lwe3207@acf4.UUCP (04/12/84)
Nf-From: acf4!lwe3207 Apr 22 12:17:00 1984 [] Sorry -- somebody accurately pointed out to me that they *are* equivalent. I had assumed that they were (because I had some other bug in my program), without testing the inequivalence. It didn't occur to me that semicolon followed by "if" had a different precedence than semicolon elsewhere.
lwe3207@acf4.UUCP (04/12/84)
Nf-From: acf4!lwe3207 Apr 22 12:18:00 1984 [not my day] For "if" in the correction read "else".
lwe3207@acf4.UUCP (04/22/84)
Nf-From: acf4!lwe3207 Apr 21 18:30:00 1984
[]
Another subtle non-equivalence:
while (a) if (b) foo; else bar;
is not a syntax error, but is not equal to
while (a) { if (b) foo; else bar; }dmmartindale@watcgl.UUCP (Dave Martindale) (04/23/84)
From: lwe3207@acf4.UUCP
Message-ID: <13900003@acf4.UUCP>
Date: Sat, 21-Apr-84 18:30:00 EST
Another subtle non-equivalence:
while (a) if (b) foo; else bar;
is not a syntax error, but is not equal to
while (a) { if (b) foo; else bar; }
1) If they aren't equivalent, what *is* the first one equivalent to?
2) Please tell us what compiler you are using when you make statements
such as this.
For the record, the 4.2BSD PCC-based compiler *does* treat the two
statements as equivalent, and I cannot see any other interpretation
allowed by the language. The "else" must be part of the if statement.jim@ism780.UUCP (04/26/84)
#R:acf4:-1390000300:ism780:12500004:000:474 ism780!jim Apr 25 10:26:00 1984 > Another subtle non-equivalence: > > while (a) if (b) foo; else bar; > > is not a syntax error, but is not equal to > > while (a) { if (b) foo; else bar; } You are wrong. These are equivalent according to a) C syntax rules. b) pcc generated code. c) common sense. There are a lot of people on this net who act like they know what they are talking about, but don't. BEWARE OF FALSE WIZARDS! -- Jim Balter, Interactive Systems (decvax!yale-co!ima!jim)
Anonymous@inmet.UUCP (05/05/84)
#R:acf4:-1390000300:inmet:5000014:000:632 inmet!Anonymous Apr 26 09:09:00 1984 [] > Another subtle non-equivalence: > > while (a) if (b) foo; else bar; > > is not a syntax error, but is not equal to > > while (a) { if (b) foo; else bar; } OK, I give up. What's the difference? According to my copy of K&R, they both have only one parse, and they differ only by the compound-statement step in the second example. Thus they shouldn't be different in meaning. Is there some C compiler that generates different code? I'd call it a bug. John M Chambers [inmet!jc] Intermetrics, Inc. 735 Concord Ave. Cambridge, MA 02138 "I don't have any solution but I certainly admire the problem."