* [Caml-list] genlex ...
@ 2003-07-07 10:13 Pietro Abate
2003-07-08 0:34 ` Pietro Abate
0 siblings, 1 reply; 2+ messages in thread
From: Pietro Abate @ 2003-07-07 10:13 UTC (permalink / raw)
To: caml-list
Hi all,
I'm re-writing the parser of my application and I'd like to write a kind
of generic parser that can parse a language that I can change dynamically...
In other words... My application is composed by a library, a set of
modules that are compiled separately and dynamically loaded and an
application that get the name of a module, a file and it parses the file
differently regarding the definions that it loads from the dynamic
module.
My point is that I want to have a generic parser that given a set of
keyword is able to parse different input.
so far I realized that I cannot do that with lex/yacc as they need a
compilation step (I could generate the lex/yacc file from the dynamic
module, but I think it's too messy...). But I can use Genlex.
Since I'm using Dynlink it's strightforward to load at runtime a
different set of keyword and initialized the lexer as :
let flexer = make_lexer Stub.variable_keyword_list
but at this point I don't know how to use the Genlex.parser to feed it
with my generic rules (every connective is declared as left assoc or non
assoc and the generic language is a calculator-like one ... )
I'd like to write something like
expr:
| LBRACK expr RBRACK { $2 }
| expr (keyword buf) expr { Tree(keyword_lookup buf,$1,$3)
| (keyword buf) expr { Leaf(keyword_lookup buf,$2)
| IDENT { Var($1)
where (keyword buf) is a function that returns 'Kwd if buf is a keyword
or fails ... Or using Genlex.parser as :
let rec parse_level1 = parser
[< e1 = parse_level0; e = parse_level11 e1 >] -> e
and parse_level11 e1 = parser
[< (keyword buf); e2 = parse_level1 >] ->
| [< >] -> e1
and parse_level0 = parser
[< (keyword buf); e = parse_level0 >] ->
....
is it possible ? does anybody can give me a small example ?
tnx,
P
--
Civilization advances by extending the number
of important operations which we can perform
without thinking. (Alfred North Whitehead)
-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: [Caml-list] genlex ...
2003-07-07 10:13 [Caml-list] genlex Pietro Abate
@ 2003-07-08 0:34 ` Pietro Abate
0 siblings, 0 replies; 2+ messages in thread
From: Pietro Abate @ 2003-07-08 0:34 UTC (permalink / raw)
To: caml-list
Hi all, I solved my problem...
On Mon, Jul 07, 2003 at 08:13:44PM +1000, Pietro Abate wrote:
> I'm re-writing the parser of my application and I'd like to write a kind
> of generic parser that can parse a language that I can change dynamically...
that's a piece of code that gives me keyword flexibility as I wanted...
(conn is an object that stores my keyword with their associativity).
let flexer = make_lexer (
conn#keywords @
["(" ; ")"; "atom" ]
);;
let find_keyword level = parser
[< 'Kwd x >] ->
let v =
try conn#lookup x
with Not_found -> raise Stream.Failure
in
if (conn#assoc v = level) then x
else raise Stream.Failure
;;
let rec parse_main = parser
[< e1 = parse_aux; e = parse_expr e1 >] -> e
and parse_expr e1 = parser
[< n = find_keyword (Connective.Left); e2 = parse_main >] ->
term_h#node n [e1;e2]
| [< >] -> e1
and parse_aux = parser
[< n = find_keyword (Connective.NonAssoc); e = parse_aux >] ->
term_h#node n [e]
| [< 'Ident s >] -> term_h#atom s
| [< 'Kwd "("; e = parse_main; 'Kwd ")" >] -> e
I haven't test it yet, but it seems to make sense and compile correctly.
p
--
Civilization advances by extending the number
of important operations which we can perform
without thinking. (Alfred North Whitehead)
-------------------
To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr
Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2003-07-08 0:34 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2003-07-07 10:13 [Caml-list] genlex Pietro Abate
2003-07-08 0:34 ` Pietro Abate
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox