[XERCESJ-1094] Xerces in infinite loop validating wrongly encoded XML 1.1 documents

• Type: Bug
• Status: Resolved
• Priority: Blocker
• Resolution: Fixed
• Affects Version/s: 2.7.1
• Fix Version/s: 2.8.0
• Environment: Linux, Solaris, jdk 1.2.2/1.4/1.5
• Assignee: Michael Glavassevich
• Reporter: Yves Lafon
• Created: 09/Aug/05 23:52
• Updated: 27/Feb/06 14:30
• Resolved: 10/Aug/05 02:20

Description

When parsing a XML1.1 document from an InputSource, where the encoding is set to iso-8859-1, with an encoding set to UTF-8 in the XML declaration, and with a iso-8859-2 character in an attribute, then xerces enters an infinite loop.
If the same character is not in the attribute, then Xerces reports an invalid XML character instead of blocking.
If the encoding of the input source is not set to iso-8859-1, Xerces works fine also.

Sample doc and modified DocumentScanner that demonstrate the issue at http://jigsaw.w3.org/Yves/xercesBug.zip
Thanks,