change from using fparse to using parse

[bendichter]

Over the last week or so I worked with the new dev of the parse library to incorporate the datetime parsing into that library: r1chardj0n3s/parse#165. This was the only feature that motivated the splinter of fparse, so we can now move our dependence back to parse with version >=1.20.0 and archive fparse.

There is one small difference in the parse implementation, which is that it infers whether the parsed value should be a datetime, a date, or a time. If no time information is present, parse returns a date object. NeuroConv expects a datetime, and I think PyNWB requires a datetime for the session_start_time, so for backwards compatibility and for better compatibility with PyNWB, I included a line that transforms dates to datetimes when extracting metadata.

I'd really like NWB to accept just a date value there, since the actual time is often missing and must be filled in with incorrect data. If NeurodataWithoutBorders/nwb-schema#542 ever gets merged we can switch this back to returning dates when only date information is given. However I don't want that issue to hold up this one so I'd like to move forward with datetimes for now.

[codecov]

Codecov Report

Merging #654 (e91a579) into main (1d3d58d) will increase coverage by 0.00%.
The diff coverage is 100.00%.

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #654   +/-   ##
=======================================
  Coverage   91.61%   91.61%           
=======================================
  Files         107      107           
  Lines        5627     5630    +3     
=======================================
+ Hits         5155     5158    +3     
  Misses        472      472           

Flag	Coverage Δ
unittests	91.61% <100.00%> (+<0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files	Coverage Δ
src/neuroconv/tools/path_expansion.py	97.56% <100.00%> (+0.19%)	⬆️