preg_replace need help with expression

Question

This is my code:

$string = '<a href="http://www.mysite.com/test" class="prevlink">&laquo; Previous</a><a href=\'http://www.mysite.com/test/\' class=\'page\'>1</a><span class=\'current\'>2</span><a href=\'http://www.mysite.com/test/page/3/\' class=\'page\'>3</a><a href=\'http://www.mysite.com/test/page/4/\' class=\'page\'>4</a><a href="http://www.mysite.com/test/page/3/" class="nextlink">Next &raquo;</a>';
$string = htmlspecialchars($string, ENT_COMPAT, 'UTF-8');
$string = preg_replace('@(&lt;a).*?(nextlink)@s', '', $string);
    echo $string;

I am trying to remove the last link:

<a href="http://www.mysite.com/test/page/3/" class="nextlink">Next &raquo;</a>';

My current output:

">Next &raquo;</a>

It removes everything from the start. I want it to remove only the one with strpos, is this possible with preg_replace and how? Thanks.

Trying to use regex for html (or xml) can be quite a deal. In your case you say to replace something that starts with .... — George Kastrinis, Aug 31 '11 at 08:29
http://stackoverflow.com/questions/1732348/regex-match-open-tags-except-xhtml-self-contained-tags — , Aug 31 '11 at 08:34
@George `$string = preg_replace('@(<a).*?(nextlink).*?(</a>)@s', '', $string);` The output is empty. Great idea, I think my prex statment is wrong. could you take a look. Thanks — Muazam, Aug 31 '11 at 08:45
You said you want to remove the last link... do you mean you want the final output to be `« Previous 123 4`? — pb149, Aug 31 '11 at 08:53
@Pete171 Correct, that's what I want. The first one. I want to remove `Next »` — Muazam, Aug 31 '11 at 08:53
@Muazam Is the final `..` tag always going to be at the end of a string? — pb149, Aug 31 '11 at 09:12
@pete171 Solved. Thanks anyway, if you want to show your solution then please do =) — Muazam, Aug 31 '11 at 09:32
@Pete171 bumperbox posted a solution below. Thanks for your effort =) — Muazam, Aug 31 '11 at 09:42

score 1 · Answer 1 · answered Aug 31 '11 at 08:39

Note: This is not a direct answer, but a suggestion to another approach.

I was told once; if you can do it in any other way, stay away from regex. I don't though, it's my white whale. Have you heard of phpQuery? It's jQuery implemented in PHP and very powerful. It would be able to do what you want in a very easy way. I know it's not regex, but perhaps it's of use to you.

If you really want to go ahead, I can recommend http://gskinner.com/RegExr/ . I think it's a great tool.

bumperbox · Accepted Answer · 2011-08-31T22:36:08.617

quite a tricky question to solve

first off, the .*? will not match like you are expecting it to.

its starts from the left finds the first match for <a, then searches until it finds nextlink, which is essentially picking up the entire string.

for that regex to work as you wanted, it would need to match from the righthand side first and work backwards through the string, finding the smallest (non-greedy) match

i couldn't see any modifiers that would do this so i opted for a callback on each link, that will check and remove any link with nextlink in it

<?php
$string = '<a href="http://www.mysite.com/test" class="prevlink">&laquo; Previous</a><a href=\'http://www.mysite.com/test/\' class=\'page\'>1</a><span class=\'current\'>2</span><a href=\'http://www.mysite.com/test/page/3/\' class=\'page\'>3</a><a href=\'http://www.mysite.com/test/page/4/\' class=\'page\'>4</a><a href="http://www.mysite.com/test/page/3/" class="nextlink">Next &raquo;</a>';

echo "RAW: $string\r\n\r\n";

$string = htmlspecialchars($string, ENT_COMPAT, 'UTF-8');

echo "SRC: $string\r\n\r\n";


    $string = preg_replace_callback(
        '@&lt\;a.+?&lt;/a&gt;@',
        'remove_nextlink',
        $string
    );


function remove_nextlink($matches) {

    // if you want to see each line as it works, uncomment this
    // echo "L: $matches[0]\r\n\r\n";

    if (strpos($matches[0], 'nextlink') === FALSE) {
        return $matches[0]; // doesn't contain nextlink, put original string back
    } else {
        return ''; // contains nextlink, replace with blank
    }
}    

echo "PROCESSED: $string\r\n\r\n";

Thanks for the explanation, I have to learn pregEx it seems very powerful. The code works, thanks a lot. =) — Muazam, Aug 31 '11 at 09:40
it's not the most elegant solution, but it does work, and writing regex to work on html is always hard work — bumperbox, Aug 31 '11 at 09:43

preg_replace need help with expression

2 Answers2