Edit Report a Bug

eregi_replace

(PHP 4, PHP 5)

eregi_replace — Remplacement par expression rationnelle insensible à la casse

Avertissement

Cette fonction est OBSOLÈTE à partir de PHP 5.3.0 et a été SUPPRIMÉE à partir de PHP 7.0.0.

Les alternatives à cette fonction incluent :

preg_replace() (avec le modificateur i (PCRE_CASELESS))

Description

eregi_replace ( string $pattern , string $replacement , string $string ) : string

Cette fonction est identique à ereg_replace(), hormis le fait qu'elle ne tient pas compte de la casse des caractères alphabétiques.

Liste de paramètres

pattern: Une expression rationnelle POSIX.
replacement: Si pattern contient des sous-masques mis entre parenthèses, replacement contiendra des sous-chaînes, sous la forme \digit, qui seront remplacées par le texte correspondant au sous-masque mis entre parenthèses ; \0 correspondra au contenu complet de la chaîne. Il est possible d'utiliser jusqu'à 9 sous-chaînes. Les parenthèses peuvent être imbriquées et seront comptées à partir de la parenthèse ouvrante.
string: La chaîne d'entrée.

Valeurs de retour

La chaîne modifiée est retournée. Si aucune correspondance n'est trouvée dans string, alors elle sera retournée sans aucune modification.

Exemples

Exemple #1 Mise en évidence des résultats de la recherche


<?php
$pattern = '(>[^<]*)('. quotemeta($_GET['search']) .')';
$replacement = '\\1<span class="search">\\2</span>';
$body = eregi_replace($pattern, $replacement, $body);
?>

Voir aussi

ereg() - Recherche par expression rationnelle standard
eregi() - Recherche par expression rationnelle insensible à la casse
ereg_replace() - >Remplacement par expression rationnelle
preg_replace() - Rechercher et remplacer par expression rationnelle standard
quotemeta() - Protège les métacaractères

add a note

User Contributed Notes 31 notes

down

carlos at braziland dot com ¶

18 years ago


@bluedragonx at gmail dot com: Now you're calling two functions instead of one. I'm trully curious if that would still be faster. 

My original post was just to highlight the need to escape characters that could be used to abuse a contact form. Something everyone should do, especially if doing it for a customer.

But bluedragonx, thanks for your input.

down

Anonymous ¶

19 years ago


@bluedragonx: You're right. And i got the order of the params wrong as well, must've been sleepy when i wrote that note. For what it's worth:

<? $_POST['email'] = str_replace("[at]", "@", strtolower($_POST['email'])); ?>

Not similar to Carlos' code though, since it'll convert all characters to lowercase.

down

Anonymous ¶

19 years ago


never mind my last post for the eregi_replace not replacing.

I just used str_replace instead and it works fine.  I must have had something wrong with my search string.  POSIX, Perl.. hmm.. yeah probably something there.

Mettedraq / Gene

down

jgriggs777 at yahoo dot com ¶

16 years ago


My apologies.  I'm a PHP newbie and the count function didn't work right if you have more than one "keyword".  I searched and spent a long time to get this to work right, so wanted to share it.

This is very easy to use by doing the following.

$text - is the text to search
$words - are the words to highlight (search text)
$the_place - is so that you can tell your users what "area" was searched.

<?php 

function highlight_this($text, $words, $the_place) { 
    $words = trim($words);
    $the_count = 0; 
    $wordsArray = explode(' ', $words); 
        foreach($wordsArray as $word) { 
         if(strlen(trim($word)) != 0)
         
         //exclude these words from being replaced
         $exclude_list = array("word1", "word2", "word3");
        // Check if it's excluded
        if ( in_array( strtolower($word), $exclude_list ) ) {
    
        } else {
            $text = str_ireplace($word, "<span class=\"highlight\">".strtoupper($word)."</span>", $text, $count);
            $the_count = $count + $the_count;
            }
            
    } 
    //added to show how many keywords were found
    echo "<br><div class=\"emphasis\">A search for <strong>" . $words. "</strong> found <strong>" . $the_count . "</strong> matches within the " . $the_place. ".</div><br>";
    
    return $text;
} 

//example of how to use
$text_to_highlight = highlight_this($text_to_highlight, $search_text, "Place you searched");

?>

down

php dot net at nr78 dot net ¶

19 years ago


@carlos at braziland dot com: Hate to be a wisecrack, but i believe the following code might be slightly faster with the same result:

<?php

$_POST['email'] = str_replace("@", "[at]", $_POST['email']);

?>

down

carlos at braziland dot com ¶

19 years ago


So here is my answer to those who are trying to submit headers through my simple contact form:

$any_form_field = eregi_replace("@","[at]",$_POST['just_one_field']); //how can anyone send emails to other email addresses without the @ symbol :)

My suggestion is to do a foreach loop in the $_POST array and remove the @ symbol from all fields (unless you want one of the fields to have the @ symbol, in which case it should be very careful and question if the symbol is necessary).

Not the best solution, but it's simple and it works.

down

chpins-php at romu92 dot freesurf dot fr ¶

21 years ago


/*new function for href2text : */

function AHREF2text($string) {
 return eregi_replace('(<a [^<]*href=["|\']?([^ "\']*)["|\']?[^>].*>([^<]*)</a>)','[\\3] (link: \\2)', $string);
}

// by Ch'Pins

down

simon_a99 at yahoo dot co dot uk ¶

22 years ago


To find a string regardless of case, you might want to use the matched string in the replacement string without changing its case.

For example, you're searching for $search = "letter" and the text being searched is $text = "post lEtTeR".  I want to change the format of the matched string.

Do this:
eregi_replace($search, "<b>\\0</b>", $text);

$text has now been changed to "post <b>lEtTeR</b>".  \\0 is the entire text matched (lEtTer in this case).

down

dan at dotgospel dot com ¶

16 years ago


/*******************************************************
Function highlights the text. You can replace the html tags with whatever you like. $x is the string, $var is the string to be highlighted. We explode the string to highlight more than one word.
*/

function highlight($x,$var) {
    $var = explode(" ",$var);
    
   for($j=0; $j< count($var); $j++){    
        $xtemp = "";
        $i=0;
        while($i<strlen($x)){
            if((($i + strlen($var[$j])) <= strlen($x)) && (strcasecmp($var[$j], substr($x, $i, strlen($var[$j]))) == 0)) {
                    $xtemp .= "<b>" . substr($x, $i , strlen($var[$j])) . "</b>";
                    $i += strlen($var[$j]);
            }
            else {
                $xtemp .= $x{$i};
                $i++;
            }
        }
        $x = $xtemp;
    }
    return $x;
}

down

augusto at fagioli dot biz ¶

16 years ago


related to the private at sfiac dot net function,
please note that you  can add a dot in the 

$c=ereg_replace".
   "("[-a-z0-9!#$%&\'*+/=?^_`{|}~]+@([.]?[a-zA-Z0-9_/-])*",".
   "$e,$c); 

having it to find email address like user.name@test.com

this is my example
$c=ereg_replace".
   "("[-a-z0-9.!#$%&\'*+/=?^_`{|}~]+@([.]?[a-zA-Z0-9_/-])*",".
   "$e,$c);

down

spiritualmind at gmail dot com ¶

17 years ago


eregi_replace seems not to deal with UTF8 chars !
I needed to utf8_decode / encode my string to parse it in eregi_replace :

<?php
  $input = "string_from_utf8_form" ;
  $output = utf8_encode(eregi_replace("(pattern)", "replacement", utf8_decode($input)) ;

 echo $output ;
?>

I think UTF8 is not totally supported by PHP.

down

Vladimir Luchaninov ¶

18 years ago


If you have plain text e-mails and links but need to make them real links

<?
function replaceLinks($text) {
    // convert support@pogoda.in into
    // <a href="mailto:support@pogoda.in">
    // support@pogoda.in</a>
    $text = ereg_replace('[-a-z0-9!#$%&\'*+/=?^_`{|}~]+@([.]?[a-zA-Z0-9_/-])*',
        '<a href="mailto:\\0">\\0</a>',$text);

    // convert http://www.pogoda.in/new_york/eng/ into
    // <a href="http://pogoda.in/new_york/eng/">
    // pogoda.in/new_york/eng/</a>
    $text = ereg_replace('[a-zA-Z]+://(([.]?[a-zA-Z0-9_/-])*)',
        '<a href="\\0">\\1</a>',$text);

    // convert www.pogoda.in/new_york/eng/ into
    // <a href="http://www.pogoda.in/new_york/eng/">
    // www.pogoda.in/new_york/eng/</a>
    $text = ereg_replace('(^| )(www([-]*[.]?[a-zA-Z0-9_/-?&%])*)',
        ' <a href="http://\\2">\\2</a>',$text);
    
    return $text;
}
?>

down

melissa at melissagirl dot com ¶

18 years ago


Here's a nice case-insensitive highlight function that ignores any regular expression characters and highlights a word and leaves it in whatever case it was before:

<?
function highlight($word, $subject) {
    $regex_chars = '\.+?(){}[]^$';
    for ($i=0; $i<strlen($regex_chars); $i++) {
        $char = substr($regex_chars, $i, 1);
        $word = str_replace($char, '\\'.$char, $word);
    }
    $word = '(.*)('.$word.')(.*)';
    return eregi_replace($word, '\1<span class="highlight">\2</span>\3', $subject);
}
?>

down

bluedragonx at gmail dot com ¶

19 years ago


Actually that wouldn't work because str_replace is case sensitive.  So any instances of [AT] [aT] or [At] would be ignored.

down

joltmans at amersel dot com ¶

19 years ago


PHP's Regex engine differs from several others in its treatment of parsing spaces. 

In many Regex languages '\s' denotes a space.
PHP does not recognize '\s', just type a space ' ' instead.

This simple example illustrates the problem:
<?php
    $string = "A sentence with   spaces";
    if (eregi("with\s*spaces", $string))
    {
        // Will never print
        echo "PHP understood \s";
    }
    else
    {
        // Will always print
        echo "PHP doesn't understand \s";
    }
?>

This example does work:
<?php
    $string = "A sentence with   spaces";
    if (eregi("with *spaces", $string))
    {
        // Will print
        echo "PHP understood ' '";
    }
?>

down

Ausvald ¶

20 years ago


Zach Johnson missed up. ereg* funcs use posix regex, not the rfc one

down

e dot boeters at planet dot nl ¶

20 years ago


This is a 'faster' way to highlight search results: 

$content = str_replace($query, "<span class=\"highlight\">" . $query . "</span>", $content);

down

julien at ratatouille dot com dot fr ¶

21 years ago


This function replace < and > symbols between <code> and </code> tags by html code for lower than (&lt;) and greather than (&gt;) elements.

function retourne_format_code($texte)
{
    $tablo=split("<code>",$texte);
    $texte="";
    $texte.=$tablo[0];
    foreach($tablo as $cle=>$valeur)
    {
        if(eregi("</code>",$valeur))
        {
            $tablo1=split("</code>",$valeur);
            $tablo1[0]=eregi_replace("<","&lt;",$tablo1[0]);
            $tablo1[0]=eregi_replace(">","&gt;",$tablo1[0]);
            foreach($tablo1 as $cle1=>$valeur1)
            {
                if($cle1==0)
                    $valeur1="<code>".$valeur1."</code>";
                $texte.=$valeur1;
            }
        }        
    }
    return $texte;
}

down

martin_goldmann at gmx dot net ¶

21 years ago


After reading the last message I wrote that de-spamizer:

function despamMailURI($myStrMail='')
{
    ?>javascript:void(location.href='mailto:'+<?=eregi_replace( "^([_\.0-9a-z-]+)@([0-9a-z][0-9a-z-]+)\.([a-z]{2,6})$", "'\\1'+'@'+'\\2'+'.'+'\\3'",$myStrMail)?>)<?
}

Usage example:
<a href="<?=despamMailURI("user@foo.bar")?>">Mail me</a>

down

iparanoid at gmx dot de ¶

21 years ago


To obtain the an email addresse in the scheme user at host dot use following function

function antispam_mail($mail) {
    return eregi_replace( "^([_\.0-9a-z-]+)@([0-9a-z][0-9a-z-]+)\.([a-z]{2,6})$", '\\1 at \\2 dot \\3',$mail );
};

Combined with a wee JavaScript (document.location='mailto:'+user+'@'+host+'.'+tld) this provides a very powerfull anti-spam mechanism while providing full mailto: link functionality.

down

buddy at directpay dot cz ¶

21 years ago


i had to solve problem conserning DB2 timestamp format. here is how to parse ANSI timestamp format to DB2 timestamp format:

$mydate = Date("Y-m-d H:i:s");   
  
$var = eregi_replace 
("([0-9]{4}-[0-9]{2}-[0-9]{2}) ([0-9]{2}):([0-9]{2}):([0-9]{2})"
,"\\1-\\2.\\3.\\4",$mydate);

echo "ANSI: $mydate, DB2 format: $var";

happy codding 
buddy

down

eder at xce dot de ¶

21 years ago


/*As php at silisoftware dot com's example works only if there is not more than one link in $string, I rewrote his expression to function with strings containing multiple links: */

function AHREF2text($string) {
  return eregi_replace('(<a [^<]*href=["|\']?([^ "\']*)["|\']?[^>]*>([^<]*)</a>)','[\\3] (link: \\2)', $string);
}

down

-1

dave at bevhost dot com ¶

12 years ago


The easiest function to replace this with is mb_eregi_replace(),
which is the mutibyte version that works on unicode characters such as chinese letters in your strings.

down

-1

jgriggs777 at yahoo dot com ¶

16 years ago


I modified the script from the note below this one so that you could show how many matches were found.  I know this should probably be posted in the str_ireplace() area, so I will post it there afterwards.

$text - is the text to search
$words - are the words to highlight (search text)
$the_place - is so that you can tell your users what "area" was searched.

<?php 
function highlight_this($text, $words, $the_place) { 
    $words = trim($words); 
    $wordsArray = explode(' ', $words); 
    foreach($wordsArray as $word) { 
        if(strlen(trim($word)) != 0) 
            $text = str_ireplace($word, "<span class=\"highlight\">".strtoupper($word)."</span>", $text, $count); 
    } 
    //added to show how many keywords were found
    echo "<br><div class=\"emphasis\">A search for <strong>" . $words. "</strong> found <strong>" . $count . "</strong> matches within the " . $the_place. ".</div><br>";
    
    //end script modification
    return $text;
} 
?>

down

-1

webmaster at hhgroups dot com ¶

16 years ago


Your remove links functions are incorrect if there are a php var in the URL, we have to add '?' character inside our pattern, I have three this function working correctly with php vars:

function transformUrl($str){
    $str=utf8_decode(urldecode($str));
    $str=eregi_replace("(^| |>)(www([.]?[a-zA-Z0-9_/-?])[^<]*)", "\\1<a href=\"http://\\2\">Look link</a>", $str);
    $str=eregi_replace("(^| |>)(http://www([.]?[a-zA-Z0-9_/-?])[^<]*)", "\\1<a href=\"\\2\">Look link</a>", $str);
    $str=eregi_replace("(^| |>)(http://([.]?[a-zA-Z0-9_/-?])[^<]*)", "\\1<a href=\"\\2\">Look link</a>", $str);
    return utf8_encode($str);
}

down

-1

furiousgreencloud at yahoo dot com ¶

20 years ago


To simply convert wild input into a sensable sting, say for a filename I use:

function cleanString($wild) {
    return ereg_replace("[^[:alnum:]+]","",$wild);
}
                                                                                
echo cleanString("@#$&*$@#H~e'{}l{}l<o\{}"); // outputs: Hello

down

-1

Rainmaker526 at hotmail.com ¶

21 years ago


I have found that some characters cannot be used by eregi_replace (or ereg_replace). When you get the REG_BADRPT error, try backslashing any special chars in your pattern string

ex.
$str = eregi_replace("*", "", $somevar)

gives the REG_BADRPT error. Change it to
$str = eregi_replace("\*", "", $somevar)

to make it work

down

-1

php at silisoftware dot com ¶

21 years ago


Transform HTML links into plain-text "links" with the URL visible

function AHREF2text($string) {
    return eregi_replace('<A .*HREF=("|\')?([^ "\']*)("|\')?.*>([^<]*)</A>', '[\\4] (link: \\2)', $string);
}

$HTMLstring = 'A link to <a href="http://www.php.net">PHP.net</A>';
echo AHREF2text($HTMLstring);
// prints:  A link to [PHP.net] (link: http://www.php.net)

down

-2

mp3spl at o2 dot pl ¶

16 years ago


Nice way to highlight that what someone search.

For example is some one search for "Wire Mobile 6"

It will highlight every word separated even if found "Wire Xp and some mobile stuff"

© and ® are importiant, because without special signs, string in second turn will convert also <span class="searchtext">
which will be already in text.

<?

$searchcolor = explode(' ',$search);
$ilosc = count($searchcolor);
$ilosc2 = "0";
while ($ilosc2 != $ilosc ) {
$name = eregi_replace ($searchcolor[$ilosc2],'©'.$searchcolor[$ilosc2].'®',$name);
$ilosc2++;
        }
$name = str_replace('©','<span class="searchtext">',$name);
$name = str_replace('®','</span>',$name);

echo $name;

?>

down

-2

php[at]drixe[dot]net ¶

17 years ago


Melissa,
This one is a better setup for your marvelous code.
It handles Melissa Magic and also Magic Melissa

function underline($subject, $word)
        {
        $mywords = explode(" ",$word);
        for ($j=0;$j<count($mywords);$j++)
            {
            $regex_chars = '\.+?(){}[]^$';
            for ($i=0; $i<strlen($regex_chars); $i++)
                {
                $char = substr($regex_chars, $i, 1);
                $mywords[$j] = str_replace($char, '\\'.$char, $mywords[$j]);
                }
            $mywords[$j] = '(.*)('.$mywords[$j].')(.*)';
            $subject =  eregi_replace($mywords[$j], '\1<span style="background:#ACA;padding:0;margin:0;">\2</span>\3', $subject);
            }
        return $subject;
        }

down

-5

a dot dotreppe at gmail dot com ¶

16 years ago


Hi everyone.

I saw a lot of highlighting functions, but no one was simple... maybe I missed a part of discussion ?



Anyway, I think this function is the easiest:



<?php

function highlight($text, $words) {

    $words = trim($words);

    $wordsArray = explode(' ', $words);

    foreach($wordsArray as $word) {

        if(strlen(trim($word)) != 0)

            $text = eregi_replace($word, '<span class="highlight">\\0</span>', $text);

    }

    return $text;

}

?>

add a note