Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.c++ > #15072 > unrolled thread

macro definition for different compile configurations

Back to article view | Back to comp.lang.c++

Contents

  macro definition for different compile configurations thomas <freshthomas@gmail.com> - 2012-03-27 00:29 -0700
    Re: macro definition for different compile configurations Victor Bazarov <v.bazarov@comcast.invalid> - 2012-03-27 07:58 -0400
    Re: macro definition for different compile configurations thomas <freshthomas@gmail.com> - 2012-03-27 06:44 -0700
      Re: macro definition for different compile configurations Qi <no@no.com> - 2012-03-27 21:53 +0800
        Re: macro definition for different compile configurations Juha Nieminen <nospam@thanks.invalid> - 2012-03-27 17:31 +0000
    Re: macro definition for different compile configurations thomas <freshthomas@gmail.com> - 2012-03-27 07:05 -0700
      Re: macro definition for different compile configurations Geoff <geoff@invalid.invalid> - 2012-03-27 09:49 -0700
        Re: macro definition for different compile configurations Stuart Redmann <DerTopper@web.de> - 2012-03-27 11:27 -0700
          Re: macro definition for different compile configurations Geoff <geoff@invalid.invalid> - 2012-03-27 11:46 -0700
            Re: macro definition for different compile configurations Stuart Redmann <DerTopper@web.de> - 2012-03-27 12:39 -0700
              Re: macro definition for different compile configurations Ian Collins <ian-news@hotmail.com> - 2012-03-28 08:49 +1300
                Re: macro definition for different compile configurations Stuart Redmann <DerTopper@web.de> - 2012-03-27 12:58 -0700
                  Re: macro definition for different compile configurations Geoff <geoff@invalid.invalid> - 2012-03-27 15:50 -0700
            Re: macro definition for different compile configurations Juha Nieminen <nospam@thanks.invalid> - 2012-03-28 08:14 +0000
              Re: macro definition for different compile configurations Stuart Redmann <DerTopper@web.de> - 2012-03-28 07:30 -0700
        Re: macro definition for different compile configurations thomas <freshthomas@gmail.com> - 2012-03-27 18:00 -0700

#15072 — macro definition for different compile configurations

Hi, 
 I'm stucked by something appears really simple at first sight.

I have a function definition with different parameter list in different compile configurations. for example.

--------------------------
CONF1:
int f(int x, int y){}
int g(int x){}

CONF2:
int f(int y){}
int g(){}
----------------------------
I'm trying to define some marco like

#if CONF1
#define PARAM(list) int x, list
#elseif CONF2
#define PARAM(list) list
#endif

But it cannot work for func g() definition in CONF1 since a "," is appended unexpectedly.

It seems to be really simple, but I am now a little dumb how to define one and solve it elegantly.

thanks. 

tom

[toc] | [next] | [standalone]

#15074

On 3/27/2012 3:29 AM, thomas wrote:
> Hi,
>   I'm stucked by something appears really simple at first sight.
>
> I have a function definition with different parameter list in different compile configurations. for example.
>
> --------------------------
> CONF1:
> int f(int x, int y){}
> int g(int x){}
>
> CONF2:
> int f(int y){}
> int g(){}

Why don't your functions declared returning a value return a value?  Are 
those supposed to be definitions or declarations?

You *do* know that in C++ you're allowed to have functions with the same 
name and different argument list, right?  That's called "overloading". 
You *can* declare/define your 'f' and 'g' all in the same scope.

> ----------------------------
> I'm trying to define some marco like
>
> #if CONF1
> #define PARAM(list) int x, list
> #elseif CONF2
> #define PARAM(list) list
> #endif

And use it HOW?  To achieve WHAT?

> But it cannot work for func g() definition in CONF1 since a "," is appended unexpectedly.

"Cannot work" for what?

> It seems to be really simple, but I am now a little dumb how to define one and solve it elegantly.

What exactly are you *trying to do*?  You're stating that your solution 
is not working, but what is *the problem* that you're trying to solve? 
(And please don't tell us that the problem is that the solution is not 
working)

V
-- 
I do not respond to top-posted replies, please don't ask

[toc] | [prev] | [next] | [standalone]

#15075

Ok...

#ifdef _A_
#define PARAM_WITH_COMMA int x,
#defien PARAM_NO_COMMA int x
#else
#define PARAM_WITH_COMMA
#define PARAM_NO_COMMA
#endif

And I have two versions of following functions
#ifdef _A_
int f(int x, int y);
int f(int x);
#else
int f(int y);
int f();
#endif

And I can write the two versions of the functions with the following declaration.
int f(PARAM_WITH_COMMA int y);
int f(PARAM_NO_COMMA);

In short I want to make my code neat and easy to read, that's to say, fewer "#ifdef" like stuff.

So I'm trying to write some macros to shorten some word usage.

I just gave an example above using two marco definitions.

Can I use just one macro to achieve the same thing?




On Tuesday, March 27, 2012 3:29:11 PM UTC+8, thomas wrote:
> Hi, 
>  I'm stucked by something appears really simple at first sight.
> 
> I have a function definition with different parameter list in different compile configurations. for example.
> 
> --------------------------
> CONF1:
> int f(int x, int y){}
> int g(int x){}
> 
> CONF2:
> int f(int y){}
> int g(){}
> ----------------------------
> I'm trying to define some marco like
> 
> #if CONF1
> #define PARAM(list) int x, list
> #elseif CONF2
> #define PARAM(list) list
> #endif
> 
> But it cannot work for func g() definition in CONF1 since a "," is appended unexpectedly.
> 
> It seems to be really simple, but I am now a little dumb how to define one and solve it elegantly.
> 
> thanks. 
> 
> tom

[toc] | [prev] | [next] | [standalone]

#15076

On 2012-3-27 21:44, thomas wrote:
> Ok...

Please top quote. And keep your line width within 76
characters.

> And I can write the two versions of the functions with the following declaration.
> int f(PARAM_WITH_COMMA int y);
> int f(PARAM_NO_COMMA);
>
> In short I want to make my code neat and easy to read, that's to say, fewer "#ifdef" like stuff.
>
> So I'm trying to write some macros to shorten some word usage.

Using macro is neither neat nor easy to read.
Whenever you can avoid macros, just avoid them.

You are using C++. C++ can do most stuff that macro can do.
When you want to write a macro, rethink again.


> I just gave an example above using two marco definitions.
>
> Can I use just one macro to achieve the same thing?

There is comma_if in Boost library. You may check Boost
preprocessor library.


-- 
WQ

[toc] | [prev] | [next] | [standalone]

#15079

Qi <no@no.com> wrote:
> C++ can do most stuff that macro can do.

  Not everything, though. For instance, there's a reason why assert()
is still a macro in C++. (Because it typically prints the file name
and line number where the assertion failed, plus the expression that
failed, that's basically impossible to achieve without a preprocessor
macro. Even if there was a way to get the line number otherwise (which
there isn't), you still couldn't stringify the expression without the
preprocessor.)

[toc] | [prev] | [next] | [standalone]

#15077

On Tuesday, March 27, 2012 3:29:11 PM UTC+8, thomas wrote:
> Hi, 
>  I'm stucked by something appears really simple at first sight.
> 
> I have a function definition with different parameter list in different compile configurations. for example.
> 
> --------------------------
> CONF1:
> int f(int x, int y){}
> int g(int x){}
> 
> CONF2:
> int f(int y){}
> int g(){}
> ----------------------------
> I'm trying to define some marco like
> 
> #if CONF1
> #define PARAM(list) int x, list
> #elseif CONF2
> #define PARAM(list) list
> #endif
> 
> But it cannot work for func g() definition in CONF1 since a "," is appended unexpectedly.
> 
> It seems to be really simple, but I am now a little dumb how to define one and solve it elegantly.
> 
> thanks. 
> 
> tom

Seems that you don't understand what I am talking about, just judging by some keyword mentioned. 

[toc] | [prev] | [next] | [standalone]

#15078

On Tue, 27 Mar 2012 07:05:35 -0700 (PDT), thomas
<freshthomas@gmail.com> wrote:

>On Tuesday, March 27, 2012 3:29:11 PM UTC+8, thomas wrote:
>> Hi, 
>>  I'm stucked by something appears really simple at first sight.
>> 
>
>Seems that you don't understand what I am talking about, just judging by some keyword mentioned. 

Try replying to the people your post is intended to reply to instead
of yourself.

You have not clearly stated your problem.

If you are coding C++ you do not need to write macros to define your
functions using compile time switches. You can simply define your
functions simultaneously in the same scope and let the compiler choose
which overloaded function to use based on the function call.

Defining:

int f(int x, int y){}
int f(int x){}
int f(){}

Calling:

i = f(x,y);
i = f(x);
i = f();

The compiler will choose the proper function that matches the
arguments in the function call.

[toc] | [prev] | [next] | [standalone]

#15080

On March, 27th, Geoff wrote:

[snip]

> Try replying to the people your post is intended to reply to instead
> of yourself.

[snip]

Gosh, it took me almost one minute to figure out what your sentence
actually means. I take it from your profile that you are a native
speaker, so I assume that such sentences are nothing strange. Still, I
wonder whether this is considered good English or not. A comma would
be a great hint for non-native speakers, but I guess that your
sentence is grammatically correct.

Regards,
Stuart

[toc] | [prev] | [next] | [standalone]

#15083

On Tue, 27 Mar 2012 11:27:14 -0700 (PDT), Stuart Redmann
<DerTopper@web.de> wrote:

>On March, 27th, Geoff wrote:
>
>[snip]
>
>> Try replying to the people your post is intended to reply to instead
>> of yourself.
>
>[snip]
>
>Gosh, it took me almost one minute to figure out what your sentence
>actually means. I take it from your profile that you are a native
>speaker, so I assume that such sentences are nothing strange. Still, I
>wonder whether this is considered good English or not. A comma would
>be a great hint for non-native speakers, but I guess that your
>sentence is grammatically correct.
>

Stuart,
I pumped it through LibreOffice just to see if it would flag the
grammar, it did not.

However, I suppose it could have been phrased better and more
formally:

"You should reply and quote the people to whom you are intending to
reply, rather than replying to yourself."

Better?

[toc] | [prev] | [next] | [standalone]

#15085

On 27 Mrz., 20:46, Geoff <ge...@invalid.invalid> wrote:
> On Tue, 27 Mar 2012 11:27:14 -0700 (PDT), Stuart Redmann
>
>
>
>
>
>
>
>
>
> <DerTop...@web.de> wrote:
> >On March, 27th, Geoff wrote:
>
> >[snip]
>
> >> Try replying to the people your post is intended to reply to instead
> >> of yourself.
>
> >[snip]
>
> >Gosh, it took me almost one minute to figure out what your sentence
> >actually means. I take it from your profile that you are a native
> >speaker, so I assume that such sentences are nothing strange. Still, I
> >wonder whether this is considered good English or not. A comma would
> >be a great hint for non-native speakers, but I guess that your
> >sentence is grammatically correct.
>
> Stuart,
> I pumped it through LibreOffice just to see if it would flag the
> grammar, it did not.
>
> However, I suppose it could have been phrased better and more
> formally:
>
> "You should reply and quote the people to whom you are intending to
> reply, rather than replying to yourself."
>
> Better?

Still looks weird to me.

"You should reply to the people to whom you are intending to reply,
rather than replying to yourself." looks somewhat better, even though
it may not be grammatically correct. Besides, it does not contain the
quoting part. Maybe like this:"You should reply to and quote the
people to whom you are intending to reply, rather than replying to
yourself."

Would that be OK?

Thanks,
Stuart

[toc] | [prev] | [next] | [standalone]

#15086

On 03/28/12 08:39 AM, Stuart Redmann wrote:
>
> "You should reply to the people to whom you are intending to reply,
> rather than replying to yourself." looks somewhat better, even though
> it may not be grammatically correct. Besides, it does not contain the
> quoting part. Maybe like this:"You should reply to and quote the
> people to whom you are intending to reply, rather than replying to
> yourself."
>
> Would that be OK?

Or simply direct the OP to

http://www.parashift.com/c++-faq-lite/how-to-post.html#faq-5.4

-- 
Ian Collins

[toc] | [prev] | [next] | [standalone]

#15088

On 27 Mrz., Ian Collins wrote:

[snip]

> Or simply direct the OP to
>
> http://www.parashift.com/c++-faq-lite/how-to-post.html#faq-5.4

Acknowledged. However, I was intending to improve my English.



In that matter, I'd rather say
"You should reply to the people whom you are intending to reply to,
[...]"
than
"You should reply to the people to whom you are intending to reply,
[...]"

Would that still be correct?

Regards,
Stuart

[toc] | [prev] | [next] | [standalone]

#15089

On Tue, 27 Mar 2012 12:58:21 -0700 (PDT), Stuart Redmann
<DerTopper@web.de> wrote:

>On 27 Mrz., Ian Collins wrote:
>
>[snip]
>
>> Or simply direct the OP to
>>
>> http://www.parashift.com/c++-faq-lite/how-to-post.html#faq-5.4
>
>Acknowledged. However, I was intending to improve my English.
>
>
>
>In that matter, I'd rather say
>"You should reply to the people whom you are intending to reply to,
>[...]"
>than
>"You should reply to the people to whom you are intending to reply,
>[...]"
>
>Would that still be correct?
>
>Regards,
>Stuart

I have to say the latter is more correct. putting the 'to' after
'reply' is splitting the infinitive. The 'to' belongs to 'whom' and
not to the verb 'reply'. But it is also more formal. You were correct
when you called me out on the 'to reply to'. You should probably also
discard the 'to the people'.

"You should reply to whom you are intending to reply, rather than
replying to yourself."

[toc] | [prev] | [next] | [standalone]

#15090

Geoff <geoff@invalid.invalid> wrote:
> "You should reply and quote the people to whom you are intending to
> reply, rather than replying to yourself."

  "Rather than replying to yourself, you should reply to the people you
are intending to reply. It makes it clearer who you are talking to."

[toc] | [prev] | [next] | [standalone]

#15154

On 28 Mrz., Juha Nieminen wrote:
> Geoff <ge...@invalid.invalid> wrote:
> > "You should reply and quote the people to whom you are intending to
> > reply, rather than replying to yourself."
>
>   "Rather than replying to yourself, you should reply to the people you
> are intending to reply. It makes it clearer who you are talking to."

The clearest so far.


I'm still thinking about which kind of trophy I should award. Were
this an MSDN forum, I could give you a positive vote, but I guess that
this topic would never have been approved by MS.

Regards,
Stuart

[toc] | [prev] | [next] | [standalone]

#15149

On Wednesday, March 28, 2012 12:49:09 AM UTC+8, Geoff wrote:
> On Tue, 27 Mar 2012 07:05:35 -0700 (PDT), thomas
> <freshthomas@gmail.com> wrote:
> 
> >On Tuesday, March 27, 2012 3:29:11 PM UTC+8, thomas wrote:
> >> Hi, 
> >>  I'm stucked by something appears really simple at first sight.
> >> 
> >
> >Seems that you don't understand what I am talking about, just judging by some keyword mentioned. 
> 
> Try replying to the people your post is intended to reply to instead
> of yourself.
> 
> You have not clearly stated your problem.
> 
> If you are coding C++ you do not need to write macros to define your
> functions using compile time switches. You can simply define your
> functions simultaneously in the same scope and let the compiler choose
> which overloaded function to use based on the function call.
> 
> Defining:
> 
> int f(int x, int y){}
> int f(int x){}
> int f(){}
> 
> Calling:
> 
> i = f(x,y);
> i = f(x);
> i = f();
> 
> The compiler will choose the proper function that matches the
> arguments in the function call.

Well, I know that overloading can make the same function name work for different parameter list. But I want to see different functions for different configurations. If I define both, it's confusing which is used when. Sometimes the parameter in difference is just meaningness in some configurations.

[toc] | [prev] | [standalone]

Back to top | Article view | comp.lang.c++

csiph-web