ant-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Bruce Atherton <br...@callenish.com>
Subject Re: ant xdocs! it ran!
Date Thu, 13 Feb 2003 22:43:30 GMT
At 01:58 PM 2/13/2003, Erik Hatcher wrote:
>On Thursday, February 13, 2003, at 04:33  PM, Bruce Atherton wrote:
>>
>>But if the superclass provides public set/add/create methods of the 
>>correct form, those would allow additional attributes and elements to be 
>>given to the task from a build file, right? Shouldn't these be documented?
>
>Right, and yes they should.  And it does do that... right?  Besides the 
><delete> task, which I haven't looked into specifically, is there a case 
>where it was not picking up superclass set/add/create methods?  I think it 
>should be doing that now.

Ahh, sorry. I looked at several tasks and found that the superclass's 
methods were not reflected in the documentation, but of course the ones I 
looked at all extended MatchingTask, which you have already said is 
special-cased. So I created a false impression for myself that none were 
being documented. Sorry about that.


>Datatypes are next up on the list once we iron out the kinks of the 
>tasks.  The infrastructure is already in place to enable that pretty easily.

Beauty way to go, eh.


>As for other pages being generated... if it cannot be directly derived 
>from Ant's source code then it doesn't make sense to generate pages with 
>this process.  Can you give me an example of what you are after here?

Nothing. I was thinking that superclasses needed their own documentation 
page, but your clarification makes it clear that MatchingTask is the only 
one with that requirement.



Mime
View raw message