It's yet another side effect of working on #2622 that could be applied separately.

it's not immediately clear what this PR does... in what way does it support match and throw expressions? (i.e. a PR description would be appreciated 😉 )

Isn't KeywordCompletor enough in the title?

I guess CHANGELOG.md should be updated too in this case.

Isn't KeywordCompletor enough in the title?

not really - for example:

This MR will ensure that the match and throw keywords are suggested in completion results,

or

This MR will ensure that match and throw keywords are recognised as expressions and allow subsequent expression completions.

or whatever - I had no idea this had anything to do with snippets 🙂 basically I don't want to have to reverse engineer (and guess) what the intent of the PR is from looking at the code.

Isn't KeywordCompletor enough in the title?

not really - for example:

This MR will ensure that the match and throw keywords are suggested in completion results,

or

This MR will ensure that match and throw keywords are recognised as expressions and allow subsequent expression completions.

or whatever - I had no idea this had anything to do with snippets 🙂 basically I don't want to have to reverse engineer (and guess) what the intent of the PR is from looking at the code.

I appreciate this idea and try to keep it in mind and apply, but it is not always clear for me what user the commit is addressed for, especially when I look at the previous commit titles from the master branch 😉. On the other side my (still poor) English may cause additional confusion - please don't hesitate to correct my possibly strange language constructions.

but it is not always clear for me what user the commit is addressed for

well you can assume that they're aimed at me or otherwise the person that's reviewing the PR 🙂 My own PRs often don't feature extensive descriptions but I'm the one merging and reviewing them so have the maintainer's privilege there.

If my PR breaks Phpactor it's my fault - if your PR breaks Phpactor it's also my fault 😄 so the clearer the intent of the PRs the better

so with this MR using an open parenthesis ( results in throw and match being suggested, which is why I'm reluctant to merge these types of changes. the code is very fragile and it's difficult to know what the consequences are.

Yes, I see that the case of cursor placed right after open parenthesis ( is often not directly recognized by tolerant parser and requires to be expressed in an indirect, a bit hacky way. Thus the need to be guarded by enough tests.

@przepompownia (not sure how else to contact you) if you like stickers let me know your address via. my git email and I'll send you a couple of Phactor ones :)

@przepompownia (not sure how else to contact you) if you like stickers let me know your address via. my git email and I'll send you a couple of Phactor ones :)

Nice to have a gadget promoting Phpactor and maybe keep to share with a few people potentially interested in :)