PageSpeed Insights 和 php 包括

Question

I'm cleaning up the web page (HTML and PHP) code to improve the PageSpeed Insights reports and I noticed a bad result using PHP include (or include_once ).

To share common code between pages I use to copy the code to external files (ie inc_header.php ) and include them in more pages. It is a consolidated practice.

Today I noticed relevant different results in PageSpeed if I test a static page or the same page using a couple of included files.

Is it true that PHP includes influence the PageSpeed Insight score? What can I do to prevent this?

Answer 1

The include can do a lot of invisible requests (database requests for example) to generate the code. If you serve a page without server side includes, you will get a lower TTFB. In other words: the results you are seeing can be exlained and are as expected.

A common trick to tackle this delay is to use a Static Site Generator. SSG's generate static pages that contain the output of these includes (preventing the extra requests).

Another trick to tackle this is to use a cache layer. A cache layer does (almost) the same as a Static Site Generator. When you cache (the output of) your request you prevent these invisible requests too.