Regular expressions problems

[russx2]

I have a page of text in which I want to replace given words for another word. Simple enough.
I have custom tags in this document, which I translate to html later. The basic rules for my replacing is that the text must be just plain text, e.g., between a closing [/closing] tag and an [opening tag ...]. The way I've coded it means this excludes replacing words between title tags, image tags etc. which is what I want.

An example of the text is here:

--
[h]History of whatever[/h]

It is well known that whatever was introduced in the late 16th century.

[image=whatever.jpg]This is a whatever image[/image]

Here I only want the 'whatever' in the sentence to be replaced, not the one between the [h] tags or between the [image] tags.

I came up with the following, but due to my poor understanding of regexp's, doesn't work!

$text = reg_replace( '#\[\/([^\[\]]*?)\]([^<]*?)(pulp)([^>]*?)\[([^\/]*?)\]#s', '[/$1]$2<a href=\"\">pulp</a>$4[$5]', $text);

I think I'm roughly there, in principal at least... can anyone give me a
hand?

TIA...
Russ

[psy]

an easier solution might be to just use str_replace

$reparr=array("[h]"=>"[newtag]",
                "[/h]"=>"[/newtag]");
$text="[h]history[/h]";

function replaceall($instr,$arr)
{
	if($instr && gettype($arr)=="array"){
		list($key,$val)=each($arr);
		$start=str_replace($key,$val,$instr);
		for($i=1; $i<count($arr); $i++){
			list($key,$val)=each($arr);
			$start=str_replace($key,$val,$start);
		}
		return $start;
	}
}

echo replaceall($text,$reparr);

just a thought. . . but that wouldnt work with the tag

[killerbishop]

A similar solution to aforementioned str_replace(), but works with regular expressions uses preg_replace(). In fact, you could make it a one-liner(!):

$text = preg_replace (array_keys($tokens), array_values($tokens), $text);

Of course, you'll need to defined $tokens as an array, like so:

$tokens = array(

   "#\\[h\\](.*?)\\[/h\\]#i" => "<B>$1</B>",
   "#\\[img=(.*?)\\](.*?)\\[/img\\]#i" => "<center><img src=$1><br>$2</center>"

);

I love elegant solutions ;-)

[russx2]

Hi,

While the above two solutions are great, and will provide a nice clean approach to replacing the tags later, it is not quite what I need. Sorry, my explaining skills are lacking! :rolleyes:

OK, I'll rephrase my original problem...

Forget about the tags included in my text for a sec. I have a database of keywords with which I want to check against each word in the page being shown and highlight (much like the google highlighing when you search for a term).

Now obviously this is as simple as a str_replace(). My problem now is that I don't want to replace matching words if they are in between tags. My [h] [/h] set of tags I used as an example above representings a heading. I don't want a matched word in a heading to be highlighted.

Likewise, the tag [image=xxx.jpg]test[/image] represents an image with the text between being the alt-text.

Now, the 'page' is available to me in a string. Replacing all the tags and such isn't a problem, it's just I want to highlight the words first.

What I think I need to do is only replace a word if, and only if, it is between a [/x] tag and a [x] tag. I think this satisifies my problem, but I am having problems actually implementing this (as can be seen from my dismal approach at a regular expression above! :p ).

So, any pointers welcome... thanks again!

Regards,
Russ

[psy]

download xmb forums (http://xmbforum.com) and check out the source code for how they do the BB Codes.. bc what youre describing sounds almost identical to the 'html' codes we use in forums

[russx2]

Thanks,

But that isn't my problem. I can perform the custom code => html conversion fine.

My problem is that I am trying to do something else - highlighting keywords (from a search term) in the text - which I do not want to occur when said word is between any tag.

Thus, only if the given word is between a closing [/x] tag and an opening [x ....] tag can I be sure it is not part of an image, a heading, a link etc.

Thanks,
Russ

[psy]

damn, that just saved me a lot of time 🙂

thanks -> i never knew those things existed