[PDFBOX-503] PDF loader causes infinite loop on non-PDF inputs

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 0.8.0-incubator
• Fix Version/s: 0.8.0-incubator
• Component/s: Parsing
• Assignee: Unassigned
• Reporter: Dave Engberg
• Created: 12/Aug/09 23:51
• Updated: 28/Apr/10 22:16
• Resolved: 01/Sep/09 17:21

Description

The current SVN head for the pdfbox incubator will experience an infinite loop in PDFParser.parseHeader() if you feed any non-PDF document to the parser. The problem is that it tries to find the PDF header within the document by skipping over any non-matching lines which don't start with a numeric digit. It relies on a readLine() function from BaseParser.java which will return an empty string when the stream is at the end-of-file. The parseHeader() call will loop on these empty lines.

I've patched this in our system by throwing an IOException from BaseParser.readLine() if the stream is already at the end-of-file at the beginning of that call.

Index: src/main/java/org/apache/pdfbox/pdfparser/BaseParser.java
===================================================================
— src/main/java/org/apache/pdfbox/pdfparser/BaseParser.java (revision 802578)
+++ src/main/java/org/apache/pdfbox/pdfparser/BaseParser.java (working copy)
@@ -1088,6 +1088,11 @@
{
StringBuffer buffer = new StringBuffer( 11 );

+ if (pdfSource.isEOF())
+

{ + throw new IOException( "Error: End-of-File, expected line"); + }

+
int c;
while ((c = pdfSource.read()) != -1)
{