Re: Parsing fully context-free grammars

  "Lowell Thomas" <lowell@coasttocoastresearch.com> wrote:


> Well, in particular, I'm looking at the two parse trees in Fig. 4.6,
> pg. 175 of the Dragon Book. If you just read the translation from the
> leaves of the trees, the first gives
>
> if(expr) then {if(expr) then {stmt} else {stmt}}
>
> the second gives
>
> if(expr) then {if(expr) then {stmt}} else {stmt}
>
> APG actually generates the first tree. But during a depth-first
> traversal of that tree, using different stack push/pop mechanisms at
> the recursive "stmt" nodes, it's not hard to match the "else" either
> of the two "then"s, effectively emulating the second tree during a
> depth-first traversal of the first.


When writing a compiler, one needs to have the parse tree built according
to the grammar specification, due to the stuff put into the actions of the
rules. So therefore, I think one will have little use for a parser
generator that cannot build the proper parse trees. But this is not
intended as an evaluation of your ideas.


--
    Hans Aberg

Return to the comp.compilers page.
Search the comp.compilers archives again.