[logback-dev] [JIRA] Resolved: (LBCORE-134) Joran XML Parser sometimes misses white space characters from element bodies

[Ceki Gulcu (JIRA)]

     [ http://jira.qos.ch/browse/LBCORE-134?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Ceki Gulcu resolved LBCORE-134.
-------------------------------

    Fix Version/s: 0.9.19
       Resolution: Fixed

Hello Michael,

Thank  you for reporting this bug, and in particular for identifying its origin. It was fixed in a recent commit:

http://github.com/ceki/logback/commit/4e5aba2f2489bd7d2a1f8284202a8cf7d3373a22


> Joran XML Parser sometimes misses white space characters from element bodies
> ----------------------------------------------------------------------------
>
>                 Key: LBCORE-134
>                 URL: http://jira.qos.ch/browse/LBCORE-134
>             Project: logback-core
>          Issue Type: Bug
>          Components: Joran
>    Affects Versions: 0.9.18
>            Reporter: Michael Franz
>            Assignee: Ceki Gulcu
>            Priority: Critical
>             Fix For: 0.9.19
>
>
> I have the following example configuration:
> 	<property name="Layout1" value="[Layout1] &quot;xyz&quot;"/>
> 	<appender name="LayoutTest_Console" class="ch.qos.logback.core.ConsoleAppender">
> 		<filter class="ch.qos.logback.classic.filter.ThresholdFilter">
> 			<level>TRACE</level>
> 		</filter>		
> 		<layout class="ch.qos.logback.classic.PatternLayout">
> 			<Pattern>
> 				[%date{yyyy-MM-dd HH:mm:ss.SSS}] ${Layout1} [Layout2] &quot;xyz&quot;%n
> 			</Pattern>
> 		</layout>
> 	</appender>
> and this produces the following output:
> [2010-01-05 13:06:34.897] [Layout1] "xyz" [Layout2]"xyz"
> Notice the missing space between [Layout2] and "xyz" in the output that was present in the configuration.
> The investigation shows a bug in ch.qos.logback.core.joran.event.SaxEventRecorder:
>   public void characters(char[] ch, int start, int length) {
>     String body = new String(ch, start, length);
>     if (body == null) {
>       return;
>     }
>     // if the body string is null
>     if (body != null) {
>       String bodyTrimmed = body.trim();
>       if (bodyTrimmed.length() == 0) {
>         return;
>       }
>     }
>     SaxEvent lastEvent = getLastEvent();
>     if (lastEvent instanceof BodyEvent) {
>       BodyEvent be = (BodyEvent) lastEvent;
>       be.append(body);
>     } else {
>       saxEventList.add(new BodyEvent(body, getLocator()));
>     }
>   }
> In my case the following calls were made to this method:
> 1. "[%date{yyyy-MM-dd HH:mm:ss.SSS}] ${Layout1} [Layout2]"
> 2. " "
> 3. """
> 4. "xyz"
> 5. """
> 6. "%n".
> Note that the calls actually performed depend on the actual body content within the file and the actually used SAX parser (in my case com.sun.org.apache.xerces.internal.jaxp.SAXParserImpl$JAXPSAXParser).
> So within the method the:
>     // if the body string is null
>     if (body != null) {
>       String bodyTrimmed = body.trim();
>       if (bodyTrimmed.length() == 0) {
>         return;
>       }
>     }
> must not occur if the last event was already a BodyEvent.

-- 
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://jira.qos.ch/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira