[CAMEL-11723] ManagedCamelContext.dumpRestsAsXml can fail if default charset is not utf-8

• Type: Bug
• Status: Resolved
• Priority: Minor
• Resolution: Fixed
• Affects Version/s: 2.18.4
• Fix Version/s: 2.18.5, 2.19.3, 2.20.0
• Component/s: camel-core, jmx
• Labels: easyfix windows

Environment

Windows, default charset is not UTF-8

• Estimated Complexity: Novice
• Assignee: Claus Ibsen
• Reporter: Niels Ull Harremoes
• Created: 29/Aug/17 09:09
• Updated: 09/Sep/17 07:53
• Resolved: 07/Sep/17 15:02

Description

I am trying to access the rest definitions using the swagger servlet.
My operation description is in Danish and includes the character "æ".

Thus, the xml string generated in line 371 of org.apache.camel.management.mbean.ManagedCamelContext.dumpRestsAsXml
contains the character "æ".

In line 376-377, wh have

       InputStream is = new ByteArrayInputStream(xml.getBytes());
        Document dom = XmlLineNumberParser.parseXml(is, new ...)

The call to xml.getBytes() uses the default encoding (CP-1252 in my case), which will encode æ as a single byte. The xml parser expects UTF-8 (since the xml starts with an xml declaration specifying encoding="UTF-8").
Therefore, it fails with "Invalid byte 2 of 3-byte UTF-8 sequence".

The fix is simple - change xml.getBytes() to xml.getBytes("UTF-8");

Issue Links

• links to: GitHub Pull Request #1906