[Post-v1.0] Generic file type to Neurodata Type mapper

[CodyCBakerPhD]

Leaving this one for @rly

NeuroConv really only supports structured proprietary formats, things such as a binary file with accompanying header where the header is specific to an acquisition system or standardized format, or a mostly-binary file with a header block, or HDF5 or MATLAB files that follow a particular and consistent structure over time, etc...

But we often see some people require conversion of data that they know how it's structured, but it might just be a very generic .pkl object, .npy array, or a headerless binary blob

Each of these file formats is fairly simple to open and read from in Python (some nuances for .pkl I might mention in the future...), but the challenge we face in NWB conversion is correct assignment of the neurodata type along with any metadata about that type

This would be easiest/best done in the GUIDE where we can walk the user through the process step by step and ask particular questions along the way

Hard to say how this would fit into the UX for the guided mode vs. free form - will leave that for @rly to think on and propose (maybe Figma it first?)

[bendichter]

Good idea, @CodyCBakerPhD. Maybe we could start by essentially walking through the logic of the modality tutorials. A few weeks back, Satra suggested we create a workflow diagram for how to convert data. It seems like that might be particularly useful here.

[CodyCBakerPhD]

Summarized and moved to #847

Re-open again whenever this starts active development