Re: Has lexing and parsing theory advanced since the 1970's?

Roger L Costello <costello@mitre.org> writes:
>That said, Flex & Bison is old. Has lexing/parsing theory advanced since the
>1970s? If yes, are there parser generators available today which are based on
>those advances in lexing/parsing theory? Or does Flex & Bison still represent
>the state-of-the-art in terms of the underlying theory it uses?


It seems to me that after the success of BNF (context-free grammars)
with Algol 60 people wanted to go further, and Algol 68 used a Van
Wijngaarden grammar (two-level grammars). But this formalism has not
been used for describing and implementing later languages, and even
Algol 60 put more in the grammar (in particular, the type difference
between flags and numbers) than later languages. Most languages use
just BNF and describe the rest of the language in prose, a few use a
formal semantics, but I guess that's not what the question was about.
So on the language definition side, there have been no advances in
grammar formalism, because none are needed.


On the implementation side, there have been many attribute grammar
generators (e.g., Ox), but few uses. In the back end various more or
less formal techniques have been used for instruction selection; e.g.,
lcc uses the tree-parser generator lburg.


- anton
--
M. Anton Ertl
anton@mips.complang.tuwien.ac.at
http://www.complang.tuwien.ac.at/anton/

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