Analysis of PHP keyword replacement classes (avoid repeated replacement, retain and restore original links)

A keyword replacement class, which is mainly used for keyword filtering or keyword search Replacement. keyword replacement is actually a str_replace () process, if you are interested, let's take a look at the php keyword replacement class (avoid repeated replacement, retain and restore the original link)

The main content of this section:

A keyword replacement class

It can be used for keyword filtering or keyword search replacement.

Implementation process analysis:

Keyword replacement is actually a str_replace () process. if it is a simple str_replace, it only takes about 2 seconds for A-word article.

Problem:

The keyword is replaced more than once. for example, a must be replaced with a, but the result may be.

Therefore, a method is required to protect the replaced tags. before processing the article, replace the label first, for example, [_ tnum _], and then restore it after the article has been processed.

Another problem is that if [_ tnum _] exists in a keyword or article, you need to exclude this. in this case, str_replace cannot be used, but preg_replace must be used for regular expression elimination.

The third question is what to do if there are two keywords a and AB. we hope to first match the long one and then the short one, so we need to sort it before matching.

In the last question, after str_replace is changed to preg_replace, it takes 5 seconds to match the same string for 10 times and strpos is faster in the string processing function. Then, use strpos to find the keyword, less than 1 second for 10 million queries. Even if it is 1 million, it takes more than 8 seconds.

A keyword matching replacement class, code:

Sample code:

<? Php/** keyword matching class * @ author ylx
 
  
* @ Packet mipang * use an instance * $ str = "The green shell egg, zafdensa, will pop up next year to open the room Lucas local Army"; * $ key = new KeyReplace ($ str, array ("xxxx" => "sadf", "next year" =>' http://baidu.com ', "Next year" =>' google. com '); * echo $ key-> getResultText (); * echo $ key-> getRuntime (); */class KeyReplace {private $ keys = array (); private $ text = ""; private $ runtime =; private $ url = true; private $ stopkeys = array (); private $ all = false; /*** @ access public * @ param string $ text specifies the article to be processed * @ param array $ keys specifies the dictionary phrase array (key => url ,...) the url can be an array. if it is an array, a * @ param array $ stopkeys of the specified stop word array (k Ey ,...) here, the words will not be processed * @ param boolean $ url true indicates replacing them with links; otherwise, only * @ param boolean $ all true indicates replacing all words found, otherwise, only replace the first time */public function _ construct ($ text = '', $ keys = array (), $ url = true, $ stopkeys = array (), $ all = false) {$ this-> keys = $ keys; $ this-> text = $ text; $ this-> url = $ url; $ this-> stopkeys = $ stopkeys; $ this-> all = $ all ;} /*** get the processed article * @ access public * @ return string text */public function getResultText () {$ Start = microtime (true); $ keys = $ this-> hits_keys (); $ keys_tmp = array_keys () ($ keys); function cmp ($, $ B) {if (mb_strlen ($ a) = mb_strlen ($ B) {return;} return (mb_strlen ($ a) <mb_strlen ($ B ))? :-;} Usort ($ keys_tmp, "cmp"); foreach ($ keys_tmp as $ key) {if (is_array ($ keys [$ key]) {$ url = $ keys [$ key] [rand (, count ($ keys [$ key])-)];} else $ url = $ keys [$ key]; $ this-> text = $ this-> r_s ($ this-> text, $ key, $ url);} $ this-> runtime = microtime (true)-$ start; return $ this-> text;}/*** get processing time * @ access public * @ return float */public function getRuntime () {return $ this-> runtime ;} /*** set the keyword ** @ access Public * @ param array $ keys array (key => url ,...) */public function setKeys ($ keys) {$ this-> keys = $ keys ;} /*** set the stop word * @ access public * @ param array $ keys array (key ,...) */public function setStopKeys ($ keys) {$ this-> stopkeys = $ keys ;} /*** set the article * @ access public * @ param string $ text */public function setText ($ text) {$ this-> text = $ text ;} /*** is used to find the keyword hit in the string * @ access public * @ return ar Ray $ keys returns the matched word array (key => url ,...) */public function hits_keys () {$ ar = $ this-> keys; $ ar = $ ar? $ Ar: array (); $ result = array (); $ str = $ this-> text; foreach ($ ar as $ k => $ url) {$ k = trim ($ k); if (! $ K) continue; if (strpos ($ str, $ k )! = False &&! In_array ($ k, $ this-> stopkeys) {$ result [$ k] = $ url;} return $ result? $ Result: array ();} /*** is used to find the stop word hit in the string * @ access public * @ return array $ keys returns the matched word array (key ,...) */public function hits_stop_keys () {$ ar = $ this-> stopkeys; $ ar = $ ar? $ Ar: array (); $ result = array (); $ str = $ this-> text; foreach ($ ar as $ k) {$ k = trim ($ k); if (! $ K) continue; if (strpos ($ str, $ k )! = False & in_array ($ k, $ this-> stopkeys) {$ result [] = $ k ;}} return $ result? $ Result: array ();} /*** process the replacement process ** @ access private * @ param string $ text replaced by * @ param string $ key keyword * @ param string $ url link * @ return string $ text processed article */private function r_s ($ text, $ key, $ url) {$ tmp = $ text; $ stop_keys = $ this-> hits_stop_keys (); $ stopkeys = $ tags = $ a = array (); if (preg_match_all ("#] +> [^ <] *] *> # Su ", $ tmp, $ m) {$ a = $ m []; foreach ($ m [] as $ k => $ z) {$ z = preg_replace ("#\## s", "\#", $ z); $ tmp = preg_replace ('#'. $ z. '# s', "[_ ". $ k. "_]", $ tmp,) ;}}; if (preg_match_all ("# <[^>] +> # s", $ tmp, $ m )) {$ tags = $ m []; foreach ($ m [] as $ k => $ z) {$ z = preg_replace ("#\## s ", "\ #", $ z); $ tmp = preg_replace ('#'. $ z. '# s', "[_ tag ". $ k. "_]", $ tmp,) ;}} if (! Empty ($ stop_keys) {if (preg_match_all ("#". implode ("|", $ stop_keys ). "# s", $ tmp, $ m) {$ stopkeys = $ m []; foreach ($ m [] as $ k => $ z) {$ z = preg_replace ("#\## s", "\#", $ z); $ tmp = preg_replace ('#'. $ z. '# s', "[_ s ". $ k. "_]", $ tmp,) ;}}$ key = preg_replace ("# ([# \ (\) \ [\] \ *]) # s ", "\\\\$", $ key); if ($ this-> url) $ tmp = preg_replace ("#(?! \ [_ S | \ [_ a | \ [_ | \ [_ t | \ [_ ta | \ [_ tag) ". $ key ."(?! Ag \ d + _ \] | g \ d + _ \] | s \ d + _ \] | _ \]) # us ", ''. $ key. '', $ tmp, $ this-> all? -:); Else $ tmp = preg_replace ("#(?! \ [_ S | \ [_ a | \ [_ | \ [_ t | \ [_ ta | \ [_ tag) ". $ key ."(?! Ag \ d + _ \] | g \ d + _ \] | s \ d + _ \] | _ \]) # us ", $ url, $ tmp, $ this-> all? -:); If (! Empty ($ a) {foreach ($ a as $ n = >$ at) {$ tmp = str_replace ("[_ ". $ n. "_]", $ at, $ tmp) ;}} if (! Empty ($ tags) {foreach ($ tags as $ n =>at at) {$ tmp = str_replace ("[_ tag ". $ n. "_]", $ at, $ tmp) ;}} if (! Empty ($ stopkeys) {foreach ($ stopkeys as $ n = >$ at) {$ tmp = str_replace ("[_ s ". $ n. "_]", $ at, $ tmp) ;}return $ tmp ;}}
 

The above is the PHP keyword replacement class introduced in this article (avoid repeated replacement, retain and restore the original link ).