Why would creating an array like array[1,2] work? I''m know that it would only create one dimension, but the only use I know for the comma is in functions. Since no ones uses commas when defineing arrays (to my knowledge), why doesn''t this produce an error? "I''ve learned something today: It doesn''t matter if you''re white, or if you''re black...the only color that REALLY matters is green" -Peter Griffin

The comma operator causes the expressions on either side of it to be evaluated, lhs first, and then returns the rhs result. Thus, doing int array[1,2]; is equivalent to doing int array[2];

[ GDNet Start Here | GDNet Search Tool | GDNet FAQ | MS RTFM [MSDN] | SGI STL Docs | Google! ]
Thanks to Kylotan for the idea!

The comma operator returns the value of the last expression in the sequence. a = (b,c,d) where b, c and d are expressions is equivalent to a = d (excepti if b and c have side-effects, of course)

Therefore array[1,2] is equivalent to array[2]

Why would they allow that though? It creates no extra functionality; its pointless and error-prone

"I''ve learned something today: It doesn''t matter if you''re white, or if you''re black...the only color that REALLY matters is green"
-Peter Griffin

Maybe, but it lets you do for( i = 0, j = 1; i < 10; ++i, j*=2 ); The 3 parameters of a for loop are expressions, not statements.

quote: Original post by matrix2113
Why would they allow that though? It creates no extra functionality; its pointless and error-prone

Don''t be silly. Array dimension declarations simply require an expression that evaluates to a constant integral value (which lets you use consts and #defines to modify compile-time behavior), while array indexing requires the expression to evaluate to an integral value (which allows you to use the return values of functions, etc as indices). Your problem is that you think the language is designed/implemented on a per-instance basis ("okay, we allow this for arrays, this for pointers...") rather than in terms of general/generic syntax.

[ GDNet Start Here | GDNet Search Tool | GDNet FAQ | MS RTFM [MSDN] | SGI STL Docs | Google! ]
Thanks to Kylotan for the idea!

quote: Original post by Fruny
Maybe, but it lets you do for( i = 0, j = 1; i < 10; ++i, j*=2 ); The 3 parameters of a for loop are expressions , not statements.

In C++ that is true, but in C the 2nd parameter is a statement and not an expression.


- Kevin "BaShildy" King
Game Programmer: DigiPen
www.mpogd.com

quote: Original post by BaShildy
In C++ that is true, but in C the 2nd parameter is a statement and not an expression.

K&R 2nd ed, section 3.5, p60

for( expr1; expr2; expr3) statement

quote: Original post by Fruny
K&R 2nd ed, section 3.5, p60

for( expr1; expr2; expr3) statement

I stand corrected. I could''ve sworn that C''s for and C++''s for had a different definition of the 2nd parameter. Was it changed as of C99 or has it always been that way?



- Kevin "BaShildy" King
Game Programmer: DigiPen
www.mpogd.com

K&R 2nd ed definitely predates C99. And I vaguely remember it being that way from the first time I read K&R, about ''95.

[ GDNet Start Here | GDNet Search Tool | GDNet FAQ | MS RTFM [MSDN] | SGI STL Docs | Google! ]
Thanks to Kylotan for the idea!