Improve error message for an invalid sniff code

[fredden]

Description

This is an implementation of the suggestions made by @gsherwood in squizlabs/PHP_CodeSniffer#2646 (comment)
I have intentionally not looked at the code changes initially suggested in squizlabs/PHP_CodeSniffer#2646 (to avoid any copyright woes).

Suggested changelog entry

Improved error message when invalid sniff codes are supplied to --sniffs or --exclude command line arguments.

Related issues/external references

Replaces / closes squizlabs/PHP_CodeSniffer#2646
Related to #319

Types of changes

• Documentation improvement

PR checklist

• I have checked there is no other PR open for the same change.
• I have read the Contribution Guidelines.
• I grant the project the right to include and distribute the code under the BSD-3-Clause license (and I have the right to grant these rights).
• I have added tests to cover my changes.
• I have verified that the code complies with the projects coding standards.
• [Required for new sniffs] I have added XML documentation for the sniff.

[jrfnl]

Thanks for this PR @fredden.

Unfortunately, there are no pre-existing tests covering the changes, so new tests will need to be written to cover this change - as the line you removed from the "PR checklist" pointed out... (that checklist item is there for a reason... I've added it back now)

Some observations when I look at the behaviour around this error message:

• If multiple erroneous names are passed via the --sniffs option, only the first error is displayed, so the user fixes that one, runs PHPCS again and will then be confronted with the next error.
  I believe this should be made more user friendly.
• If the list of sniffs contains a stray comma, like --sniffs=A.B.C,,D.E.F, this will be reported as "The specified sniff code "" is invalid".
  I believe we can filter out/ignore empty sniff codes altogether without bothering the end-user with this.

As this PR introduces a dedicated method to validate the sniff codes (👍🏻), I think we should take this opportunity to make the fix comprehensive.

Other than that, I find the messaging a bit "wordy", so I wonder whether it can be made more succinct.

[fredden]

I have discussed this with @jrfnl. I intend to add tests to cover the existing behaviour in a separate pull request. I'll mark this as a draft until that change is complete. I'll come back to this when that change is done and make any necessary changes here at that time.