feat: modify command in SKEL (SK-167)

[drmorr0]

Related Links

• SK-167

Description and Rationale

• modify is now a fully-supported SKEL function (originally we called this apply but I decided late in the process that modify was better)
• Checked additional variable errors/issues in the AST generation instead of the engine
• build module updated to support make debug-test
• fix the rhs_to_value function in the event that the first component isn't a variable
• updated the docs

How

Checking defined variables in AST generation

It turns out we were already tracking a defined_vars set in the ast code, so it was a "relatively" simple matter of plumbing that through to more places. It's a little subtle/tricky, because in the left-hand-side of a resource test, you can only reference the variable that's defined in that test, but in the right-hand side you can reference variables that were defined in previous tests. In other words, this is invalid:

$x := some.path | $y == foo

But this is valid:

$x := some.path | exists($x) && $y := some.other.path | $y == $x

We also got rid of the var_prefixed_resource_path rule in the grammar, because it turns out a) it over-complicated the grammar, and b) it disallowed some valid expressions.

Modify SKEL command

This was more-or-less straightforward, we just needed to ensure that all the different variants in the assignment worked correctly (e.g., path = val, $x.path = val, path = $y, and $x.path = $y).

We created a reify_pointers function which takes an abstract pointer and the variable context, and converts that to a list of concrete pointers that match the object with all the variables filled in. This actually simplifies the logic for the remove command as well, and fixes a couple corner cases I didn't think about before (specifically, if the variable substitution doesn't work because we didn't reference a particular variable in the abstract pointer, that would have silently gone through before; I don't think it actually would do anything but still probably isn't ideal).

Test Steps

• New unit and integration tests written
• manual testing against some real traces using a file like the following, confirmed that the node selectors were updated as expected

modify(@t >= 1h && spec.template.spec.nodeSelector."simkube.dev/foo" == "bar", spec.template.spec.nodeSelector."simkube.dev/foo" = "baz");

Other notes

When I was writing this before, I forgot about the klabel macro, which is super-handy. Would love to get some of this stuff extracted out into a shared library at some point.

• I certify that this PR does not contain any code that has been generated with GitHub Copilot or any other AI-based code generation tool, in accordance with this project's policies.

[codecov]

Codecov Report

❌ Patch coverage is 97.22222% with 5 lines in your changes missing coverage. Please review.
✅ Project coverage is 78.70%. Comparing base (f7b6f42) to head (7f22473).
⚠️ Report is 1 commits behind head on main.

Files with missing lines	Patch %	Lines
sk-cli/src/skel/ast.rs	97.56%	2 Missing ⚠️
sk-cli/src/skel/engine.rs	97.46%	2 Missing ⚠️
sk-cli/src/skel/mod.rs	0.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #227      +/-   ##
==========================================
+ Coverage   78.23%   78.70%   +0.46%     
==========================================
  Files          63       63              
  Lines        3690     3789      +99     
==========================================
+ Hits         2887     2982      +95     
- Misses        803      807       +4     

☔ 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.

[linear]

SK-167 Support `apply` in the SKEL parser