Re: Parsing fully context-free grammars

haberg@math.su.se (Hans Aberg) wrote:


> 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.


Actually, a situation where it is convenient using a different parse than
of the language grammar, is when one has a large number of operator
precedences (like in Prolog) which cannot be conveniently parsed via a
parser generated by a Yacc style static grammar parser generator. In the
parse, one puts the operators and values on a stack or compute the
semantic value as parenthesizes and operator precedences so require.


In article 05-09-067, "Lowell Thomas"
<lowell@coasttocoastresearch.com> wrote:


> It seems to me that this could be generalized to say, in effect, that
> any tree from the forest can be emulated by any other. Does anyone
> know of a contradiction to this?


So there are indeed situations where this is useful and used. But I do not
know of any general exploration of the topic.


--
    Hans Aberg

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