James Strachan wrote:
There's a few ways this could go. We could add a mechanism so that Tag
libraries could restrict the use of other libraries within themselves. e.g.
a specific TagLibrary could dsiable or enable some/all libraries from being
usable within one/all of its tags.

Another option could be for an individual tag to validate its body inside
the doTag() method. By calling getBody() and then traversing the tree of
TagScripts, a specific Tag could validate that the body does not contain
dynamic expressions or Tags of certain kinds etc.

The latter would be pretty easy to implement in your specific example I
think. We could maybe provide some kind of helper class (TagWalker?) to aid
Tag implementors wishing to traverse over their bodies, validating child
tags.
both seem useful. However, in my case, where I dont want any scripting at all, I could also call getParent() from within my element2, and assert that the class is that of element1, right?

In general, I have the feeling that it would be nice if we could delegate more error checking and validation into the runtime, based on declarations provided by the taglib - instead of having to perform it in every tag implementation (or not at all).

James
-------
http://radio.weblogs.com/0112098/
----- Original Message -----
From: "Christian Sell" <[EMAIL PROTECTED]>
To: "Jakarta Commons Users List" <[EMAIL PROTECTED]>
Sent: Monday, February 03, 2003 9:51 PM
Subject: Re: [Jelly] restrict use of tag libraries


Paul Libbrecht wrote:

Christian,

What do you mean by "no taglibs allowed here" ?

I mean that writing

<element1 xmlns="com.yadda.Yadda">

*will* go through, but writing

<element1 xmlns="com.yadda.Yadda" xmlns:y="core">

will *not* go through. In fact, what I really want is to make sure that
there is no scripting from other taglibs between my element1 and
element2 - the namespaces are not really the problem.

<

Do you mean no taglib-declaration ?
Do you mean only the following taglibs' tags ?

In both cases, this has to do with parsing. So a schema, as James
mentionned, or some other (e.g. forcing a SAX filter between the parser
and the jelly document-handler).

I still dont see what role a "schema" would play here


Remember: jelly scripts are first parsed then executed (possibly several
times).
Hope it helps.

Paul


Le Lundi, 3 f�vr 2003, � 19:35 Europe/Berlin, Christian Sell a �crit :


<element1 xmlns="com.yadda.Yadda"> <!-- no taglibs allowed here -->
   <element2 xmlns:j="jelly:core" xmlns:log="jelly:log">
       <j:something/>
       .....
   </mytags:element2>
</mytags:element1>


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]





---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

__________________________________________________
Do You Yahoo!?
Everything you'll ever need on one web page
from News and Sport to Email and Music Charts
http://uk.my.yahoo.com

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]




---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to