| WORKAROUND for me has been to archive the XML in addition to publishing it. Not ideal, but it has worked well. Also, In pipeline jobs, I have fancier logic to make this less ugly. It does the following:
- Publishes XML via the JUnit plugin (normally), ie:
junit allowEmptyResults: true, testResults: '/TEST-.xml'*
- Deletes any successful junit XML (I have a pipeline library method to scan files for failures="0" and errors="0", omitting code for brevity). This is optional and is done to conserve disk space.
- Zips and archives the original glob which is now only failure files:
zip archive: true, dir: '', glob: '/TEST-.xml', zipFile: junit-failures-xml.zip*
Hope this helps. |