C++Talk.NET C++ language newsgroups   Archives   FAQ   Search   Memberlist   Usergroups   Register   Profile   Log in to check your private messages   Log in  Which method would you prefer?          C++Talk.NET Forum Index -> C++ Language (Moderated) View previous topic :: View next topic   Author Message Dhruv Guest Posted: Wed Dec 24, 2003 10:56 pm    Post subject: Which method would you prefer? Given that there are people with decadeds of expertise and experience on this particular newsgroup, I'd like to know whether you would prefer templates or macros to evaluate compile time expressions? eg: //Assume you want to compute the greater of 2 ints, that are compile time constant. Ok, the operatin to be performed will not be so simple, but just for simplicity, assume that it is some complicated expression that can be evaluated at compile time. template <unsigned int lhs, unsigned int rhs> struct MAX_UI { enum { value = lhs > rhs ? lhs : rhs }; }; v/s: #deifine MAX_MACRO(lhs,rhs) lhs > rhs ? lhs : rhs Usage: const int x = MAX_UI<23, 45>::value; const int y = MAX_MACRO(23, 45); Also, I'd like to know the particular use of unnamed enums. From what I gather, they are somewhat similar to unnamed namespaces? Or have I gotten it absolutely wrong? Merry Christmas! -Dhruv. [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Joshua Lehrer Guest Posted: Thu Dec 25, 2003 7:06 am    Post subject: Re: Which method would you prefer? "Dhruv" <dhruvbird (AT) gmx (DOT) net> wrote Quote: template <unsigned int lhs, unsigned int rhs struct MAX_UI { enum { value = lhs > rhs ? lhs : rhs }; }; v/s: #deifine MAX_MACRO(lhs,rhs) lhs > rhs ? lhs : rhs Usage: const int x = MAX_UI<23, 45>::value; const int y = MAX_MACRO(23, 45); I would prefer the former, as you can't use it on run-time values. The latter could be used, by accident, with run-time values, and we all know how bad that can be. On the other hand, the template, as you presented it, can only be used for integral values. You'd have to add a bit of magic to let it work for floating point values. joshua lehrer factset reseach systems NYSE:FDS [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Francis Glassborow Guest Posted: Thu Dec 25, 2003 2:33 pm    Post subject: Re: Which method would you prefer? In message <pan.2003.12.24.18.06.44.881772 (AT) gmx (DOT) net>, Dhruv <dhruvbird (AT) gmx (DOT) net> writes Quote: Given that there are people with decadeds of expertise and experience on this particular newsgroup, I'd like to know whether you would prefer templates or macros to evaluate compile time expressions? eg: //Assume you want to compute the greater of 2 ints, that are compile time constant. Ok, the operatin to be performed will not be so simple, but just for simplicity, assume that it is some complicated expression that can be evaluated at compile time. template <unsigned int lhs, unsigned int rhs struct MAX_UI { To start with I would not use a name without lowercase letters for something which was not a pre-processor id. Quote: enum { value = lhs > rhs ? lhs : rhs }; }; v/s: #deifine MAX_MACRO(lhs,rhs) lhs > rhs ? lhs : rhs As macros do not respect scope I would avoid them like the plague. Quote: Usage: const int x = MAX_UI<23, 45>::value; const int y = MAX_MACRO(23, 45); Also, I'd like to know the particular use of unnamed enums. I would not do that but it is to some extent a matter of style. Quote: From what I gather, they are somewhat similar to unnamed namespaces? Not really. unnamed namespaces provide scopes enum (named or not) do not provide scopes. Quote: Or have I gotten it absolutely wrong? Misguided rather than absolutely wrong. There is no benefit to not naming an enum type, and it is generally a good idea to avoid the pre-processor if there is a sensible alternative. -- Francis Glassborow ACCU Author of 'You Can Do It!' see http://www.spellen.org/youcandoit or http://www.robinton.demon.co.uk Happy Xmas, Hanukkah, Yuletide, Winter/Summer Solstice to all. [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Dhruv Guest Posted: Thu Dec 25, 2003 9:28 pm    Post subject: Re: Which method would you prefer? On Thu, 25 Dec 2003 02:06:45 -0500, Joshua Lehrer wrote: [...] Quote: I would prefer the former, as you can't use it on run-time values. The latter could be used, by accident, with run-time values, and we all know how bad that can be. But, wouldn't the compiler issue an error when that hapens? I guess you are talking aboout something like this: int x = 0; cin>>x; const int y = x; Right? Now, we cannot use y as compile time constant, or can we? Quote: On the other hand, the template, as you presented it, can only be used for integral values. You'd have to add a bit of magic to let it work for floating point values. How would I make it work for floats? Regards, -Dhruv. [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Dhruv Guest Posted: Thu Dec 25, 2003 9:29 pm    Post subject: Re: Which method would you prefer? On Thu, 25 Dec 2003 09:33:31 -0500, Francis Glassborow wrote: [...] Quote: I would not do that but it is to some extent a matter of style. From what I gather, they are somewhat similar to unnamed namespaces? Not really. unnamed namespaces provide scopes enum (named or not) do not provide scopes. Or have I gotten it absolutely wrong? Misguided rather than absolutely wrong. There is no benefit to not naming an enum type, and it is generally a good idea to avoid the pre-processor if there is a sensible alternative. I thought that they behaved similar to unnamed namespaces in the fact that you can access the enum value without the :: operator as in: enum Named { x = 1, y = 2 }; enum { Unnamed = 23 }; Usage would be: Named:; Unnamed; Right? Regards, -Dhruv. [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Dhruv Guest Posted: Fri Dec 26, 2003 12:53 pm    Post subject: Re: Which method would you prefer? On Thu, 25 Dec 2003 09:33:31 -0500, Francis Glassborow wrote: [...] Quote: Misguided rather than absolutely wrong. There is no benefit to not naming an enum type, and it is generally a good idea to avoid the pre-processor if there is a sensible alternative. Oops, I should have written: enum Named { x = 1, y = 2 }; enum { Unnamed = 23 }; Usage would be: Named (0); Named (1); etc... Or: x, y, etc... Unnamed; Ok, so that clears my doubts... I was wrong in the earlier post, but now I've understood where I was wrong! Regards, -Dhruv. [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Carl Barron Guest Posted: Fri Dec 26, 2003 1:07 pm    Post subject: Re: Which method would you prefer? Dhruv <dhruvbird (AT) gmx (DOT) net> wrote: Quote: Given that there are people with decadeds of expertise and experience on this particular newsgroup, I'd like to know whether you would prefer templates or macros to evaluate compile time expressions? eg: //Assume you want to compute the greater of 2 ints, that are compile time constant. Ok, the operatin to be performed will not be so simple, but just for simplicity, assume that it is some complicated expression that can be evaluated at compile time. template <unsigned int lhs, unsigned int rhs struct MAX_UI { enum { value = lhs > rhs ? lhs : rhs }; }; v/s: #deifine MAX_MACRO(lhs,rhs) lhs > rhs ? lhs : rhs Usage: const int x = MAX_UI<23, 45>::value; const int y = MAX_MACRO(23, 45); Macros are dangerous!! MAX_MACRO above is particularly dangerous if used with expressions that produce side effects , like MAX_MACRO(++x,++y); one of the two variables will be incremented TWICE with MAX_MACRO !! MACROS are visible in a file from the point of introduction to the end of the file,including any files that include a file containing this macro. This can result in redefinition errors changing what the macro does in different portions of a file. a useful such redfinition is includeing <cassert> more than once in a file to selecctively define what the assert() macro does. This is documented and well known but if done accidently with some other problem specific macro have fun finding it give a core dump 'miles' from the macro's invocation... it the macro is simple like #define PI 3.1415928 use const double PI = 3.1415928; for the same effect and no problems with #define PI 0 siliently in a header and then computing x = y/PI + x/(x+1).... redefining a const double is an error and will be flagged!!! if the macro defines an expression, placing that expression in a template avoids multiple revaluation of expressions with side_effects, and if it is inlined even avoids a real function call. template <class T> inline T my_max(const T &x,const T &y) { return x>y?x:y; } avoids a possible evaluate expression twice [in error] problem. if you really want it for only compile time integral constants then template <int N,int M> struct compile_time_max { static const int value = N>M ?N:M; }; compile_time_max<3,4>::value is a const int = 4 at compile time. [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Francis Glassborow Guest Posted: Fri Dec 26, 2003 11:48 pm    Post subject: Re: Which method would you prefer? In message <pan.2003.12.25.17.01.49.345785 (AT) gmx (DOT) net>, Dhruv <dhruvbird (AT) gmx (DOT) net> writes Quote: I thought that they behaved similar to unnamed namespaces in the fact that you can access the enum value without the :: operator as in: enum Named { x = 1, y = 2 }; enum { Unnamed = 23 }; Usage would be: Named:; Though one common compiler accepts that it is wrong because an enum does not introduce a scope. Strictly Named: is illformed. -- Francis Glassborow ACCU Author of 'You Can Do It!' see http://www.spellen.org/youcandoit or http://www.robinton.demon.co.uk Happy Xmas, Hanukkah, Yuletide, Winter/Summer Solstice to all. [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Joshua Lehrer Guest Posted: Mon Dec 29, 2003 10:57 am    Post subject: Re: Which method would you prefer? "Dhruv" <dhruvbird (AT) gmx (DOT) net> wrote Quote: On Thu, 25 Dec 2003 02:06:45 -0500, Joshua Lehrer wrote: [...] I would prefer the former, as you can't use it on run-time values. The latter could be used, by accident, with run-time values, and we all know how bad that can be. But, wouldn't the compiler issue an error when that hapens? No. I meant that the macro could be used on run-time values, which can be bad: MAX_MACRO(func1(),func2()); if func1() and/or func2() is long running. The compiler would not generate a warning on this. -joshua lehrer factset research systems NYSE:FDS [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] Back to top Display posts from previous: All Posts1 Day7 Days2 Weeks1 Month3 Months6 Months1 Year Oldest FirstNewest First         C++Talk.NET Forum Index -> C++ Language (Moderated) All times are GMT Page 1 of 1   Jump to: Select a forum C/C++----------------C LanguageC Language (Moderated)C++ Language (Moderated)C++ language (comp.lang.c++)C++ language, library and standards Non-English C/C++ Newsgroups----------------C++ (German)C++ (French)  You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum Powered by phpBB © 2001, 2006 phpBB Group SEO toolkit © 2004-2006 webmedic.