Do not create conda error reports for common / expected exceptions

[jaimergp]

Description

Closes #5263

This is what we would need to close that issue without registering new base classes in conda exception handler.

CondaError calls super() so we can't compose new exception types that inherit both from CondaError and e.g. CalledProcessError. This prevents us from having a mixed exception that doesn't change which type of exception is raised (for the purpose of try/except idioms or isinstance checks).

I'd rather have this:

class BuildScriptException(CondaError, CalledProcessError):
  ...

and then call:

try:
  subprocess_for_build_script()
except CalledProcessError from exc:
  raise BuildScriptException(exc, ...) from exc

Achieving this ^ will require changes in conda/conda though (either changing the CondaError class so it doesn't call super(), or having a different base class for the purpose of exception handling like BaseCondaError).

Checklist - did you ...

• Add a file to the news directory (using the template) for the next release's release notes?
• Add / update necessary tests?
• Add / update outdated documentation?

[codspeed-hq]

CodSpeed Performance Report

Merging #5264 will not alter performance

_{Comparing jaimergp:error-gracefully (d870fed) with main (9c0ceec)}

Summary

✅ 3 untouched benchmarks

[jaimergp]

pre-commit.ci autofix

[jezdez]

@kenodegard @jaimergp How does this overlap with #5255?

[jaimergp]

Mostly compatible. We need to agree on a base exception for all conda-build exceptions, or just subclass CondaError for each one. I chose the former, but Ken's does:

class CondaBuildUserError(CondaError):
  ...

vs my idea:

class CondaBuildError(CondaError):
  ...

class CondaBuildUserError(CondaBuildError):
  ...

I don't mind either, but we need to agree on one.

[kenodegard]

I don't mind either with how we define CondaBuildUserError

I only opted for CondaBuildUserError(CondaError) to avoid any potential side-effects of inheriting directly from CondaBuildError 🤷🏼‍♂️

We've also decided it would be better to split #5255 into smaller PRs for easier review and testing so merging this ahead of reworking all of that would be good

[beeankha]

Suggested change

	Test whether settings from .condarc files are needed.
	Test whether settings from .condarc files are heeded.

I think this wording edit needs to be reverted, since this test seems to be checking whether or not the settings in the .condarc files are actually being taken into account/paid attention to (?)