PyK: Misc enhancements to KCFG and anti-unification

[nishantjr]

• KCFG: Shorten log message for easier readability

  This brings the more important parts of the messages into focus.

• CTerm.anti_unify: Don't call ml_pred_to_bool unless necessary

  Since these values are only used if keep_values is set, we do not need to always
  compute them. In addition, it avoids exceptions when the predicate cannot be
  converted to a bool expression: e.g. when using #Ceil as a side condition

  This is only a work-around (it would
  still break for a #Ceil side condition, if keep_values=True), but lets us move forward without risking
  breaking other projects that rely on the original behaviour.

• pyk:KCFGShow.to_module: Thread imports and defunc_with parameters through to KCFG.to_module

  This lets us use those options for printing the modules.

[ehildenb]

@nishantjr it looks ilke you updated the Haskell and LLVM submodules both in your PR, is that intentional? In general, you shouldn't do that, because it missed updating all of the things like nix flake, pyproject.toml, and dependency list files. Can you back those changes out?

[nishantjr]

@nishantjr it looks ilke you updated the Haskell and LLVM submodules both in your PR, is that intentional? In general, you shouldn't do that, because it missed updating all of the things like nix flake, pyproject.toml, and dependency list files. Can you back those changes out?

Ugh, accidentally committed after a rebase. Fixed.

[nishantjr]

@ehildenb I've pushed another change that exposes KCFG.to_module's imports parameter in KCFGShow.

[ehildenb]

@nishantjr , looking at the CI failures, I think this test is the most telling one: TestKonvertCellMap.test_kast_to_kore[account-cell-map]. I guess this comes from the change in how you're handling cell-maps, maybe that change isn't needed? If it is needed, please add a test demonstrating why it's needed, so we can make sure that one is passing and current ones are passing.

[ehildenb]

I re-triggered teh CI run, it was due to the haskell server becoming unavailable for some reason. Maybe changing the way that modules are built is causing some issues? But I figured it's worth re-triggering the run.

Can you update the PR description so we get a good commit message?

[nishantjr]

Thanks Everett. How does it look now? I've backed out one of the commits; will add in the test for it next week.