HTML API: Fix splitting single text node

[sirreal]

Fix an issue where a < character will always break a text node.

It should not break a text node if we will not start another node type, i.e. we find < followed by !, /, ? or an ascii alpha character.

Trac ticket: https://core.trac.wordpress.org/ticket/60385

---

This Pull Request is for code review only. Please keep all other discussion in the Trac ticket. Do not merge this Pull Request. See GitHub Pull Requests for Code Review in the Core Handbook for more details.

[github-actions]

Test using WordPress Playground

The changes in this pull request can previewed and tested using a WordPress Playground instance.

WordPress Playground is an experimental project that creates a full WordPress instance entirely within the browser.

Some things to be aware of

• The Plugin and Theme Directories cannot be accessed within Playground.
• All changes will be lost when closing a tab with a Playground instance.
• All changes will be lost when refreshing the page.
• A fresh instance is created each time the link below is clicked.
• Every time this pull request is updated, a new ZIP file containing all changes is created. If changes are not reflected in the Playground instance,
  it's possible that the most recent build failed, or has not completed. Check the list of workflow runs to be sure.

For more details about these limitations and more, check out the Limitations page in the WordPress Playground documentation.

Test this pull request with WordPress Playground.

[sirreal]

@dmsnell Please take a look at this.

[sirreal]

I don't love this, I think it's doing character code comparison but I couldn't find conclusive docs.

I considered ctype_alpha but apparently that's locale dependent so best avoided. We want ascii alpha.

We could also apply ord and compare (65 <= ord( $char ) <= 122) and comment the magic numbers.

[dmsnell]

based on the PHP docs for the comparison operators I think this is valid as-is, as long as we're comparing strings to strings

[dmsnell]

Good catch @sirreal. While this shouldn't have caused any problems in an application, it's more correct combining these. I wanted to re-examine the rules and fallback to the existing "failure to find a tag" but I think having this up-front has its merit.

Did you examine trapping this inside of next_tag() in the case that parse_next_tag() returns false? I'm a bit surprised that this wasn't behaving differently, but I guess that loop runs to the end.

My one big remaining question is whether we expect this to be a normal case or a rare case. That maybe would suggest whether to eagerly re-compute the next character, as we do in this patch, or put it at the end of the loop in parse_next_tag(). We can adjust as we see fit.

[dmsnell]

Merged in [57489]
0b800d7

[sirreal]

I wanted to re-examine the rules and fallback to the existing "failure to find a tag" but I think having this up-front has its merit.

My first thought was that we should find the text node once we've found the start of the next node, but that seemed like a much larger refactor. What I don't like about this approach is that we're effectively identifying the start of the next node in multiple places.

I think it's fine for now, this is a smaller change with the 6.5 release on the horizon and as test coverage increases we can explore larger changes before the next release.

Did you examine trapping this inside of next_tag() in the case that parse_next_tag() returns false?

I did not.

My one big remaining question is whether we expect this to be a normal case or a rare case.

I don't have data to answer that. My gut suggests this is unusual, but I have no idea.

That maybe would suggest whether to eagerly re-compute the next character, as we do in this patch, or put it at the end of the loop in parse_next_tag(). We can adjust as we see fit.

I think you're describing the approach I'd like, but again I think that would be a larger refactor and perhaps best left for the next release cycle.