Is there a C compiler that does not compile this?

advertisements

I was hanging out in my profiler for a while trying to figure out how to speed up a common log parser which was bottlenecked around the date parsing, and I tried various algorithms to speed things up.

The thing I tried that was fastest for me was also by far the most readable, but potentially non-standard C.

This worked quite well in GCC, icc, and my really old and picky SGI compiler. As it's a quite readable optimization, where doesn't it do what I want?

static int parseMonth(const char *input) {
    int rv=-1;
    int inputInt=0;
    int i=0;

    for(i=0; i<4 && input[i]; i++) {
        inputInt = (inputInt << 8) | input[i];
    }

    switch(inputInt) {
        case 'Jan/': rv=0; break;
        case 'Feb/': rv=1; break;
        case 'Mar/': rv=2; break;
        case 'Apr/': rv=3; break;
        case 'May/': rv=4; break;
        case 'Jun/': rv=5; break;
        case 'Jul/': rv=6; break;
        case 'Aug/': rv=7; break;
        case 'Sep/': rv=8; break;
        case 'Oct/': rv=9; break;
        case 'Nov/': rv=10; break;
        case 'Dec/': rv=11; break;
    }
    return rv;
}


Solaris 10 - SPARC - SUN Compiler.

Test code:

#include <stdio.h>

static int parseMonth(const char *input) {
    int rv=-1;
    int inputInt=0;
    int i=0;

    for(i=0; i<4 && input[i]; i++) {
        inputInt = (inputInt << 8) | input[i];
    }

    switch(inputInt) {
        case 'Jan/': rv=0; break;
        case 'Feb/': rv=1; break;
        case 'Mar/': rv=2; break;
        case 'Apr/': rv=3; break;
        case 'May/': rv=4; break;
        case 'Jun/': rv=5; break;
        case 'Jul/': rv=6; break;
        case 'Aug/': rv=7; break;
        case 'Sep/': rv=8; break;
        case 'Oct/': rv=9; break;
        case 'Nov/': rv=10; break;
        case 'Dec/': rv=11; break;
    }

    return rv;
}

static const struct
{
    char *data;
    int   result;
} test_case[] =
{
    { "Jan/", 0 },
    { "Feb/", 1 },
    { "Mar/", 2 },
    { "Apr/", 3 },
    { "May/", 4 },
    { "Jun/", 5 },
    { "Jul/", 6 },
    { "Aug/", 7 },
    { "Sep/", 8 },
    { "Oct/", 9 },
    { "Nov/", 10 },
    { "Dec/", 11 },
    { "aJ/n", -1 },
};

#define DIM(x) (sizeof(x)/sizeof(*(x)))

int main(void)
{
    size_t i;
    int    result;

    for (i = 0; i < DIM(test_case); i++)
    {
        result = parseMonth(test_case[i].data);
        if (result != test_case[i].result)
            printf("!! FAIL !! %s (got %d, wanted %d)\n",
                   test_case[i].data, result, test_case[i].result);
    }
    return(0);
}

Results (GCC 3.4.2 and Sun):

$ gcc -O xx.c -o xx
xx.c:14:14: warning: multi-character character constant
xx.c:15:14: warning: multi-character character constant
xx.c:16:14: warning: multi-character character constant
xx.c:17:14: warning: multi-character character constant
xx.c:18:14: warning: multi-character character constant
xx.c:19:14: warning: multi-character character constant
xx.c:20:14: warning: multi-character character constant
xx.c:21:14: warning: multi-character character constant
xx.c:22:14: warning: multi-character character constant
xx.c:23:14: warning: multi-character character constant
xx.c:24:14: warning: multi-character character constant
xx.c:25:14: warning: multi-character character constant
$ ./xx
$ cc -o xx xx.c
$ ./xx
!! FAIL !! Jan/ (got -1, wanted 0)
!! FAIL !! Feb/ (got -1, wanted 1)
!! FAIL !! Mar/ (got -1, wanted 2)
!! FAIL !! Apr/ (got -1, wanted 3)
!! FAIL !! May/ (got -1, wanted 4)
!! FAIL !! Jun/ (got -1, wanted 5)
!! FAIL !! Jul/ (got -1, wanted 6)
!! FAIL !! Aug/ (got -1, wanted 7)
!! FAIL !! Sep/ (got -1, wanted 8)
!! FAIL !! Oct/ (got -1, wanted 9)
!! FAIL !! Nov/ (got -1, wanted 10)
!! FAIL !! Dec/ (got -1, wanted 11)
$

Note that the last test case still passed - that is, it generated a -1.

Here's a revised - more verbose - version of parseMonth() which does work the same under both GCC and Sun C compiler:

#include <stdio.h>

/* MONTH_CODE("Jan/") does not reduce to an integer constant */
#define MONTH_CODE(x)   ((((((x[0]<<8)|x[1])<<8)|x[2])<<8)|x[3])

#define MONTH_JAN       (((((('J'<<8)|'a')<<8)|'n')<<8)|'/')
#define MONTH_FEB       (((((('F'<<8)|'e')<<8)|'b')<<8)|'/')
#define MONTH_MAR       (((((('M'<<8)|'a')<<8)|'r')<<8)|'/')
#define MONTH_APR       (((((('A'<<8)|'p')<<8)|'r')<<8)|'/')
#define MONTH_MAY       (((((('M'<<8)|'a')<<8)|'y')<<8)|'/')
#define MONTH_JUN       (((((('J'<<8)|'u')<<8)|'n')<<8)|'/')
#define MONTH_JUL       (((((('J'<<8)|'u')<<8)|'l')<<8)|'/')
#define MONTH_AUG       (((((('A'<<8)|'u')<<8)|'g')<<8)|'/')
#define MONTH_SEP       (((((('S'<<8)|'e')<<8)|'p')<<8)|'/')
#define MONTH_OCT       (((((('O'<<8)|'c')<<8)|'t')<<8)|'/')
#define MONTH_NOV       (((((('N'<<8)|'o')<<8)|'v')<<8)|'/')
#define MONTH_DEC       (((((('D'<<8)|'e')<<8)|'c')<<8)|'/')

static int parseMonth(const char *input) {
    int rv=-1;
    int inputInt=0;
    int i=0;

    for(i=0; i<4 && input[i]; i++) {
        inputInt = (inputInt << 8) | input[i];
    }

    switch(inputInt) {
        case MONTH_JAN: rv=0; break;
        case MONTH_FEB: rv=1; break;
        case MONTH_MAR: rv=2; break;
        case MONTH_APR: rv=3; break;
        case MONTH_MAY: rv=4; break;
        case MONTH_JUN: rv=5; break;
        case MONTH_JUL: rv=6; break;
        case MONTH_AUG: rv=7; break;
        case MONTH_SEP: rv=8; break;
        case MONTH_OCT: rv=9; break;
        case MONTH_NOV: rv=10; break;
        case MONTH_DEC: rv=11; break;
    }

    return rv;
}

static const struct
{
    char *data;
    int   result;
} test_case[] =
{
    { "Jan/", 0 },
    { "Feb/", 1 },
    { "Mar/", 2 },
    { "Apr/", 3 },
    { "May/", 4 },
    { "Jun/", 5 },
    { "Jul/", 6 },
    { "Aug/", 7 },
    { "Sep/", 8 },
    { "Oct/", 9 },
    { "Nov/", 10 },
    { "Dec/", 11 },
    { "aJ/n", -1 },
    { "/naJ", -1 },
};

#define DIM(x) (sizeof(x)/sizeof(*(x)))

int main(void)
{
    size_t i;
    int    result;

    for (i = 0; i < DIM(test_case); i++)
    {
        result = parseMonth(test_case[i].data);
        if (result != test_case[i].result)
            printf("!! FAIL !! %s (got %d, wanted %d)\n",
                   test_case[i].data, result, test_case[i].result);
    }
    return(0);
}

I wanted to use MONTH_CODE() but the compilers did not cooperate.