Path: csiph.com!xmission!news.snarked.org!border2.nntp.dca1.giganews.com!nntp.giganews.com!news.iecc.com!.POSTED.news.iecc.com!nerds-end
From: luser droog <mijoryx@yahoo.com.dmarc.email>
Newsgroups: comp.compilers
Subject: Re: Supporting multiple input syntaxes
Date: Thu, 13 Aug 2020 21:36:13 -0700 (PDT)
Organization: Compilers Central
Lines: 49
Sender: news@iecc.com
Approved: comp.compilers@iecc.com
Message-ID: <20-08-006@comp.compilers>
References: <20-08-002@comp.compilers> <20-08-004@comp.compilers>
Mime-Version: 1.0
Content-Type: text/plain; charset="UTF-8"
Injection-Info: gal.iecc.com; posting-host="news.iecc.com:2001:470:1f07:1126:0:676f:7373:6970"; logging-data="99725"; mail-complaints-to="abuse@iecc.com"
Keywords: C, parse
Posted-Date: 15 Aug 2020 10:44:32 EDT
X-submission-address: compilers@iecc.com
X-moderator-address: compilers-request@iecc.com
X-FAQ-and-archives: http://compilers.iecc.com
In-Reply-To: <20-08-004@comp.compilers>
Xref: csiph.com comp.compilers:2558

On Thursday, August 13, 2020 at 5:22:51 PM UTC-5, Hans-Peter Diettrich wrote:
> Am 13.08.2020 um 00:20 schrieb luser droog:
> > I've got my project successfully parsing the circa-1975 C syntax
> > from that old manual. I'd like to add parsers for K&R1 and c90
> > syntaxes.
> >
> > How separate should these be? Should they be complete
> > separate grammars, or more piecewise selection?
>
> IMO this depends widely on the usage of the parser output (diagnostics,
> backend...). C90 is much stricter than K&R, requires more checks. Do you
> need extensive error diagnostics, or do you assume that all source code
> is free of errors?
>
>
> > https://github.com/luser-dr00g/pcomb/blob/master/pc9syn.c
>
> You seem to implement an LL(1) parser? My C98 Parser is LL(2), i.e. an
> LL(1) parser with one or two locations where more lookahead is required.
> Also identifiers are classified as typenames and others prior to their
> usage.
>

Yes, it's basically LL(1) with backtracking. There's one part of the
grammar I'm using that's left-recursive and I still need to work that
out.

> For real-world testing (recommended!) a preprocessor is required and a
> copy of the standard libraries of existing compiler(s).
>
> Your test_syntax() source misses "=" from the variable declarations
> (initializers). What about pointer syntax/semantics? If you add these
> (and other) syntax differences conditionally (version specific) to your
> code, which way would look better to you? Which way will be safer to
> maintain?
>

That's actually correct for the 1975 dialect: no '=' to initialize
variables. I think it's pretty ugly without it, but it could be
removed anyway for the AST.

>
> Nice code BTW :-)
>

Thanks! I think I need to sidetrack a bit and work up some primitives
for pattern matching and decomposition to make the backend easier.
I'll report back if/when it can do more tricks.