[TAP5-2219] XML-Parsing broken if system encoding is not UTF-8

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 5.4
• Fix Version/s: 5.4
• Component/s: tapestry-core
• Environment: Windows (probably linux too but most systems run with utf-8)
• Assignee: Jochen Kemnade
• Reporter: Michael Wyraz
• Created: 07/Nov/13 19:08
• Updated: 22/Jun/14 22:01
• Resolved: 27/Apr/14 13:55

Description

Tapestry reads tml files in XMLTokenStream.openStream(), does some doctype "magic" and writes all the file content's into a ByteOutputStream. On T5.3 reading/writing was done using the system's caracter encoding. With T5.4 that was changed for the reader but not for the writer. This messes up all non-ascii characters if system encoding is not utf-8.

Solution: Change
PrintWriter writer = new PrintWriter(bos);
To
PrintWriter writer = new PrintWriter(new OutputStreamWriter(bos, "UTF8"));

in XMLTokenStream.openStream().

TODO: Unit test for XMLTokenStream (follows soon)

Issue Links

• duplicates: TAP5-1778 Template parsing dependent on JVM default charset
• is duplicated by: TAP5-1741 Parsing template which contains Chinese (Maybe other double bytes) words throws MalformedByteSequenceException: Invalid byte 3 of 3-byte UTF-8 sequence.
• is duplicated by: TAP5-1778 Template parsing dependent on JVM default charset
• is related to: TAP5-2293 Closure minimizer produces invalid output on non-UTF-8 systems
• relates to: TAP5-1741 Parsing template which contains Chinese (Maybe other double bytes) words throws MalformedByteSequenceException: Invalid byte 3 of 3-byte UTF-8 sequence.