[WIP] Add MIRS file reader

[katherinekolman]

This adds a reader for files from MIRS.

• Tests added
• Tests passed
• Passes flake8 satpy
• Fully documented
• Add your name to AUTHORS.md if not there already

[ghost]

DeepCode's analysis on #b990f9 found:

• ⚠️ 1 warning 👇

Top issues

Description	Example fixes
Missing close for open, add close or use a with block. Occurrences: mirs.py:72	🔧 Example fixes

👉 View analysis in DeepCode’s Dashboard | Configure the bot

[coveralls]

Coverage decreased (-0.3%) to 89.794% when pulling b990f9a on katherinekolman:mirs_reader into f80c568 on pytroll:master.

[codecov]

Codecov Report

Merging #1285 into master will decrease coverage by 0.28%.
The diff coverage is 18.89%.

@@            Coverage Diff             @@
##           master    #1285      +/-   ##
==========================================
- Coverage   90.07%   89.79%   -0.29%     
==========================================
  Files         218      219       +1     
  Lines       31580    31707     +127     
==========================================
+ Hits        28447    28471      +24     
- Misses       3133     3236     +103     

Impacted Files	Coverage Δ
satpy/readers/mirs.py	18.89% <18.89%> (ø)

---

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 f80c568...b990f9a. Read the comment docs.

[joleenf]

@djhoese In this commit, the file metadata is taking precedence over metadata in the yaml for the metop file used for testing. (IMG_SX.M2.D17037.S1601.E1607.B0000001.WE.HR.ORB.nc). Therefore, standard_name is added when reading the file metadata. I thought this line in the available_datasets meant that it is possible to append new information to the metadata:

This method is not guaranteed that it will be called before any other file type’s handler. The availability “boolean” not being None does not mean that a file handler called later can’t provide an additional dataset, but it must provide more identifying (DataID) information to do so and should yield its new dataset in addition to the previous one.

Can the boolean be True, False or None? Or is it only either True or None? I think the available_dataset() information above is saying, "Even if the availability "boolean" is True rather than None, a file handler could provide additional datasets." After that line, what is the more identifying DataID? Is this referring to the more definitive DataIDs like Brightness Temperature data defined by frequency and polarization "btemp_23v" versus the wider scope of BT? If so, then that is working correctly in this reader, because both the BT and the definitive brightness temperature channels are being yielded. Is there a way to keep variable metadata from a yaml and use metadata from a file, or is it either/or?
Also, in "Migrating to Xarray" the example shows using the "coord" attribute. Is assign_coords the only concern in this issue report?

[joleenf]

@djhoese Previous comment was supposed to be under #1511

[djhoese]

Replaced by #1511