464: allow for 0x node prefix values

[migillett]

A potential answer to issue 464. This allows users to run commands using a different prefix (0x) from the meshtastic CLI.

The code change here checks for if the node ID supplied can be turned into an int or not. If so, just pass the node id directly in. If not, it must be a string. Grab the last 8 characters of the node id and force-format the id to have a ! prefix. That way users can provide any prefix they want to the meshtastic CLI.

[ianmcorvidae]

This doesn't look bad, aside from the CI errors (looks like it wants a tuple, not a list, for the metavar). However, I'd prefer if this handling was moved into _sendPacket, which is what currently processes the !xxxxxxxx into integers, and it's a bit lower level. Not everything flows through sendData, though most does.

[migillett]

@ianmcorvidae ok! I can move this into the _sendPacket function. That's no problem. Do we still want to keep the documentation for [!xxxxxxxx, xxxxxxxxxx] or drop those changes?

[ianmcorvidae]

I think those metavar changes are fine but it looks like mypy wants it to be a tuple instead of a list. That should just mean ("!XXXXXXXX", "0xXXXXXXXX") instead of ["!XXXXXXXX", "0xXXXXXXXX"] for those, I believe.

I do also notice that it's complaining that int | str isn't indexable due to the code you have ensuring the presence of the !. With integrating it into _sendPacket that may go away, though. If you want to check the stuff that CI does, but locally:

poetry run pylint meshtastic examples/ --ignore-patterns ".*_pb2.pyi?$"
poetry run mypy meshtastic/
poetry run pytest

[migillett]

@ianmcorvidae ok, I think that got it. I tried several things but the metavars deal would constantly throw errors no matter what I did. Once I went back to the old metavars definition of metavar="!xxxxxxxx" the tests passed successfully. I instead added a bit to the help deal telling about using ! or 0x prefixes. Technically no prefixes are required at all with how we're parsing hexadecimal strings with this change though. Want to see what you think about that one.

I also ran a manual test locally between my nodes using ! and 0x prefixes and both worked

[migillett]

I could also do a if nodeId.startswith('!') or nodeId.startswith('0x'): too. but I didn't want to hard-code prefixes unless necessary. But I'm game for either.

[migillett]

just saw the build errors, I'm on it.

[migillett]

@ianmcorvidae sorry about that. fixed the issues the linter was yelling about.

[ianmcorvidae]

Perfect, no worries. I'll kick it off again in case there's anything else. The linter can get a bit annoying at times so that's understood, haha

[ianmcorvidae]

Looks like that's all resolved now. Thanks for the contribution!

[codecov]

Codecov Report

Attention: Patch coverage is 62.50000% with 3 lines in your changes missing coverage. Please review.

Project coverage is 60.58%. Comparing base (fc3b81d) to head (23ea19c).
Report is 40 commits behind head on master.

Files with missing lines	Patch %	Lines
meshtastic/mesh_interface.py	40.00%	3 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #731      +/-   ##
==========================================
+ Coverage   60.35%   60.58%   +0.22%     
==========================================
  Files          24       24              
  Lines        3970     4031      +61     
==========================================
+ Hits         2396     2442      +46     
- Misses       1574     1589      +15     

Flag	Coverage Δ
unittests	60.58% <62.50%> (+0.22%)	⬆️

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

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[migillett]

@ianmcorvidae you're welcome! hopefully the first of many.