[SOLVED] creating SEO strings

[tgavin]

I'm working on stripping all of the special characters, etc so that I have SEO friendly urls. I'm not getting the results I want.

Starting with the title Cheech & Chong's Nice Dreams. I would like for it to be 'Cheech-and-Chongs-Nice-Dreams

Nothing I've tried is working.

 

<?php
function create_slug($string) {
return ereg_replace("[^A-Za-z0-9]", "-", $string);
}
$title = "Cheech & Chong's Nice Dreams";
$title_slug = create_slug($title);
echo $title_slug;

// or
return return preg_replace('/\s/', '-', $str);

// i've tried more
$return = ereg_replace("[^A-Za-z0-9]", "-", $string);
$return = preg_replace('/\s/', '-', $return);
return str_replace('--', '-', $return);
?>

 

ugh.

[dsaba]

<?php
$str = 'Cheech &     Chong\'s    Nice     Dreams';
$str = str_replace('&', 'and', $str);
$str = preg_replace('~[^\w]+~', '-', $str);
echo $str;
?>

 

 

notice how multiple non-alphanumeric chars are replaced by a SINGLE '-'

[tgavin]

<?php
$str = 'Cheech &     Chong\'s    Nice     Dreams';
$str = str_replace('&', 'and', $str);
$str = preg_replace('~[^\w]+~', '-', $str);
echo $str;
?>

 

 

notice how multiple non-alphanumeric chars are replaced by a SINGLE '-'

And how!

 

Perfect. That leaves it wide open to add anything else I may need. Thanks!

[wizman45]

Its amazing what you find out. I had the exact problem. Thanks for posting the solution.

[effigy]

[^\w]+

 

FYI: This can be shortened to \W+.

[BrandonK]

Very nice idea tgavin.  That's a lot more efficient than what I was using before.  I added one more step to the replacement for slightly better results (strips punctuation):

<?php
$str = 'Cheech &     Chong\'s    Nice     Dreams';
$str = str_replace('&', 'and', $str);
$str = preg_replace(array('/[\p{P}]+/', '/[^\w]+/'), array('', '-'), $str);
echo $str;

 

This way you don't get Cheech-and-Chong-s-Nice-Dreams, but instead Cheech-and-Chongs-Nice-Dreams. YMMV.

 

\p{P} matches all punctuation (PHP 4.4.0+ and 5.1.0+)

[tgavin]

Very nice idea tgavin.  That's a lot more efficient than what I was using before.  I added one more step to the replacement for slightly better results (strips punctuation):

<?php
$str = 'Cheech &     Chong\'s    Nice     Dreams';
$str = str_replace('&', 'and', $str);
$str = preg_replace(array('/[\p{P}]+/', '/[^\w]+/'), array('', '-'), $str);
echo $str;

 

This way you don't get Cheech-and-Chong-s-Nice-Dreams, but instead Cheech-and-Chongs-Nice-Dreams. YMMV.

 

\p{P} matches all punctuation (PHP 4.4.0+ and 5.1.0+)

Effigy, thank you!

 

Brandon, thank you too. I was using this because I don't know what the hell I'm doing

<?php
function create_slug($str) {
$str = str_replace("'", '', $str);
$str = str_replace('&', 'and', $str);
$str = str_replace('!', '', $str);
$str = str_replace('?', '', $str);
$str = str_replace('(', '', $str);
$str = str_replace(')', '', $str);
$str = str_replace('.', '', $str);
$str = preg_replace('[^\w]+', '-', $str);
return strtolower($str);
}
?>

Throw that in with mod rewrite and you have an SEO gold mine!

[BrandonK]

I discovered a bug with the regex that I posted.  Since I am planning on using this function to replace my current method for creating slugs, I ran it through some of my odd product names...

 

If your product name has a hyphen, it is stripped by the \p{P} so:

"Cheech-Chong Up In Smoke" becomes "CheechChong-Up-In-Smoke"

 

After a lot of trial and error, I came up with this:

 

<?php
$str = 'Cheech &     Chong\'s    Nice     Dreams';
$str = str_replace('&', 'and', $str);
$str = preg_replace(array('/[^\w\s-_\\/]+/', '/[^\w]+/'), array('', '-'), $str);
echo $str;

 

It loses some of its simplicity that we had earlier, but it allows the above string to come out as "Cheech-Chong-Up-In-Smoke" as well as handling backslashes like I needed it to.  If someone else has an idea to clean it up, I'd be happy to run it through my tests.

 

P.S. For those that are interested, here's what I found out about the \p variations:

not matched by \p{P}: `~$^+=<>
{Pc}:	_
{Pd}:	-
{Pe}:	)]}
{Pf}: ????? - no clue, probably has to do with the placement of the punctuation
{Pi}: ????? - no clue, probably has to do with the placement of the punctuation
{Po}:	!@#%&*:"?\;',./
{Ps}:	({[

Not much more info here: http://www.php.net/manual/en/reference.pcre.pattern.syntax.php

[tgavin]

I noticed that bug too. After trial and error, scratching head, pulling hair I finally realized why my previous database records were being rewritten

 

The solution I had wasn't very elegant. However, it did keep the title intact, sans the punctuation, and made everything lowercase. The problem I see in your fix is that if somebody was looking for 'up in smoke', they would be more apt to type 'cheech-and-chongs', not 'Cheech-Chong'. It's also easier to read in a search result, and to remember. Making it lowercase just simplifies everything. Plus, I *think* it's common practice too.

[BrandonK]

I threw Cheech-Chong in there to show that it preserves existing hyphens.  You initial example still works with Cheech & Chong.  I tend to make urls lowercase because servers can be case sensitive.  I do think that search engines are smart enough to not mistake Cheech with cheech, but its always better to error on the side of caution.  strtolower() is easy to add.

[tgavin]

little problem. I noticed that numbers are disappearing. If I use shrek, then shrek2, the original shrek is overwritten.

[BrandonK]

They shouldn't be.  You'd have to post your code, but numerics should stay:

<?php
$str = 'Shrek2 vs Shrek 2';
$str = str_replace('&', 'and', $str);
$str = preg_replace(array('/[^\w\s-_\\/]+/', '/[^\w]+/'), array('', '-'), $str);
echo $str;

[tgavin]

my bad. It was something else. I forgot that I had made this post or I would have mentioned it. Sorry.