get_page_by_path() enhancements

[Denis-de-Bernardy]

The attached file works around a few queries in get_page_by_path() when memcached is used on a site, by trying to fetch page parents in the object cache before fetching the needed data in the database.

Related: #11292 and #10381.

Is there any good reason why get_page() or get_post() couldnt be used inside that loop to take care of the query & caching?

Replying to dd32:

Is there any good reason why get_page() or get_post() couldnt be used inside that loop to take care of the query & caching?

Depends if you want to do a SELECT * or not... This one can safely get checked in, I've been using it on my own site for months.

Replying to dd32:

Is there any good reason why get_page() or get_post() couldnt be used inside that loop to take care of the query & caching?

get_page() is used below, when desired page is found. While it's not found, database is queried inside 2 TWO loops over and over again searching for the desired page. And when that's the case, all we need is post_parent, post_name and ID.

If that post is in cache, there's no need to query database, just get it from cache and move on. Faster and secure. I vote for this patch to be commited :)

Please don't change reported versions. They should reflect the version used in the report, or in cases where it is researched, the earliest instance where the report applies.

See also: #14357

Patch refreshed. +1 on this one

This looks good.

(In [16206]) Eliminate some queries in get_page_by_path() by consulting cache. Props Denis-de-Bernardy, wojtek.szkutnik. fixes #11373