is there a rule that says the attribute's value will be first parsed as HTML?

[winterheat]

i wonder if there is a rule that says:

the attribute's value will be first parsed as HTML

?

for example:

<a href="foo.php?a=1&amp;b=2">

the value will be converted into foo.php?a=1&b=2


and

<a href="#" onclick="changeIt( &quot;hello world &quot; ); return false;">Click me</a>

the value part will be converted into changeIt("hello world"); return false;

is there such a rule? hm... i wonder if any book or webpage actually state this rule? (as many of us learn HTML / CSS / Javascript by ourselves... i haven't heard of this rule before). thanks.

[felgall]

If it is in the HTML file and recognised as PCDATA then it will be parsed as HTML. Only CDATA sections don't get parsed as HTML.

[winterheat]

If it is in the HTML file and recognised as PCDATA then it will be parsed as HTML. Only CDATA sections don't get parsed as HTML.

so do you mean anything that is part of the HTML code is PCDATA?
what about javascript code inside of <script> tag?

because the following code will work. so the code inside <script> is CDATA?

<a href="#" id="link1">Click me</a>
<a href="#" id="link2">Click me</a>    
<div id="divContent"></div>  
 
 
<script>
 
document.getElementById('link1').onclick = function() {
	changeIt("hello world"); return false;
}
 
document.getElementById('link2').onclick = function() {
	changeIt("hello &quot; world"); return false;
}
 
function changeIt(title) {
	document.getElementById('divContent').innerHTML = title + " " + new Date();	
}
 
</script>

[felgall]

In HTML code inside a script tag is CDATA, in XHTML it is PCDATA - you need to look at the doctype definition to see what is defined as PCDATA and what as CDATA.

[winterheat]

In HTML code inside a script tag is CDATA, in XHTML it is PCDATA - you need to look at the doctype definition to see what is defined as PCDATA and what as CDATA.

is it that you need to give the MIME type as XHTML too? also, if IE doesn't support XHTML, then usually we serve it as HTML anyway?

[winterheat]

a friend of mine suggested (see the color text):

<a href="#" onclick='changeIt("hello &amp;quot; world"); return false;'>Click me</a>
<div id="divContent"></div>


<script>

function changeIt(title) {
document.getElementById('divContent').innerHTML = title + " " + new Date();
}

</script>


seems like lazy but it works...

[AutisticCuckoo]

i wonder if there is a rule that says:

the attribute's value will be first parsed as HTML

?

Yes, HTML attribute values are parsed as HTML, so you'll need to use HTML-style escapes (entity referenes or numeric character references) to 'hide' special characters.

is it that you need to give the MIME type as XHTML too?

Yes, otherwise you're not using XHTML at all. If you serve anything as text/html it is HTML, and HTML parsing rules apply.

a friend of mine suggested (see the color text):

<a href="#" onclick='changeIt("hello &amp;quot; world"); return false;'>Click me</a>

Yes, the easiest way to handle quotes inside an attribute value is to use the 'other' quotation marks around the whole value:

<img src="car.jpg" alt="Mary's car">
<img src="car.jpg" alt='My "new" car'>

However, if you need both double quotes and apostrophes inside an attribute value you'll have to escape one of them:

<img src="car.jpg" alt="Mary's &quot;new&quot; car">
<img src="car.jpg" alt='Mary&#39;s "new" car'>