[RESOLVED] Replacing characters in a string

Master_Zen

Hello,

I am writing a search engine with PHP which takes the search string and checks it inside a mySQL table using a LIKE statement with the variable enclosed in %s.

The engine is for an aviation site. So for example, I would like things like "F-16", "F.16", "F16", and "F 16" all to generate the same hits, say, by turning it into F%16.

I believe the correct PHP function is ereg_replace but I have no idea how to actually use it because I am unfamiliar with all the characters used in these types of functions (all the / } [ whatever).

Also, if anyone has a link to a good tutorial so I can learn this, please do so.

Thanks

big_nerd

Master Zen,

The "F16" would be difficult unless you KNEW the format. For the rest though, you can use preg_replace().

<?php
$str = array("F-16", "F.16", "F16", "F 16");
$sstr = "F-16";

foreach ($str as $key => $value) {  // for an array of values
	$str = preg_replace("/[^A-Za-z0-9]/","%",$value);
	echo "The Origional $value - New: $str<br>\n";
}

$newstr = preg_replace("/[^A-Za-z0-9]/","%",$sstr); // For a single value
echo $newstr;
?>

This will take out anything that is not 0-9 or a-z and replace it with a % sign.

dagon

you should consider mysql's Full-Text Search Functions

Master_Zen

Thanks guys, I've been studying Regex from various tutorials online since yesterday and I've been souping up the formula.

I've trying to find a way by which I can use preg_replace to modify the user-inputed string for searching the database in the following ways:

1 - If the string includes a non-alfanumeric digit, convert it to "". So, strings like F-15, F.15, F 15 would all end up F_15 which would be the actual string I search the db with (since is the wildcard).

2 - If the string does not include a digit, then separate it. So, F15 would be F_15.

This is the code

$newsearch = preg_replace("/^([a-z+])(\W?)(\d+)/i","\\1_\\3", $oldsearch);

first pattern searches for the first letter digit or digits (the "F"). Second pattern searches for a possible existing non-alphanumeric character (the "-" or "." or space or none). Third pattern searches for the rest of the string regardless of what appears.

In the replacement it makes the new string start with the first pattern, followed by a "_" then continue with the third pattern, ommitting the second completely.

Works like a charm. 😃

big_nerd

Excellent! Hope I helped a bit.

Don't forget to Mark the Thread Resolved once your done (under Thread Tools).

Master_Zen

Thank you very much 🙂