https://github.com/processing/processing/issues/1013
In short, change your way of parsing XML...

Thanks for your reply.

Hello,

Looks like you haven't understood the comments of the issue.
loadXML() is fine, and parseXML() won't change the behavior with regard to whitespace.
The idea is not to try and process each and every node of the XML file, but to search and get only those you are interested in.
Use getChildren() with a name as parameter to get the tags with this name.

Thanks. I got it.

But for us, we do not want to hardcode the name of the elements into the code. We just parse the xml file and create our tree structure in our code.

I just detect the space with "#text" name and go to the next child.