{"id":536,"date":"2006-05-05T21:48:59","date_gmt":"2006-05-05T19:48:59","guid":{"rendered":"http:\/\/frenchfragfactory.net\/ozh\/archives\/2006\/05\/05\/speeding-things-with-php-output-buffer\/"},"modified":"2007-05-08T15:54:55","modified_gmt":"2007-05-08T13:54:55","slug":"speeding-things-with-php-output-buffer","status":"publish","type":"post","link":"https:\/\/planetozh.com\/blog\/2006\/05\/speeding-things-with-php-output-buffer\/","title":{"rendered":"Speeding Things with PHP Output Buffer"},"content":{"rendered":"

While playing a bit with Xdebug<\/a> (a profiling tool for PHP), I've come across something I didn't know and that I frankly found a bit odd : how to use output buffering (PHP function ob_start()<\/em>) to speed up some stuff.<\/p>\n

In a few words, just so you know if this is obvious noob stuff for you and you can skip this article : <\/p>\n

\n\n\t\t
\n\n\t\t
\n\t\t\t\t\t\t< View plain text<\/span> ><\/a>\n\t\t\t\t\t<\/div>\n\n\t\t
php<\/div>\n\n\t\t\n\t\t
\n\n\t<\/div>\n\t\n\t
\n\t\t
  1. ob_start<\/span>(<\/span>)<\/span>;<\/span><\/div><\/li>\n
  2. echo<\/span> a lot of stuff;<\/span><\/div><\/li>\n
  3. $result<\/span> =<\/span> ob_get_contents<\/span>(<\/span>)<\/span>;<\/span><\/div><\/li>\n<\/ol>\t<\/div>\n\n<\/div>\n\n

    can be a lot faster than a straight<\/p>\n

    \n\n\t\t
    \n\n\t\t
    \n\t\t\t\t\t\t< View plain text<\/span> ><\/a>\n\t\t\t\t\t<\/div>\n\n\t\t
    php<\/div>\n\n\t\t\n\t\t
    \n\n\t<\/div>\n\t\n\t
    \n\t\t
    1. $result<\/span> =<\/span> a lot of stuff<\/div><\/li>\n<\/ol>\t<\/div>\n\n<\/div>\n\n

      <\/p>\n

      Say you have a very complex routine that errrr… prints a string with a lot of 'a'.<\/p>\n

      Straight and simple code would be :<\/p>\n

      \n\n\t\t
      \n\n\t\t
      \n\t\t\t\t\t\t< View plain text<\/span> ><\/a>\n\t\t\t\t\t<\/div>\n\n\t\t
      php<\/div>\n\n\t\t\n\t\t
      \n\n\t<\/div>\n\t\n\t
      \n\t\t
      1. $s<\/span> =<\/span> ''<\/span>;<\/span><\/div><\/li>\n
      2. for<\/span> (<\/span>$i<\/span>=<\/span>0<\/span>;<\/span> $i<\/span> <<\/span> $loops<\/span>;<\/span> $i<\/span>++<\/span>)<\/span> $s<\/span> .=<\/span> 'a'<\/span>;&<\/span>#91;\/php&#93;<\/span><\/div><\/li>\n
      3.  <\/div><\/li>\n
      4. Now,<\/span> with output bufferization you could make it like :<\/span><\/div><\/li>\n
      5. &<\/span>#91;php&#93;$s = '';<\/span><\/div><\/li>\n
      6. ob_start<\/span>(<\/span>)<\/span>;<\/span><\/div><\/li>\n
      7. for<\/span> (<\/span>$i<\/span>=<\/span>0<\/span>;<\/span> $i<\/span> <<\/span> $loops<\/span>;<\/span> $i<\/span>++<\/span>)<\/span> echo<\/span> 'a'<\/span>;<\/span><\/div><\/li>\n
      8. $s<\/span> =<\/span> ob_get_contents<\/span>(<\/span>)<\/span>;<\/span><\/div><\/li>\n
      9. ob_end_clean<\/span>(<\/span>)<\/span>;&<\/span>#91;\/php&#93;<\/span><\/div><\/li>\n
      10.  <\/div><\/li>\n
      11. Let's benchmark a bit these two snippets :<\/span><\/div><\/li>\n
      12.  <\/div><\/li>\n
      13. &#91;php&#93;<?php<\/span><\/div><\/li>\n
      14. $iterations = 50000;<\/span><\/div><\/li>\n
      15.  <\/div><\/li>\n
      16. function straight($loops = 100) {<\/span><\/div><\/li>\n
      17.     for ($i=0; $i < $loops; $i++) $s .= '<\/span>a';<\/span><\/div><\/li>\n
      18.     return $s;<\/span><\/div><\/li>\n
      19. }<\/span><\/div><\/li>\n
      20.  <\/div><\/li>\n
      21. function buffer($loops = 100) {<\/span><\/div><\/li>\n
      22.     ob_start();<\/span><\/div><\/li>\n
      23.     for ($i=0; $i < $loops; $i++) echo '<\/span>a';<\/span><\/div><\/li>\n
      24.     $s = ob_get_contents();<\/span><\/div><\/li>\n
      25.     ob_end_clean();<\/span><\/div><\/li>\n
      26.     return $s;<\/span><\/div><\/li>\n
      27. }<\/span><\/div><\/li>\n
      28.  <\/div><\/li>\n
      29. $lotsofa1 = straight($iterations);<\/span><\/div><\/li>\n
      30. $lotsofa2 = buffer($iterations);<\/span><\/div><\/li>\n
      31. ?><\/span><\/div><\/li>\n<\/ol>\t<\/div>\n\n<\/div>\n\n

        Results<\/strong> : on my test machine, straight code took about 60 ms<\/em> to execute while buffered code took 50 ms<\/em>. Holy beep, the weirdo syntax is 15% faster !? I guess you save time by manipulating once the result variable, when you get the buffer content, but that was quite surprising for my PHP beginner self.<\/p>\n

        While I was at it, I pushed it a bit further and tried to think about interesting ways to use this. I thought : maybe it's faster for example to include()<\/em> a file into a buffer than simply using file_get_content()<\/em> ?<\/p>\n

        To test this, I used the following :<\/p>\n

        \n\n\t\t
        \n\n\t\t
        \n\t\t\t\t\t\t< View plain text<\/span> ><\/a>\n\t\t\t\t\t<\/div>\n\n\t\t
        php<\/div>\n\n\t\t\n\t\t
        \n\n\t<\/div>\n\t\n\t
        \n\t\t
        1. <?php<\/span><\/div><\/li>\n
        2.  <\/div><\/li>\n
        3. function<\/span> bench_readfile(<\/span>$file<\/span>)<\/span> {<\/span><\/div><\/li>\n
        4.     ob_start<\/span>(<\/span>)<\/span>;<\/span><\/div><\/li>\n
        5.     readfile<\/span>(<\/span>$file<\/span>)<\/span>;<\/span><\/div><\/li>\n
        6.     $s<\/span> =<\/span> ob_get_contents<\/span>(<\/span>)<\/span>;<\/span><\/div><\/li>\n
        7.     ob_end_clean<\/span>(<\/span>)<\/span>;<\/span><\/div><\/li>\n
        8.     return<\/span> $s<\/span>;<\/span><\/div><\/li>\n
        9. }<\/span><\/div><\/li>\n
        10.  <\/div><\/li>\n
        11. function<\/span> bench_filegetcontents(<\/span>$file<\/span>)<\/span> {<\/span><\/div><\/li>\n
        12.     return<\/span> file_get_contents<\/span>(<\/span>$file<\/span>)<\/span>;<\/span><\/div><\/li>\n
        13. }<\/span><\/div><\/li>\n
        14.  <\/div><\/li>\n
        15. function<\/span> bench_include(<\/span>$file<\/span>)<\/span> {<\/span><\/div><\/li>\n
        16.     ob_start<\/span>(<\/span>)<\/span>;<\/span><\/div><\/li>\n
        17.     include<\/span>(<\/span>$file<\/span>)<\/span>;<\/span><\/div><\/li>\n
        18.     $s<\/span> =<\/span> ob_get_contents<\/span>(<\/span>)<\/span>;<\/span><\/div><\/li>\n
        19.     ob_end_clean<\/span>(<\/span>)<\/span>;<\/span><\/div><\/li>\n
        20.     return<\/span> $s<\/span>;<\/span><\/div><\/li>\n
        21. }<\/span><\/div><\/li>\n
        22.  <\/div><\/li>\n
        23. $content1<\/span> =<\/span> bench_readfile(<\/span>'file1.txt'<\/span>)<\/span>;<\/span><\/div><\/li>\n
        24. $content2<\/span> =<\/span> bench_filegetcontents(<\/span>'file2.txt'<\/span>)<\/span>;<\/span><\/div><\/li>\n
        25. $content3<\/span> =<\/span> bench_include(<\/span>'file3.txt'<\/span>)<\/span>;<\/span><\/div><\/li>\n
        26.  <\/div><\/li>\n
        27. ?><\/span><\/div><\/li>\n<\/ol>\t<\/div>\n\n<\/div>\n\n

          Files file1.txt, file2.txt and file3.txt were copies of the same file, but I wanted to be sure that no function would benefit from cached data.<\/p>\n

          Results<\/strong> : the include() block took 51 ms<\/em> to execute, readfile()<\/em> took 7.3 ms and file_get_contents()<\/em> won it in just 2 ms.<\/p>\n

          Conclusion<\/strong> : Well, no surprise here. Hopefully built-in function are sometimes optimized enough :\u00c3\u017e<\/p>\n","protected":false},"excerpt":{"rendered":"

          While playing a bit with Xdebug (a profiling tool for PHP), I've come across something I didn't know and that I frankly found a bit odd : how to use output buffering (PHP function ob_start()) to speed up some stuff. In a few words, just so you know if this is obvious noob stuff for […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[],"tags":[2,10,74,164],"class_list":["post-536","post","type-post","status-publish","format-standard","hentry","tag-code","tag-php","tag-stats","tag-xdebug"],"_links":{"self":[{"href":"https:\/\/planetozh.com\/blog\/wp-json\/wp\/v2\/posts\/536","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/planetozh.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/planetozh.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/planetozh.com\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/planetozh.com\/blog\/wp-json\/wp\/v2\/comments?post=536"}],"version-history":[{"count":0,"href":"https:\/\/planetozh.com\/blog\/wp-json\/wp\/v2\/posts\/536\/revisions"}],"wp:attachment":[{"href":"https:\/\/planetozh.com\/blog\/wp-json\/wp\/v2\/media?parent=536"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/planetozh.com\/blog\/wp-json\/wp\/v2\/categories?post=536"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/planetozh.com\/blog\/wp-json\/wp\/v2\/tags?post=536"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}