Ævar Arnfjörð Bjarmason <ava...@gmail.com> writes:

>> +CFLAGS += -pedantic
>> +# don't warn for each N_ use
>> +CFLAGS += -DUSE_PARENS_AROUND_GETTEXT_N=0
>> +endif
>
> ...and set this to "no" not "0" since we document that that's the way to
> toggle it off in the Makefile, i.e. let's be consistent.

The Make variable USE_PARENS_AROUND_GETTEXT_N is described as taking
"yes" or "no".

    # Define USE_PARENS_AROUND_GETTEXT_N to "yes" if your compiler happily
    # compiles the following initialization:
    #
    #   static const char s[] = ("FOO");
    #
    # and define it to "no" if you need to remove the parentheses () around the
    # constant.  The default is "auto", which means to use parentheses if your
    # compiler is detected to support it.

But the knob on the CFLAGS set by these variables take 1 or 0

    ifeq (yes,$(USE_PARENS_AROUND_GETTEXT_N))
            BASIC_CFLAGS += -DUSE_PARENS_AROUND_GETTEXT_N=1
    else
    ifeq (no,$(USE_PARENS_AROUND_GETTEXT_N))
            BASIC_CFLAGS += -DUSE_PARENS_AROUND_GETTEXT_N=0
    endif
    endif

And the code that uses the CFLAGS knob 

    /* Mark msgid for translation but do not translate it. */
    #if !USE_PARENS_AROUND_GETTEXT_N
    #define N_(msgid) msgid
    #else
    ...
    #define N_(msgid) (msgid)
    #endif

pays attention to the truth/false in usual C preprocessor sense.
Your "no" happens to serve as 0 just like "yes" would.

So I think you suggestion is a bad one that makes a misleading
result.

[Footnote]

*1* The following shows all "not X" except for "not one".

#include <stdio.h>

#define ZERO 0
#define ONE 1
#define YES yes
#define NO no
#undef UNDEF

const char *msgs[] = {
#if !ZERO
        "not zero",
#endif
#if !ONE
        "not one",
#endif
#if !YES
        "not yes",
#endif
#if !NO
        "not no",
#endif
#if !UNDEF
        "not undef",
#endif
        NULL
};

int main(int ac, char **av)
{
        const char **cp = msgs;

        while (*cp) {
                printf("%s\n", *cp);
                cp++;
        }
        return 0;
}




Reply via email to