Remove unnecessary string decode in agri_l1 reader

[jactry]

This fixs a regression introduced by commit 14d9681.

• Passes flake8 satpy
• Add your name to AUTHORS.md if not there already

[codecov]

Codecov Report

Merging #1459 (4327226) into master (a18b6e2) will decrease coverage by 0.00%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master    #1459      +/-   ##
==========================================
- Coverage   90.52%   90.52%   -0.01%     
==========================================
  Files         238      238              
  Lines       34153    34149       -4     
==========================================
- Hits        30916    30912       -4     
  Misses       3237     3237              

Flag	Coverage Δ
behaviourtests	4.53% <ø> (+<0.01%)	⬆️
unittests	91.00% <ø> (-0.01%)	⬇️

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

Impacted Files	Coverage Δ
satpy/readers/agri_l1.py	99.15% <ø> (-0.03%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update a18b6e2...4327226. Read the comment docs.

[djhoese]

I believe @zxdawn recently ran into this issue as well. The main issue is newer versions of h5py loading/decoding strings differently. After investigating a bit with @zxdawn on slack, we weren't really sure why these strings were encoded with gbk anyway since they are all ASCII characters.

I'm OK with this PR, but am still curious why these strings ever needed to be decoded specifically with gbk in the first place.

[zxdawn]

@djhoese I can test different versions of h5py and report here later.

[mraspaud]

@zxdawn any progress here?

[zxdawn]

@mraspaud It's not related to the version of h5py.

I have tested satpy==0.23.0, h5py==2.10.0 and satpy==0.23.0, h5py==3.1.0, they both works well without decoding:

'band_names': 'band12(band number is range from 1 to 14)'
'long_name': '10.8um channel 4km image data layer'

Then, if I test satpy==0.21.0, h5py==2.10.0, these attrs are bytes without decoding:

'band_names': b'band12(band number is range from 1 to 14)'
'long_name': b'10.8um channel 4km image data layer'

Anyway, LGTM for the newest satpy.

[djhoese]

Thanks. Looks good.