allow data_dir option in PREP section to be a list of directories

[NickleDave]

instead of forcing user to e.g. write a bash script with a bunch of commands + separate .toml files

[NickleDave]

instead of changing / adding option(s), just make it so the data_dir option can be either a string, i.e. a single directory path, or a list of strings, i.e. multiple paths to directories

this would mean that functions in io.dataframe would need to "know" that data_dir can be a list

I think it should happen with in the dataframe.from_files function itself and not any "lower" functions, that get called by from_files

e.g. here from_files should just loop over data_dir (if it is a list) and concatenate the results into a single (flattened) list, instead of audio.files_from_dir needing to handle both strings and lists for data_dir https://github.com/NickleDave/vak/blob/78cc5cb764302e6fc927e206ccaf2eac3e6eff5b/src/vak/io/dataframe.py#L132

but later on when data_dir gets passed to spect.to_dataframe I think that spect might actually need to handle both a str and lists, since it only gets called once https://github.com/NickleDave/vak/blob/78cc5cb764302e6fc927e206ccaf2eac3e6eff5b/src/vak/io/dataframe.py#L183

@yardencsGitHub please let me know if you have opinions on this planning to add this weekend to make it easier to finish up the behavioral expt part

[NickleDave]

Need to also:

• [ ] make it so PrepConfig.data_dir can be a single directory or a list of directories
  • create a validator just for data_dir that validates this
• [ ] also make it so that the core.prep parameter data_dir can be a single directory or a list of directories
  • need to do the same validation we did for the PrepConfig attribute -- trying to maintain ability to call core functions outside of the cli