Eliminating the white space from the xref macro

[MichalMaler]

The correct way of using xref macro is documented in https://quarkus.io/guides/doc-reference#anchors-reference
This PR just cleans it up so that contributors can not get a bad example directly from the source code.

Kudos to @jmartisk for creation of the regex to catch them all.

[gastaldi]

Good catch!

[cescoffier]

Question: Should we have a linter with such kind of rules?

[michelle-purcell]

Question: Should we have a linter with such kind of rules?

Great idea @cescoffier . We could investigate adding this to the Vale ruleset for Quarkus.
@MichalMaler - WDYT? let's try this out... 🧪
BTW... a PR is on the way to fix most of the false-positives too.

[gsmet]

Does it hurt to have a whitespace? Just wondering how much we should invest in it.

[MichalMaler]

Does it hurt to have a whitespace? Just wondering how much we should invest in it.

I guess it doesn't hurt, but I like to keep things unified if possible.
The recommendation to not to use the space after comma come directly from you; you told me this while i was drafting the cross-ref guide (March 22)

[gsmet]

Well, I don't recall saying that tbh. I said it was working with or without equally but the doc is without and that our example should be consistent with the AsciiDoctor doc.

I'm not saying we shouldn't merge this, it's good. I wouldn't make it a priority to have a tooling to check that though, if it doesn't cause any problems, given all the other issues we have that actually cause problems.

[MichalMaler]

Well, I don't recall saying that tbh. I said it was working with or without equally but the doc is without and that our example should be consistent with the AsciiDoctor doc.

I'm not saying we shouldn't merge this, it's good. I wouldn't make it a priority to have a tooling to check that though, if it doesn't cause any problems, given all the other issues we have that actually cause problems.

But you are saying in next sentence that you dont think it matters.
So, as I said, this is just a cosmetic tweak, clean-up, unification, to reach some consistency. :)

Vale can use this as a recommendation, or I can do this check once a year, or drop this completely.
No hard feelings if this will not got merged.

[gsmet]

Looks like I had a good recollection of what I said :)

I'm not sure it's coming from these changes but the doc build is failing with:

[INFO] asciidoctor: INFO: possible invalid reference: duplicated-context
[INFO] asciidoctor: INFO: possible invalid reference: duplicated-context

[gsmet]

We need to understand what's going on with ^ but we can merge once it's fixed.

[MichalMaler]

Looks like I had a good recollection of what I said :)

You are my greatest Quarkus hero, would never said something you didn't :))

All I did was use this regex to search for the space after a comma inside <<>> in the repo .adoc files with:
(<<[A-Za-z-]+,)
And replace it with:
$1

The truth is, this operation had two iterations (but it was maybe caused by the fact that my IDEA didn't find all occurrences when I hit the FindNReplace button for the first time.

I double-checked all stuff I touched and they are all in the <<>>, thus should not touched the context or and ID.
My local build with mvn clean package finished OK too.

[gsmet]

Merging as the failure is not related. I will fix it in an upcoming PR.

[github-actions]

🙈 The PR is closed and the preview is expired.

[gsmet]

See #35089

[michalvavrik]

This PR just cleans it up so that contributors can not get a bad example directly from the source code.

Yeah, that's exactly how I work :-) Thanks!