Herbie 0.0.5

[blaylockbk]

New Name! HRRR-B 🠖 Herbie

I updated the GitHub repository name to Herbie and I'm slowly removing the old hrrrb API (but it's still there).

The most significant change is that the vision of Herbie has expanded. Herbie is being built do download many different model types, not just the HRRR model.

• Rename package to herbie. "Herbie is your model output download assistant with a mind of its own." Yes, this is named after a favorite childhood movie series.
• Implement new Herbie class
• Drop support for hrrrx (experimental HRRR no longer archived on Pando and ESRL is now developing RRFS)
• Added ability to download and read RAP model GRIB2 files.
• Less reliance on Pando, more on aws and google.
• New method for searchString index file search. Uses same regex search patterns as old API.
• Filename for GRIB2 subset includes all GRIB message numbers.
• Moved default download source to config file setting.
• Check local file copy on init. (Don't need to look for file on remote if we have local copy)
• Option to remove grib2 file when reading xarray if didn't already exist locally (don't clutter local disk).
• Attach index file DataFrame to object if it exists.
• If full file exists locally, use remote idx file to cURL local file instead of remote. (Can't create idx file locally because wgrib2 not available on windows)
• Added GFS data, though it isn't implemented as cleanly as HRRR or RAP
• Renamed 'field' argument to 'product'
• ✨ Moved the source URL templates to their own classes in the models folder
• Renamed GitHub repository to Herbie (changed from HRRR_archive_download)
• Added RRFS, NBM, GFS, RAP as models Herbie can download
• Reworked read_idx() to support index files with additional info (specifically for the NBM).

---

This discussion was created from the release Herbie 0.0.5.