Delete all line breaks, add one each <, get the data you want per line and you’re good.

Of course, only for targeted data extraction. …which should be the default in any parser, really.