
www.Usenet.com
| <-- __Chronological__ --> | <-- __Thread__ --> |
Alan Jiang <[EMAIL PROTECTED]> wrote: > For example: (all variables are int type) > a-c+d is expressed as tree (a-c)+d, but I hope IBURG can transform it > [..] > Does it possible for IBURG rules to do the above job? Or it is just > above the ability of IBURG? IBurg can do it, but it is an unusual application of IBurg. Typically, IBurg matches part of a tree and generates code for it in a different representation. You could generate a new tree but you can't re-write the tree that you are matching in place (at least I believe so). -- Christian
| <-- __Chronological__ --> | <-- __Thread__ --> |