i am sure the elastic search is better than the mysql text search but there is no reason I can see for wordpress to not use the text search. The same pre processing that needs to be done with the elastic search can be applied to creating the text search "knowledge base". That by itself, without any need for complex 3rd party install and configuration will drive forward wordpress search by light years.

... in the end of the day, people that need a good front end search should be willing to fork the extraa money for it, but everyone needs a backend search.

back to the out of memory problem. So I made my shortcode handling better, but than the core shortcode handling failed.

First time I see a situation in which outputting is a much better strategy than returning values.

@TomJNowell I used to use the Revelansii plugin to get better results, but that killed the server performance so I've just accepted it for what it is.

Tom, completely unrelated to that. What is the preferred approach of commenting on WordPress code? Is there a specific php commenting standard out there that WordPress developers should follow?

Just in case - back in 2009-2012 I used sphinxsearch.com + custom action to transform pdf/docs/xls to make all content on our site searchable. After that we also switched to Google CSE, but its discontinues since last year =(((

I recommend using the mPDF library to convert PDFs. It works surprisingly well. I made a convert post to PDF plugin recently that I should release.

@ChristineCooper PHPDocs and common sense

@ChristineCooper just use the wordpress coding standard linter and it will make sure you are commenting the "wordpress way"

@Butuzov yeh sphinx was all the rage few years ago. I worder what happened to it as iit seems like people speek only about elastic search now

is there a simple way to remove html that isn't wrapping anything?

like <span class="some-class"></span>

or <p id="some-id"></p>

Regex I imagine -> regex101.com/r/m075e7/1