Add support for open ephys version 0.6

[JuliaSprenger]

Have you had exposure to open ephys version 0.6 yet? There appear to be some new file naming conventions, and interesting new corner cases, e.g. "CH{n}" name collisions! I've had to patch things over by parsing the "Continuous_Data.openephys" xml file, and then shimming that output to get consistent outcomes. I do hope there's an easier way that occurs to you!

Originally posted by @miketrumpis in https://github.com/NeuralEnsemble/python-neo/issues/1055#issuecomment-1063241131

[samuelgarcia]

@miketrumpis : do you have a dataset for dealing this problem ?

[miketrumpis]

Apologies for the lack of transparency here, as this is not public data. I've shared a description of the problem in the fork I made. My intention was to PR, but I had to cut some corners for my particular use cases, which sacrificed some generalizability. Here is a description of the problem, highlighting the new path layout and variations on the XML structure that I'm using to decode the data streams.

Readme file here: https://github.com/miketrumpis/python-neo/tree/oephys-hacks/neo/rawio

[JuliaSprenger]

Hi @miketrumpis, thanks for pointing to your custom fixes. However for a general fix it is essential to have a test dataset as reference. Do you know any public dataset in the new format or will you generate one in the future?

[zm711]

I think we fixed this recently, but if you test the current version and it is still not working please open a new issue with additional info (we now have a test file for new version and it seems to work so if it doesn't work for you we might need to tweak it a bit more).