Require short ternary (Elvis operator) syntax

Big +1, surprised this isn't already a thing!

Would be +1, but still requires the actually proposed changes, of course.

Honestly I don't know what to propose, it's essentially "turn on SlevomatCodingStandard.ControlStructures.RequireShortTernaryOperator and let phpcs enforce it". Happy for others to come up with the wording for the coding standards.

+1

Added proposed changes to issue summary

Added acbramley and alexskrypnyk as supporters (confirmed in slack)

Thanks for proposing the text, looks good!
Added myself as a supporter.

Next step should be the change record it seems, after that it can be discussed in the next meeting.

Added a change record to keep it moving :)

[#3446111]

Setting to RTBC so it gets reviewed by the committee.

Updating status on behalf of @quietone as per #13.

Can someone check if there are any sniffs to enforce this change?

@quietone the sniff is already implemented in 📌 Require short ternary (Elvis operator) syntax Postponed

My apologies for the oversight.

This was announced in blog post → published 9 Oct.

There has been no feedback from the blog post.

Im now a -1 on this issue, copying my comment from the core issue:

I've been making use of phpstan strict ruleset for a couple years now, and I have to mention that I agree with one of their rules, that would be at odds with this issue. That is, to never use ?:. Because it has weak comparison semantics.

Justification, from docs:

Disallow short ternary operator (?:) - implies weak comparison, it's recommended to use null coalesce operator (??) or ternary operator with strict condition.

So, since this issue is about core itself, not coding standards. I'd support enabling this rule from strict, and/or in the interim finding a codesniffer rule that disallows it.

So I'd be advocating for SlevomatCodingStandard.ControlStructures.DisallowShortTernaryOperator

The proposed text talks about a specific scenario which is already a non-strict comparison. So disallowing the short ternary syntax does not improve this. It may encourage people to learn more as to why it's disallowed, and refactoring to use a strict comparison, but in most cases people would probably just expand it out to the longer syntax.

I've updated the documentation and published the change record.

Reviewed the documentation page, that still had visibile html tags, fixed those. I think this is done now.

@dpi, I think it best to discuss using SlevomatCodingStandard.ControlStructures.DisallowShortTernaryOperator in a meeting or a separate issue.

Updating credit. I've included myself as well.

OK. this is now complete

Thanks everyone!

Automatically closed - issue fixed for 2 weeks with no activity.