[antlr-interest] advocacy of C++ support in ANTLR 3.x

[Gavin Lambert]

At 06:23 3/04/2008, Jim Idle wrote:
>The adaptor is an adaptor, or the default tree nodes would not 
>work, but in C there is no class inheritance and I cannot protect 
>functions/pointers that must be public, from being used anywhere. 
>The Java version of the tree node also has addChild etc. It needs 
>to because the tree adaptor needs to ask it to add children. 
>However, if you don't call the treenode methods, then they won't 
>be used directly in either C or Java or any other target.

I think Tom's point is that in the Java version the tree nodes are 
simple objects, and the adaptor class is the only known interface 
-- you ask the adaptor class to create a node, to add a child to a 
node, etc, but when an actual node is referred to it's always an 
'object', not a 'CommonTreeNode'.  The grammar generated code 
never directly touches the tree nodes themselves.

This same approach ought to still be possible in C -- substituting 
'void *' for 'object' and adding some memory tracking/disposal 
semantics, of course.

So the default ANTLR3_COMMON_TREE_ADAPTOR creates and manipulates 
ANTLR3_COMMON_TREE objects, but the interface specified by 
ANTLR3_BASE_TREE_ADAPTOR only ever uses 'void *'.  This means that 
a user can completely reimplement the interface to use their own 
custom node type totally divorced from ANTLR.  The 
ANTLR3_BASE_TREE structure simply goes away, as it's no longer 
useful.  (Or maybe you could still have two levels; one which 
implements the basic tree semantics and one that adds the token 
payload.  But I'm not sure that's especially useful.)