preg_replace

Worker Joined: Dec 03, 2005 Posts: 153

Hi All

I have an input box for members name, it works ok

I have the following code

Code:

 $mnane = preg_replace("/[^a-zA-Z0-9\s]/", "", $mnane);

It work ok, till someone puts a members name in with underscore _ within it, ie. com_thor the code removes the underscore and joins the name together.

I also have a Location input, which is 40 Characters, when we put in a location then use a comma to seperate town name with area ie. birmingham, west midlands it removes the comma.

How can I allow the use of underscore and comma's within the code

Cheers
Scorp

Posted: Fri Jan 20, 2012 3:02 pm

I guess if you must use \\ for all your caracters you wanna add to show username with underscore. Try this one.

Code:

$mnane = preg_replace("/[^a-zA-Z0-9_-\\[\\]\\{\\}\\=-]/", "", $mnane);

Posted: Fri Jan 20, 2012 5:00 pm

Hi unicornio

I dont use \\ and the code did not work, however, if I use this code:

Code:

 "/[^a-zA-Z0-9_-\s]/", ""

The underscore works, Its just the commas I need to allow now.

Cheers scorp

Posted: Fri Jan 20, 2012 5:58 pm

Code:

    /[a-zA-Z0-9,]+/ matches if any of the characters are alphanumeric + comma.


    /^[a-zA-Z0-9,]+$/ matches if all of the characters are alphanumeric + comma.

Code:

 $mnane = preg_replace("/^[a-zA-Z0-9,]*$/", "", $mnane);

Code:




    / : regex delimiters.


    ^ : start anchor


    [..] : Char class


    0-9 : any digit


    a-z : any alphabet


    , : a comma. comma is not a regex metachar, so you need not escape it


    + : quantifier for one or more. If an empty input is considered valid, change + to *


    $ : end anchor


    i : to make the matching case insensitive.

Try this one and let me know. If it doesn't work then someone else can help u
Shocked

Posted: Fri Jan 20, 2012 6:25 pm

Hi unicornio

I have just added the comma within the code:

Code:

"/[^a-zA-Z0-9,_-\s]/", ""

And now I can use the comma and underscore

The code you shown:

Code:




    / : regex delimiters. 


    ^ : start anchor 


    [..] : Char class 


    0-9 : any digit 


    a-z : any alphabet 


    , : a comma. comma is not a regex metachar, so you need not escape it 


    + : quantifier for one or more. If an empty input is considered valid, change + to * 


    $ : end anchor 


    i : to make the matching case insensitive.

I have been looking around the net for a explanation to preg_replace and the breakdown of it, that code explains, nice one

Cheers Scorp

Posted: Fri Jan 20, 2012 7:07 pm

I'm glad you solved it. It was my pleasure to help u. See u around.

registered · Posted: Sat Jan 21, 2012 9:03 am

RN 2.5.0 uses the following instead (takes Unicode into account):

/[^\p{L}\p{N}\p{Pd}\p{Pc}]/

This translates to:

Any Letter
Any Number
Any dash separator (e.g., "-")
Any connector (e.g., "_")