Add validate command

[surge119]

Summary

Added command to Fixit to check if a config file is valid

Test Plan

Created invalid fixit.toml files such as:

[tool.fixit]
enable = [
    "fixit/rules:DeprecatedABCImport",
]
disable = [
  "fixit.rules",
]

root = true

and ran the new command against them:

# Example from above
$ hatch run fixit validate-config .fixit.toml
Failed to parse rule `fixit/rules:DeprecatedABCImport`: ConfigError: invalid rule name 'fixit/rules:DeprecatedABCImport'

# Another example with invalid TOML
$ hatch run fixit validate-config .
Invalid config: TOMLDecodeError: Expected '=' after a key in a key/value pair (at line 9, column 4)

# pyproject.toml of this project:
$ hatch run fixit validate-config pyproject.toml
Failed to parse rule `.examples.noop`: CollectionError: could not import rule(s) .examples.noop from examples

# Example of config with error and rules with errors:
Failed to parse rule `fixit.rules:DeprecatedABCImport`: NameError: name 'a' is not defined
Failed to parse rule `fixit/rules:DeprecatedABCImport`: ConfigError: invalid rule name 'fixit/rules:DeprecatedABCImport'
Failed to parse rule `fixit.rules`: NameError: name 'a' is not defined

[amyreese]

I think I would prefer this to take a direct path to a config (eg, pyproject.toml or fixit.toml), read the "raw" config with read_configs([path]), and then manually call parse_rule() for each string in every enable/disable option (including in every override block). Then call find_rules() on each qualified rule and log errors/warnings for anything that fails to load.

That way, we're not validating the config for a single code path in the repo, but instead validating all possible rules that are referenced from the file, including in all override blocks, to make sure that rules are correctly expressed, findable, and importable.

Bonus points if the warnings/errors can reference the specific block/enable/disable section to make it clear to the user which element is causing the problem.

[amyreese]

Looks great overall. One last request: can we add a simple smoke test that tests this command against a bogus config file and make sure it produces some baseline expected output? You should be able to follow examples in tests/smoke.py for validating expected output on stdout.