From mboxrd@z Thu Jan  1 00:00:00 1970
Message-ID: <cd10e193dd8f20c43d30c91b7a8dd4a5@plan9.jp>
To: 9fans@cse.psu.edu
From: Joel Salomon <JoelCSalomon@Gmail.com>
Date: Sun, 18 Feb 2007 15:19:42 -0500
MIME-Version: 1.0
Content-Type: text/plain; charset="UTF-8"
Subject: [9fans] cc lexer bug?
Content-Transfer-Encoding: quoted-printable
Topicbox-Message-UUID: 114e35d6-ead2-11e9-9d60-3106f5b1d025

cpu% cat t.c
void foo (void)
{
	double d;
	d =3D 08.7;
	USED(d);
}
cpu% 8c t.c
t.c:4 syntax error, last name: 8.7
cpu%=20

This came up as I=E2=80=99m making my lexer for C able to scan numbers.  =
I
tried to understand ken=E2=80=99s code, but it gets very hairy right arou=
nd
/sys/src/cmd/cc/lex.c:751 =E2=80=94 and I think there=E2=80=99s a bug.  O=
r, at least,
an undocumented departure from the ANSI standard; Harbison & Steele
(5E) suggest that =E2=80=9C08.7=E2=80=9D is a valid floating point consta=
nt.

As far as my lexer is concerned (http://www.tip9ug.jp/who/chesky/comp/lex=
.c,
if anyone cares), it=E2=80=99s using line-at-a-time buffering courtesy of
Brdstr(2), so I=E2=80=99m back to thinking that regcomp(2) + strtod(2) et=
 al.
is the way to go.  It won=E2=80=99t handle hex floats, but who cares?

--Joel