What's the difference between function declaration and function prototype?

Hill Pang <pangbw@gmail.com> writes: > I can find some answer from web, but I can't find their definitions in > the C99 standard so far. > > Which page is it in? A function prototype is a *kind* of function declaration, one that specifies the types of the parameters. A non-prototype declaration has empty parentheses; it's also known as an old-style or K&R-style declaration, and it's obsolescent. The term "prototype" is defined in C99 and C11 6.2.1p2: A *function prototype* is a declaration of a function that declares the types of its parameters. -- Keith Thompson (The_Other_Keith) kst-u@mib.org <http://www.ghoti.net/~kst> Will write code for food. "We must do something. This is something. Therefore, we must do this." -- Antony Jay and Jonathan Lynn, "Yes Minister"

On 8/7/2012 1:30 PM, Varun Tewari wrote: > There a little difference, but in modern context, they are same. That seems misleading. Keith Thompson quoted the actual definition: "A function prototype is a declaration of a function that declares the types of its parameters." The very definition suggests the existence of function declarations that omit the parameter types, and indeed if you look elsewhere in the Standard you will find the rules for declaring functions this way. By "in modern context," I imagine you mean something like "always use prototypes, because a compiler that can't handle them cannot possibly be `modern'." That's good advice: There is almost never a reason to omit the parameter types, and I'd venture that even when a reason exists it's probably a weak one. The Standard describes non-prototype function declarations as an "obsolecent feature" of the language, and even if it never gets all the way to "obsolete" the message is clear: Don't Do That. Still, when language feature F comes in forms F1 and F2, with F1 discouraged and F2 strongly recommended, I think someone who asks "What's the difference between F and F2" deserves more of an answer than "Never mind." -- Eric Sosman esosman@ieee-dot-org.invalid

karora129@gmail.com writes: > prototype and declaration both are same That's incorrect. As I wrote in my response last week: The term "prototype" is defined in C99 and C11 6.2.1p2: A *function prototype* is a declaration of a function that declares the types of its parameters. -- Keith Thompson (The_Other_Keith) kst-u@mib.org <http://www.ghoti.net/~kst> Will write code for food. "We must do something. This is something. Therefore, we must do this." -- Antony Jay and Jonathan Lynn, "Yes Minister"

On 08/16/2012 05:37 AM, karora129@gmail.com wrote: > prototype and declaration both are same Are you claiming that "int func();" is a function prototype? Or are you claiming that it's not a function declaration? -- James Kuyper

On Aug 7, 6:55=A0pm, Eric Sosman <esos...@ieee-dot-org.invalid> wrote: > On 8/7/2012 1:30 PM, Varun Tewari wrote: > > There a little difference, but in modern context, they are same. <snip> > =A0 =A0 =A0By "in modern context," I imagine you mean something like > "always use prototypes, because a compiler that can't handle them > cannot possibly be `modern'." =A0That's good advice: There is almost > never a reason to omit the parameter types, and I'd venture that > even when a reason exists it's probably a weak one. maintaining K&R style code. Yes it's still about. It's only a couple of years since I modified a K&R program. > =A0The Standard > describes non-prototype function declarations as an "obsolecent > feature" of the language, and even if it never gets all the way to > "obsolete" the message is clear: Don't Do That. <snip>

On 8/16/2012 8:18 AM, Nick Keighley wrote: > On Aug 7, 6:55 pm, Eric Sosman <esos...@ieee-dot-org.invalid> wrote: >> On 8/7/2012 1:30 PM, Varun Tewari wrote: > >>> There a little difference, but in modern context, they are same. > > <snip> > >> By "in modern context," I imagine you mean something like >> "always use prototypes, because a compiler that can't handle them >> cannot possibly be `modern'." That's good advice: There is almost >> never a reason to omit the parameter types, and I'd venture that >> even when a reason exists it's probably a weak one. > > maintaining K&R style code. Yes it's still about. It's only a couple > of years since I modified a K&R program. [...] Until recently, the code I maintain had to be able to compile on HP-UX. The C compiler that came with the system would recognize prototypes, and tell you that you needed to buy their other C compiler if you needed prototype support. As I recall, there is now a gcc for HP-UX, eliminating that hurdle. -- Kenneth Brody

Kenneth Brody <kenbrody@spamcop.net> writes: > On 8/16/2012 8:18 AM, Nick Keighley wrote: > > On Aug 7, 6:55 pm, Eric Sosman <esos...@ieee-dot-org.invalid> wrote: > >> On 8/7/2012 1:30 PM, Varun Tewari wrote: > > > >>> There a little difference, but in modern context, they are same. > > > > <snip> > > > >> By "in modern context," I imagine you mean something like > >> "always use prototypes, because a compiler that can't handle them > >> cannot possibly be `modern'." That's good advice: There is almost > >> never a reason to omit the parameter types, and I'd venture that > >> even when a reason exists it's probably a weak one. > > > > maintaining K&R style code. Yes it's still about. It's only a couple > > of years since I modified a K&R program. > [...] > > Until recently, the code I maintain had to be able to compile on > HP-UX. The C compiler that came with the system would recognize > prototypes, and tell you that you needed to buy their other C compiler > if you needed prototype support. > > As I recall, there is now a gcc for HP-UX, eliminating that hurdle. I was using gcc in HP-UX back in the 90s. I know I had access to a range between 2.5 and 2.95, which covers most of the decade. Phil -- > I'd argue that there is much evidence for the existence of a God. Pics or it didn't happen. -- Tom (/. uid 822)