Optimize PHP html_entity_decode function

[ArtUkrainskiy]

Improvements affect the C function traverse_for_entities:

• Use memchr to search for '&' instead of scanning character by character.
• Use memchr to locate ';' to determine potential entity boundaries instead of process_named_entity_html, avoiding unnecessary per-character validations.
• Use memcpy instead of character-by-character copying.
• Refactor code for improved structure and readability.

Benchmark branch - https://github.com/ArtUkrainskiy/php-src/tree/html_entity_decode/benchmark

------------------------------------------------------------------------------
|                      Test |     old avg(ns) |     new avg(ns) |    diff(%) |
------------------------------------------------------------------------------
|                      4k & |            5949 |           21115 |    -71.98% |
------------------------------------------------------------------------------
|             only entities |            8279 |           10202 |    -18.80% |
------------------------------------------------------------------------------
|        400 valid entities |            6439 |            5861 |      7.80% |
------------------------------------------------------------------------------
|        200 valid entities |            4891 |            3178 |     38.12% |
------------------------------------------------------------------------------
|        200 invalid entity |            4777 |            3181 |     37.29% |
------------------------------------------------------------------------------
|             200 ampersand |            4809 |            1221 |    198.35% |
------------------------------------------------------------------------------
|        100 valid entities |            4188 |            1777 |    124.49% |
------------------------------------------------------------------------------
|         50 valid entities |            2885 |             979 |    193.50% |
------------------------------------------------------------------------------
|        String ends with & |            2428 |             176 |   1221.69% |
------------------------------------------------------------------------------

As you can see, the speedup depends on the number of entities and & characters in the string — the fewer there are, the more noticeable the performance improvement.

In edge cases, where the string consists entirely of & characters or valid HTML entities, performance actually worsens. However, I don't think this is a common scenario.

Either way, I plan to continue optimizing and implement & scanning using SIMD instructions, which should significantly improve performance even in high-entity-density cases.

[bukka]

@Girgias are you going to review the logic as well? Just checking if I should look into this or if you are happy to handle it all?

[Girgias]

@Girgias are you going to review the logic as well? Just checking if I should look into this or if you are happy to handle it all?

Please do review the logic, I only had a cursory glance :)

[bukka]

Ok I will check it out next week if no one is quicker.

[dragoonis]

Nice idea @ArtUkrainskiy :-) 👍

[bukka]

I think it looks reasonable except that introduction of valid_entity boolean and checking that everywhere which doesn't look like performance optimization to me. I understand that it was probably meant to make code more readable but not sure if it's worth it in this case. Might be worth to check if it has any impact.

[bukka]

This looks better. The comments are pretty much only for minor issue / optimizations. Overall I think it looks good.

[bukka]

I went through this again and think it looks good. Doesn't make sense to hold it because of few NITs which I can easily address during the merge. I will try do a bit of testing in about two weeks time and merge it then.

[bukka]

I have done a bit of testing. Also fixed few nits and squash / rebased it so think it should be good enough. I will do one last round of testing in a couple of weeks and if I don't find anything, I will merge it.

[bukka]

I did a bit more checking and it seems all fine so merged to master. Thanks for the contribution!