Removing characters from phone numbers

mrgrammar

I have a database field (phone) that contains phone numbers. Some of the numbers have characters and some don't and the characters are not all the same. Some have parenthesis and dashes while others have only dashes.

I want to remove all the characters but leave the numbers so that I end up with just a 10 digit number. How do I remove just the left/right parenthesis and the dashes?

justsomeone

preg_replace

I don't have access to a php environment just now to test this, so bear in mind that this may not be exactly right, but should be within a few tweaks of it.

let's say your potentially-punctuation-ridden phone number is in a variable called $phoneNumber.

You could do (I think)
$phoneNumber = preg_replace("\D","",$phoneNumber);

Note that the second parameter is an empty string ie double-quote dou0pbele-quote with nothing in between.

This uses the power of regular expressions to replace all non-numeric characters "\D", with nothing "".

It doesn't do anything to enforce the length of the string, but if you need to, you can use strlen() for that.

Drakla

I don't think you can do a REGEXP replace just with a MySQL query, so I think you're going to have to pull out their primary key and the number, perform a [man]preg_replace[/man] on them to replace all non digit (which is /D in a pattern) with nothing (which is '' by the way) and then update it back.

justsomeone

SNAP! 🙂

Drakla

That is SO freaky - I was wondering whether to write SNAP in the thread, then bottled out. I was also trying to come up with one of my all-in-wonder single SQL command solutions, but I'm having none of it, as I've only come up with something pathetically complicated.

mrgrammar

Originally posted by Drakla
...you're going to have to pull out their primary key and the number, perform a [man]preg_replace[/man] on them to replace all non digit (which is /D in a pattern) with nothing (which is '' by the way) and then update it back.

preg_replace worked just fine. What I realized after I began checking the data is that I have to keep alpha characters too. Therefore, I came up with this:

$newphone = preg_replace("/(/","",$qlead[phone]);
$newphone = preg_replace("/)/","",$newphone);
$newphone = preg_replace("/-/","",$newphone);
$newphone = preg_replace("/./","",$newphone);

Two questions:

How do I simplify the above code? The manual says you can use an array, but I'm just starting to figure out arrays.
Where do I find a comprehensive list of all the codes like \D?

(I would have put the above code in PHP brackets, but doing so doesn't show the \ in front of the characters.)

mrgrammar

$newphone = preg_replace("/(/","",$qlead[phone]);
$newphone = preg_replace("/)/","",$newphone);
$newphone = preg_replace("/-/","",$newphone);
$newphone = preg_replace("/./","",$newphone);

How about this? ($oldphone is from the database)

$search = Array("/(/", "/)/", "/-/", "/./");
$newphone = preg_replace($search,"",$oldphone);

Drakla

My fave regex tutorial is at http://www.regular-expressions.info/tutorial.html

http://uk.php.net/manual/en/reference.pcre.pattern.syntax.php

Has the \D things, but there but look at the cheat sheet at http://www.regexlib.com/CheatSheet.htm

Do a google search for "The Regex Coach" to get a nice free tool, but I test most with this old tat I knocked up, I'll attach - it's old, but good enough

It automatically puts in backticks for the delimiters, so don't do /pattern/ just put in pattern.

LordShryku

If you're only going to strip the specific characters, [man]str_replace[/man] would be better, since you don't actually need pattern matching. But if you just want to remove everything except letters and numbers, pattern match specifically for that.

$qlead['phone'] = '887S#-%d34';
echo preg_replace('/[^A-Z0-9]/i', '', $qlead['phone']);

The \D is specific to preg(I think). The above pattern would work though ereg as well, just removing the /i in place for eregi for case insensitivity.