cpptool-develop

[Cpptool-develop] ASTNodeTypeFeatures and ASTNodePropertyFeatures...

[Baptiste L. <gai...@fr...>]

    I'm not really happy with the ASTNodeTypeFeatures and
ASTNodePropertyFeatures  features. They seems to only be used to 'optimize'
the visit (mayHave...). In practice, this means that each time we add a
visitor for a specific kind of entity, some features are added to speed up
the visit. This sound plain wrong to me (you have to change most of the
ASTNodeTypes construction to do that).

    Also, doing it that way make it difficult to see what is being visited
as you have to search through all the node types for some specific features.

    I'm thinking of removing feature support for ASTNodeType and
ASTNodeProperty (or a least remove its usage for driving visitation).
ASTNodePropertyFeatures is in fact unused. An alternative to drive the node
visitation would probably be to declare what really need to be visited in
the visitor, much like what has been done in the IdentifierResolver.

What do you think ?

Baptiste.

Re: [Cpptool-develop] ASTNodeTypeFeatures and ASTNodePropertyFeatures...

[Sven R. <rei...@ma...>]

I'm not very happy with this neither. Actually, I've been thinking about 
how the whole ASTNode hierarchy can be improved (with all the properties, 
it's not very "abstract" anymore). I'm not sure yet whether inheritance 
(in form of a State/Strategy) can help us here.

I'll keep thinking about this over the holidays.

Sven.

-- 
Sven Reichard
Dept. of Math. Sci.
University of Delaware
rei...@ma...

Re: [Cpptool-develop] ASTNodeTypeFeatures and ASTNodePropertyFeatures...

[Baptiste L. <gai...@fr...>]

At the current time, there is an obstacle to using a hierarchy for
ASTNode.It is the node mutation which change the type of the node.

Baptiste.

----- Original Message -----
From: "Sven Reichard" <rei...@ma...>
To: "CppTool Mailing List" <Cpp...@li...>
Sent: Friday, December 20, 2002 12:11 AM
Subject: Re: [Cpptool-develop] ASTNodeTypeFeatures and
ASTNodePropertyFeatures...


> I'm not very happy with this neither. Actually, I've been thinking about
> how the whole ASTNode hierarchy can be improved (with all the properties,
> it's not very "abstract" anymore). I'm not sure yet whether inheritance
> (in form of a State/Strategy) can help us here.
>
> I'll keep thinking about this over the holidays.
>
> Sven.
>
> --
> Sven Reichard
> Dept. of Math. Sci.
> University of Delaware
> rei...@ma...

Re: [Cpptool-develop] ASTNodeTypeFeatures and ASTNodePropertyFeatures...

[Sven R. <rei...@ma...>]

On Fri, 20 Dec 2002, Baptiste Lepilleur wrote:

> At the current time, there is an obstacle to using a hierarchy for
> ASTNode.It is the node mutation which change the type of the node.
> 
> Baptiste.

That's why I mentioned a State pattern.
Sven.

-- 
Sven Reichard
Dept. of Math. Sci.
University of Delaware
rei...@ma...