Regex / replace html tags.

[Jezternz]

Okay!

I want to make a function where you can input bascly the body of an html page.
So basicly formatted html. Then I want my function to go through it and add some code to the Open Tags Only (not close tags),
I would like to just have another replace with regex method and replace the html exit tags, however this is a problem as I do not know what the text inside is.
Is there anyway I can do this without making my own custom loop? If so please help! if not, I would apreciate a hand starting off the loop. Thanks

I have made myself a function that does half the work. Only problem is it will do both open and close tags:

function convert_body(body){
 	var newbody = body.replace(/[>]/gi, ' onclick="select(this)">');
 	return newbody;
}

Currently if I input
[HTML]
<div id="container" >
<p class="yellow"> text</p>
<a href="#" class="yellow"> link</a>
</div>
[/HTML]
I get:
[HTML]
<div id="container" onclick="select (this)">
<p class="yellow" onclick="select (this)">text</p onclick="select (this)">
<a href="#" class="yellow" onclick="select (this)">link</a onclick="select (this)">
</div onclick="select (this)">
[/HTML]
When I want:
[HTML]
<div id="container" onclick="select (this)">
<p class="yellow" onclick="select (this)">text</p>
<a href="#" class="yellow" onclick="select (this)">link</a>
</div>
[/HTML]

[acoder]

See this link (grabbing HTML tags).

[Jezternz]

yeh I read that however there was a problem because when I searched for closetag with a wildcard text (ie div or a) when I went to replace it I could not retrieve that wildcard text.

Anyways found a different method to do it:

execute_all_children('document.body');

function execute_all_children(element){
	var elementdir = eval(element);
 	if(elementdir.nodeType == 1){
	 	// Do changes to every element (background in this eg)
 	 	elementdir.style.backgroundColor = 'black';
	}		 	
 	var i = 0;
 	while(i<elementdir.childNodes.length){		
		execute_all_children(element+'.childNodes['+i+']');	
		i++;	
	}
}

[acoder]

So you used recursion. That's all good. It is possible though to capture the text using parentheses.

[Jezternz]

is it more effiecient? could you start me off if it is? cheerz

[acoder]

How about:
[code=javascript]body.replace(/<([A-Z][A-Z0-9]*)\b([^>]*)>/gi, '<$1 $2 onclick="select (this)">');[/code]It matches the opening tag and the rest of the content before it closes and then uses backreferences from the capturing parentheses.

[Jezternz]

wow gonna have to go back to that other site and look that up, try work out exactly what all that means. thankyou so much acoder, I will defnitly look at replacing my other code. This looks far more efficient.

[acoder]

It just matches the opening tag only, but with parentheses for re-using in the replace. The $1 and $2 match the parenthesized (captured) input.

[Jezternz]

oh ok, that now makes alot more sence to me. thats cool, thanks acoder!

[acoder]

In fact, looking back at it, you can match the whole opening tag in one go:[code=javascript]body.replace(/<([A-Z][A-Z0-9]*\b([^>]*)>/gi, '<$1 onclick="select (this)">');[/code]but anyway, you're welcome.