Replace portion of text string

[gamefreak13]

I know how to do it, but I am not happy with my method because it may not work every time.

 

<?php

$input = 'http://profile.myspace.com/index.cfm?fuseaction=user.viewprofile&friendid=00000&MyToken=f6e43fc-c41a-4ebb-81bd-68f091107868';

echo substr_replace($input,"",-45);  

?>

 

This obviously does what I want it to do. It removes the "&MyToken....." part. However, if there were to ever be anything after it, then it would strip incorrectly.

 

This also works. Preg_replace.. but the problem with this is that it strips everything after "&MyToken...". I think this is the proper code but I need the proper regex.

 

<?php

$str = 'http://profile.myspace.com/index.cfm?fuseaction=user.viewprofile&friendid=00000&MyToken=f6e43fc-c41a-4ebb-81bd-68f091107868&foo=bar';

$reg = '%&MyToken=(.+)%';

$str = preg_replace($reg,"",$str);

echo $str;

?>

 

My goal is to remove only the "&MyToken=f6e43fc-c41a-4ebb-81bd-68f091107868" part without risking removing other important content.

[thebadbad]

This should do the trick (using a positive lookahead):

 

<?php

$str = 'http://profile.myspace.com/index.cfm?fuseaction=user.viewprofile&friendid=00000&MyToken=f6e43fc-c41a-4ebb-81bd-68f091107868&foo=bar';

$reg = '%&MyToken=(.*?)(?=&)%';

$str = preg_replace($reg,"",$str);

echo $str;

?>

[gamefreak13]

Works great, but I have to ask.. does the (?-&) part say "stop at the next ampersand" ? Just wondering.. I like to understand the code I work with.

 

$reg = '%&MyToken=(.*?)(?=&)%';

 

Thanks!

 

EDIT: Uh oh.. the above code only works if there actually is something after the MyToken part (there is 99% chance not going to be). It's good that we have both codes for both occasions, but I need them both to work. Is there a way to say "stop at next ampersand or at end"?

[thebadbad]

Oh sorry, didn't think of that.

 

something(?=&) matches "something" ending with an &, without matching the & part. So it removes the string up until the &foo=bar, while still keeping &foo=bar.

 

This should work:

 

$reg = '%&MyToken=(.*?)((?=&)|$)%';

 

Swapped (?=&) with ((?=&)|$) meaning it will either do the (?=&) or $ (which means end of string). I.e. if an ampersand is found, it stops there, but if the end of the string is found (= no ampersand), that's alright too

[gamefreak13]

Works great in all variations! Thanks!

 

<?php

// Different variations

$str1 = 'http://profile.myspace.com/index.cfm?fuseaction=user.viewprofile&friendid=00000&MyToken=f6e43fc-c41a-4ebb-81bd-68f091107868&foo=bar';
$str2 = 'http://profile.myspace.com/index.cfm?fuseaction=user.viewprofile&friendid=00000&MyToken=f6e43fc-c41a-4ebb-81bd-68f091107868';
$str3 = 'http://profile.myspace.com/index.cfm?fuseaction=user.viewprofile&friendid=00000&foo=bar';
$str4 = 'http://profile.myspace.com/index.cfm?fuseaction=user.viewprofile&friendid=00000';


echo preg_replace("%&MyToken=(.*?)((?=&)|$)%","",$str1);


echo "<hr>";


echo preg_replace("%&MyToken=(.*?)((?=&)|$)%","",$str2);


echo "<hr>";


echo preg_replace("%&MyToken=(.*?)((?=&)|$)%","",$str3);


echo "<hr>";


echo preg_replace("%&MyToken=(.*?)((?=&)|$)%","",$str4);


?>