I have a directory full of xml (xbrl) files to process, over 5,700. I imagine that I will have to process them in a BPO, get a directory count and grab them one at a time. Has anyone done something like this, to confirm that I'm heading in the best direction. I imagine that I'll open each file, parse it and navigate to the correct nodes and access the relevant data, close it, process the data and then repeat.

Any suggestions welcome.