nvim api commands test

[gigamaax]

I started trying to fix the tests, case by case but I'm not sure if the direction I went for the first test is something you'd agree with. I also don't know enough about message pack, this codebase's implementation of messagepack, nor the internals of nvim_get_commands to know if this is dumb. so I figured I'd open up a pr and get some feedback early on.

very specifically, this pr addresses only two failing tests:

    --- FAIL: TestAPI/Command (0.00s)
        --- FAIL: TestAPI/Command/Commands (0.00s)
            --- FAIL: TestAPI/Command/Commands/Nvim (0.00s)
                api_test.go:2251: msgpack: cannot convert Nil to string
            --- FAIL: TestAPI/Command/Commands/Batch (0.00s)
                api_test.go:2267: msgpack: cannot convert Nil to string

as far as I can tell, what's happening is strings are sometimes nil. I'm not sure what the root case for that is, or if it's even really an issue. it appears to mostly only matter on the omitempty fields, which makes me feel like it should be alright if there's nothing there. so, rther than consume the error, I just said it was an empty string.

the other issue is the actual expect case. and this is the thing that I'm most worried is potentially wrong:
for v0.9, no commands are returned. for 10, three are returned, and they appear to be treesitter commands. for nightly, assuming 11 as well, but I'm still running 10 so I can't confirm, 4 are returned. it's the same three in v10, EditQuery, Inspect, and InspectTree. the last one is Open.

I'd imagine that { builtin = false } should filter those out, but maybe since they're not vim commands and more specific to neovim, they get special treatment? I coldn't find specific docs about any of the commands to figre it out.

let me know what you think. I'm happy to make any adjustments. also, let me know if you'd prefer a draft pr next time :)

[codecov]

Codecov Report

Attention: Patch coverage is 0% with 2 lines in your changes missing coverage. Please review.

Project coverage is 83.1%. Comparing base (dd77a91) to head (96d7cca).
Report is 9 commits behind head on main.

Files with missing lines	Patch %	Lines
msgpack/decode.go	0.0%	2 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##            main    #183     +/-   ##
=======================================
- Coverage   83.3%   83.1%   -0.3%     
=======================================
  Files         14      14             
  Lines       3132    3134      +2     
=======================================
- Hits        2610    2605      -5     
- Misses       522     529      +7     

Flag	Coverage Δ
linux-1.18-nightly	?
linux-1.18-stable	83.1% <0.0%> (-0.3%)	⬇️
linux-1.19-nightly	?
linux-1.19-stable	?
linux-1.20-nightly	?
linux-1.20-stable	?
macos-1.18-nightly	?
macos-1.18-stable	?
macos-1.19-nightly	?
macos-1.19-stable	?
macos-1.20-nightly	?
macos-1.20-stable	?
windows-1.18-nightly	?
windows-1.18-stable	?
windows-1.19-nightly	?
windows-1.19-stable	82.2% <0.0%> (-0.3%)	⬇️
windows-1.20-nightly	?
windows-1.20-stable	82.2% <0.0%> (-0.3%)	⬇️

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

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[justinmk]

Thank you! Yeah this looks fine to me. To see what the test is doing, run nvim -u NONE and then run:

lua =vim.api.nvim_get_commands({})

on Nvim 0.11, that indeed returns 4 commands. So this particular test is sensitive to the Nvim version.

[gigamaax]

ok, cool. thank you!

I've updated the keymap test to be less specific about the version of neovim and focus a little bit more on the direct actions taken within the test. I think that should make it a lot less fragile with new verisons.

I'm going to close this for now while I work on the other failing tests and reopen when they pass