Harmonize help commands' --format to --output-format with deprecation warnings

[akx]

Summary

Since --format was changed to --output-format for check, it feels like it makes sense for the same to work for the auxiliary commands.

This

• adds the same deprecation warning that used to be a thing in Rename format option to output-format #7514 (and un-became a thing in Remove support for providing output format via format option #7984)

Fixes #7990.

Test Plan

• cargo run --bin=ruff -- rule --all --output-format=json works
• cargo run --bin=ruff -- rule --format=json works with warnings

[MichaReiser]

Nice. I linked the related issue. Would it be possible to log a warning when using --format that mention that it is deprecated in favor of --output-format?

[akx]

Ah, I didn't even know that there was an issue about this 😁

Sure, I can canonicalize it all to output_format and add warnings, but will it be weird that the help commands support a subset of the main check output formats with the same flag name?

[github-actions]

PR Check Results

Ecosystem

✅ ecosystem check detected no linter changes.

[MichaReiser]

Sure, I can canonicalize it all to output_format and add warnings, but will it be weird that the help commands support a subset of the main check output formats with the same flag name?

I think that's fine, especially considering that it used to be the same before? @zanieb what's your take?

[akx]

@MichaReiser Done deal ✨

[zanieb]

I think this needs a bit more discussion before it goes out.

I think it's okay for a subset of formats to be supported on the CLI, but it doesn't really make sense when using the environment variable. Like.. if you've set your output format in your environment to something not supported by another command it'll just fail?

In combination with that concern, it seems problematic to add support for RUFF_FORMAT to commands to that did not previously have it as all use of those commands could start to fail.

In general, I'm also not sure it makes sense to add support for deprecated options to new commands i.e. adding --format support to ruff version.

[charliermarsh]

I think it's okay for a subset of formats to be supported on the CLI, but it doesn't really make sense when using the environment variable. Like.. if you've set your output format in your environment to something not supported by another command it'll just fail?

Can you expand on this?

[zanieb]

Well we have the output formats supported by ruff check: text, json, json-lines, junit, grouped, github, gitlab, pylint, azure

And then we have the output formats supported by "help" commands: text, json

If we fail on unknown formats and use RUFF_FORMAT/RUFF_OUTPUT_FORMAT for both of these I believe this will break existing usage and be confusing for future users.

export RUFF_OUTPUT_FORMAT=github
ruff check    # OK
ruff version  # FAILS with invalid output format

I don't think the help formats should use the same variable at least, and maybe they shouldn't support environment variables at all.

[charliermarsh]

Hmm I see. How about leaving the --format and --output-format support for these commands (for consistency, for now -- we'd then deprecate --format for all of them at once), but omitting the environment variable support?

[akx]

@charliermarsh @zanieb Done: removed the envvars, removed --format from version (which hadn't had it in the first place).

[zanieb]

Thank you for making the changes! I have some minor comments.

[zanieb]

Perhaps resolve_help_output_format would be a better function name? No strong feelings though.

[charliermarsh]

Thank you as always @akx!