|Parsers for run-time modifiable grammars firstname.lastname@example.org (Shankar Unni) (1991-05-03)|
|Re: Parsers for run-time modifiable grammars email@example.com (1991-05-03)|
|Re: Parsers for run-time modifiable grammars firstname.lastname@example.org (Thomas Schoebel) (1991-05-06)|
|Re: Parsers for run-time modifiable grammars email@example.com (1991-05-06)|
|From:||Shankar Unni <firstname.lastname@example.org>|
|Organization:||HP India Software Op., Bangalore|
|Date:||Fri, 03 May 91 12:21:39 +0530|
Recently, in this newsgroup, I saw a request from someone about run-time
modifiable grammars. I am aware of the two papers published in May and
November 1990 respectively in the ACM SIGPLAN Notices, but they deal with
a restricted aspect of the matter.
I am looking for a general-purpose parser or parser-generator tool that can
handle languages that allow the programmer to modify his input syntax on
the fly. The specific language I am looking at is "ASN.1" (CCITT standard
X.208), which has an interesting "MACRO" facility which actually allows the
user to embed productions in EBNF notation in the macro definition.
Subsequently, when the macro name is recognized, its arguments are parsed
using the specified productions in the macro definition - which means that
the end of the argument list probably cannot be recognized unless the macro
instance is parsed with the augmented grammar.
This is also a somewhat specific use of run-time modifiable grammars (the
"augmented" grammar is in scope only while parsing the macro instance), but
needs a *much* more powerful mechanism than those described in the SIGPLAN
papers last year.
If folks could mail me any references on this topic, I am willing to
summarize and post a followup to this newsgroup. I would also appreciate
full reference information (title, authors, etc) of the two ACM SIGPLAN
papers, as I can't lay my hands on the issues at the moment...
Shankar Unni E-Mail:
HP India Software Operation, Bangalore Internet: email@example.com
Phone : +91-812-261254 x417 UUCP: ...!hplabs!hpda!shankar
[Irons' IMP72 allowed you to add BNF to the grammar on the fly. It used
Earley's algorithm, the only bottom up parser I know that can handle
ambiguous grammars and that doesn't require rebuilding the whole parser when
you add productions. -John]
Return to the
Search the comp.compilers archives again.