(hotfix): avoid error with large totim in returning budget dataframes

[mnfienen]

When totim is very large (like hundreds of years) the get_dataframes method for mflistfile fails when trying to convert totime to a pandas.datetime object. This fix just uses a try-except to try converting to datetime and, if fails, just leaves totim as a float. @jlarsen-usgs - are there other methods that return dataframes with a time index? I suggest this fix could apply for them as well.

[codecov]

Codecov Report

Merging #1677 (99109c1) into develop (d99e0cc) will increase coverage by 0.2%.
The diff coverage is 77.7%.

@@            Coverage Diff            @@
##           develop   #1677     +/-   ##
=========================================
+ Coverage     71.6%   71.8%   +0.2%     
=========================================
  Files          253     253             
  Lines        55881   55885      +4     
=========================================
+ Hits         40053   40178    +125     
+ Misses       15828   15707    -121     

Impacted Files	Coverage Δ
flopy/utils/mflistfile.py	70.7% <50.0%> (-0.3%)	⬇️
flopy/modpath/mp7.py	85.0% <100.0%> (+<0.1%)	⬆️
flopy/utils/gridutil.py	98.4% <0.0%> (+0.7%)	⬆️
flopy/utils/binaryfile.py	80.1% <0.0%> (+2.8%)	⬆️
flopy/utils/compare.py	30.4% <0.0%> (+12.7%)	⬆️

[jdhughes-dev]

You just need to run isort and black on this PR.