Re: Verification that a CFL is LL(1) / SLR

"Charles E. Bortle, Jr." <cbrtjr@ix.netcom.com>
5 Apr 2000 22:23:40 -0400

          From comp.compilers
Related articles
Verification that a CFL is LL(1) / SLR avi.tal@altavista.net (Avi Tal) (2000-04-01)
Re: Verification that a CFL is LL(1) / SLR torbenm@diku.dk (2000-04-03)
Re: Verification that a CFL is LL(1) / SLR cbrtjr@ix.netcom.com (Charles E. Bortle, Jr.) (2000-04-05)
| List of all articles for this month |
From: "Charles E. Bortle, Jr." <cbrtjr@ix.netcom.com>
Newsgroups: comp.compilers
Date: 5 Apr 2000 22:23:40 -0400
Organization: MindSpring Enterprises
References: 00-04-023 00-04-049
Keywords: parse

> Write a grammar for your language. Use the standard LL(1) and SLR
> parse-table constructions. If you succeed with no conflicts, this is a
> proof that your grammar, and hence your language, is LL(1) / SLR.
>
> Note that this is not a decision procedure: Some grammars for LL(1) or
> SLR languages may not themselves be LL(1) or SLR. AFAIK, there is no
> sure way to decide if a grammar has an equivalent LL(1) or SLR
> grammar, and hence if the language is LL(1) or SLR.
>


Also, for some grammars, for example for the Pascal dangeling else,
the grammar may not be LL(1) but if you structure it and the parser
driver correctly, the correct rule with be taken anyway, so it becomes
a somewhat academic issue.
--
Charles cbrtjr@ix.netcom.com
* http://pw2.netcom.com/~cbrtjr/wrdthing.html *

Post a followup to this message

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