[antlr-interest] Complaints about BaseAST implementation

[Akhilesh Mritunjai]

Hi

The problem is bit more than that. If the fields are
removed from BaseAST (there is no reason for them to
be there), following classes are affected:

BaseAST
CommonAST
ParserTree
ParseTreeRule
ParseTreeToken

I finally forked the code, and made changesin ANTLR
rather than copying/pasting the algorithms and making
a new MyBaseAST class that I might need to update
everytime a new release comes over. And I figured that
I can submit the diffs to Terence after I test it.

In our testing so far, there haven't been any
problems. The fix works like a charm and now our
appliction handles multi-million node trees with
default heap settings :-)

- Akhilesh

--- Andy Tripp <atripp at jazillian.com> wrote:
> Looks to me that the only places where the "right"
> and "down" fields are 
> misused are in
> the addChild() and getNumberOfChildren() methods of
> BaseAST.
> Here is what they should look like:
> 
>  /**Add a node to the end of the child list for this
> node */
>     public void addChild(AST node) {
>         if (node == null) return;
>         BaseAST t = getFirstChild();
>         if (t != null) {
>             while (t.right != null) {
>                 t = t.getNextSibling();
>             }
>             t.right = (BaseAST)node;
>         }
>         else {
>             this.down = (BaseAST)node;
>         }
>     }
> 
>     /** How many children does this node have? */
>     public int getNumberOfChildren() {
>         BaseAST t = getFirstChild();
>         int n = 0;
>         if (t != null) {
>             n = 1;
>             while (t.getNextSibling() != null) {
>                 t = t.getNextSibling();
>                 n++;
>             }
>             return n;
>         }
>         return n;
>     }
> 
> 



		
__________________________________ 
Yahoo! Music Unlimited 
Access over 1 million songs. Try it free.
http://music.yahoo.com/unlimited/