# Benchmark.bm do |bm|
 			# 	bm.report("serialize") {1000.times {serialize_objects_and_references} }
 			# 	bm.report("serialize - old") {1000.times {old_serialize_objects_and_references} }
@@ -314,11 +313,11 @@
 				##########################################
 				when str = @scanner.scan(/\%/)
 					#is a comment, skip until new line
 					loop do
 						# break unless @scanner.scan(/[^\d\r\n]+/)
-						break if @scanner.check(/([\d]+[\s]+[\d]+[\s]+obj[\n\r\s]+\<\<)|([\n\r]+)/) || @scanner.eos? # || @scanner.scan(/[^\d]+[\r\n]+/) || 
+						break if @scanner.check(/([\d]+[\s]+[\d]+[\s]+obj[\n\r\s]+\<\<)|([\n\r]+)/) || @scanner.eos? # || @scanner.scan(/[^\d]+[\r\n]+/) ||
 						@scanner.scan(/[^\d\r\n]+/) || @scanner.pos += 1
 					end
 					# puts "AFTER COMMENT: #{@scanner.peek 8}"
 				##########################################
 				## Parse a Name
@@ -362,26 +361,31 @@
 					fresh = true
 					if @scanner.matched[-1] == 'r'
 						if @scanner.skip_until(/<</)
 							data = _parse_
 							@root_object ||= {}
-							@root_object[data.shift] = data.shift while data[0]						
+							@root_object[data.shift] = data.shift while data[0]
 						end
 						##########
 						## skip untill end of segment, maked by %%EOF
 						@scanner.skip_until(/\%\%EOF/)
+						##########
+						## If this was the last valid segment, ignore any trailing garbage
+						## (issue #49 resolution)
+						break unless @scanner.exist?(/\%\%EOF/)
+
 					end
-					
+
 				when @scanner.scan(/[\s]+/)
 					# Generally, do nothing
 					nil
 				when @scanner.scan(/obj[\s]*/)
 					# Fix wkhtmltopdf PDF authoring issue - missing 'endobj' keywords
 					unless fresh || (out[-4].nil? || out[-4].is_a?(Hash))
 						keep = []
-						keep << out.pop # .tap {|i| puts "#{i} is an ID"} 
-						keep << out.pop # .tap {|i| puts "#{i} is a REF"} 
+						keep << out.pop # .tap {|i| puts "#{i} is an ID"}
+						keep << out.pop # .tap {|i| puts "#{i} is a REF"}
 
 						if out.last.is_a? Hash
 							out << out.pop.merge({indirect_generation_number: out.pop, indirect_reference_id: out.pop})
 						else
 							out << {indirect_without_dictionary: out.pop, indirect_generation_number: out.pop, indirect_reference_id: out.pop}
@@ -391,11 +395,11 @@
 						out << keep.pop
 						out << keep.pop
 					end
 					fresh = false
 				else
-					# always advance 
+					# always advance
 					# warn "Advnacing for unknown reason... #{@scanner.peek(4)}" unless @scanner.peek(1) =~ /[\s\n]/
 					warn "Warning: parser advnacing for unknown reason. Potential data-loss."
 					@scanner.pos = @scanner.pos + 1
 				end
 			end
@@ -418,11 +422,11 @@
 				@parsed.delete_if {|obj| obj[:Type] == :Catalog}
 				@parsed << catalogs
 
 				raise "Unknown error - parsed data doesn't contain a cataloged object!" unless catalogs
 			end
-			case 
+			case
 			when catalogs.is_a?(Array)
 				catalogs.each {|c| catalog_pages(c, inheritance_hash ) unless c.nil?}
 			when catalogs.is_a?(Hash)
 				if catalogs[:is_reference_only]
 					if catalogs[:referenced_object]
@@ -576,6 +580,6 @@
 		# 		end
 		# 	end
 		# end
 
 	end
-end
\ No newline at end of file
+end