Add comments explaining performance improvement in TypedData

I would like to suggest a comment like this for this code part.

Reading the original comment on [#14908519] mentions this was changed for performance. Believe the phrase "is the fastest." covers that point. So think this is good.

Hi!
Are any additional changes needed here?

I was about to say no, nothing else to do, but I think there is:

1. The comment should link to where the performance optimization is documented, that seems to be https://blog.blackfire.io/php-7-performance-improvements-encapsed-string...

2. Are we sure that this is still a performance improvement in PHP 8? i.e. that the concatenation operator hasn't had similar improvements since?

Thanks @catch for your comments.

1. Added a link to the blog article in the description.

2. I'm not sure if any performance changes were made for the PHP 8 for the string concatenations. I couldn't find any mentions about that.

Link has been added to comment per recommended.

I know we don't mention interpolation (will admit no idea what that is) but these were the best links I think I could find

https://gist.github.com/Atulin/406881ee184aede02475b832147f7c12

https://wpshout.com/php-concatenate-strings/#gref

Here we’re using the PHP concatenate operator, . again, rather than relying on PHP string interpolation. There is good-and-bad in both of these options. And all three work fine. The performance overhead of using string interpolation on double-quoted strings is small enough that I don’t think you should worry about it. But you will hear people point to the (true) fact that the third example in this section is (very marginally) faster

PHP STRING INTERPOLATION? USE SINGLE OR DOUBLE QUOTES? section

Not sure if that helps answer the question?

The conclusions in 📌 Evaluate replacing all usage of sprintf by dot concatenation or variable inside string Active contradict the comments here.

From reading 📌 Evaluate replacing all usage of sprintf by dot concatenation or variable inside string Active this still something we want to do?

#12 is right, this should explicitly mention this is being done because the code can run in the critical path.

Directly added mentions that this code can run in the critical path.

Committed and pushed efecede14e to 11.x and d4f1ab7f8e to 10.2.x and 841980744f to 10.1.x. Thanks!

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