Strategy to parse own tags to html

meeiw

I'm planning on how to parse a string into HTML.
The string will contain some ownspecial [markers] that will render the string differently.

$string="[bold]boldtext[italic]bold and italic text[/bold] italic text[/italic]"

will render the text into

$renderedString="<b>boldtext<i>bold and italic text</b>  italic text</i>"

If the string contains incorrect markers they will be rendered as plain text, like;

$string="[bold]boldtext[italic]bold and italic [/bold]text[/bold] italic text[/italic]"
$renderedString="<b>boldtext<i>bold and italic</b> text &#nr;/bold&#nr;  italic text</i>"

The strategy I first came up with was the following:

Find a legal marker that has an closing marker as well
Render the text inside markers to html
Go to next marker and do it again
[/list=3]This ain't gonna work. If I render the content inside to markers I migth render another legal marker to html, so i have must switches to not edit legal markers and then I must in some very painful way implode all complex strings into one...
And I starting to avoid this bad strategy. And as I can see this forum has some kind of parser to do this. Whats the most common way to achieve this?

Mordecai

str_replace().

meeiw

Ok. But how will I use str_replace()?
A dummy script below won't check after any incorrect markers.

$string="[bold]boldtext[/bold][/bold]";
$renderedString1=replace_string("[bold]", "<b>",$string);
$renderedString2=replace_string("[/bold]", "</b>",$string);

It feels like I need some preg_replace to validate and then renderer the string in someway, or else I'm just really annoying and doesn't understand how to use str_replace. Hopefully I'm not.

meeiw

Finally I last got a solution (i think) to my problem. Here is how I solved it;

//The string to be parsed
$str='[    bold ]blaom[/bold][ link =http://www.aftonbladet.se ]ab aaaa[ italic ] a[/link] g//g/[/italic]hf g[underrubrik]ab </html>ag aa[/underrubrik] sd fh dsf[/link]  sdfj dsfjh jh dsfj hdfs';

$search_pattern=array(
	'#\[\s*(bold)\s*\](.*?)\[/\s*\1\s*\]#',
	'#\[\s*(italic)\s*\](.*?)\[/\s*\1\s*\]#',
	'#\[\s*(underrubrik)\s*\](.*?)\[/\s*\1\s*\]#',
	'#\[\s*(link)\s*=(.+?)\s*\](.*?)\[/\s*\1\s*\]#'
	);
$replace_pattern=array(
	'<B>$2</B>',
	'<I>$2</I>',
	'<H2>$2</H2>',
	'<A HREF="$2">$3</A>'
	);

$str=htmlentities($str);
$hits=preg_replace($search_pattern, $replace_pattern,$str);

echo"<h1>String</h1>";
echo $str;
echo"<h1>Parsing</h1>";
echo $hits;
echo"<br>";

If you need some more own tags simply add a regexp in the search array and a replacement exp in the replace array.

Anyhow Mordecai posted a tip to use str_replace instead. Does this win much in speed in executing the script or is my script fast enough? Any opinions here appreciated.
Thanks in advance.