We have sizable number of large junitResult.xml files, 500GB in total. These are rarely read, and contain lots of repeating output. Compressing the files with gzip yields a large disk space saving:

> du -hs *
141M junitResult.xml
6.6M junitResult.xml.gz

I propose to add gz support to hudson.XmlFile to read these gzipped files. The support will be read only, the files can be compressed via scripts / other means.

A pull request will be opened shortly after this ticket is submitted.