void
gawk_exit(int status)
{
if (fatal_tag_valid) {
exit_val = status;
longjmp(fatal_tag, 1);
}
final_exit(status);
}
int
main(int argc, char **argv)
{
/*
* The + on the front tells GNU getopt not to rearrange argv.
*/
const char *optlist = "+F:f:v:W;m:bcCd::D::e:E:gh:i:l:L:nNo::Op::MPrStVY";
bool stopped_early = false;
int old_optind;
int i;
int c;
char *scan, *src;
char *extra_stack;
int have_srcfile = 0;
SRCFILE *s;
/* do these checks early */
if (getenv("TIDYMEM") != NULL)
do_flags |= DO_TIDY_MEM;
#ifdef HAVE_MCHECK_H
#ifdef HAVE_MTRACE
if (do_tidy_mem)
mtrace();
#endif /* HAVE_MTRACE */
#endif /* HAVE_MCHECK_H */
#if defined(LC_CTYPE)
setlocale(LC_CTYPE, "");
#endif
#if defined(LC_COLLATE)
setlocale(LC_COLLATE, "");
#endif
#if defined(LC_MESSAGES)
setlocale(LC_MESSAGES, "");
#endif
#if defined(LC_NUMERIC) && defined(HAVE_LOCALE_H)
/*
* Force the issue here. According to POSIX 2001, decimal
* point is used for parsing source code and for command-line
* assignments and the locale value for processing input,
* number to string conversion, and printing output.
*
* 10/2005 --- see below also; we now only use the locale's
* decimal point if do_posix in effect.
*
* 9/2007:
* This is a mess. We need to get the locale's numeric info for
* the thousands separator for the %'d flag.
*/
setlocale(LC_NUMERIC, "");
init_locale(& loc);
setlocale(LC_NUMERIC, "C");
#endif
#if defined(LC_TIME)
setlocale(LC_TIME, "");
#endif
#if MBS_SUPPORT
/*
* In glibc, MB_CUR_MAX is actually a function. This value is
* tested *a lot* in many speed-critical places in gawk. Caching
* this value once makes a speed difference.
*/
gawk_mb_cur_max = MB_CUR_MAX;
/* Without MBS_SUPPORT, gawk_mb_cur_max is 1. */
/* init the cache for checking bytes if they're characters */
init_btowc_cache();
#endif
(void) bindtextdomain(PACKAGE, LOCALEDIR);
(void) textdomain(PACKAGE);
(void) signal(SIGFPE, catchsig);
#ifdef SIGBUS
(void) signal(SIGBUS, catchsig);
#endif
(void) sigsegv_install_handler(catchsegv);
#define STACK_SIZE (16*1024)
emalloc(extra_stack, char *, STACK_SIZE, "main");
(void) stackoverflow_install_handler(catchstackoverflow, extra_stack, STACK_SIZE);
#undef STACK_SIZE
myname = gawk_name(argv[0]);
os_arg_fixup(&argc, &argv); /* emulate redirection, expand wildcards */
if (argc < 2)
usage(EXIT_FAILURE, stderr);
/* initialize the null string */
Nnull_string = make_string("", 0);
/* Robustness: check that file descriptors 0, 1, 2 are open */
init_fds();
/* init array handling. */
array_init();
/* init the symbol tables */
init_symbol_table();
output_fp = stdout;
/* we do error messages ourselves on invalid options */
opterr = false;
/* copy argv before getopt gets to it; used to restart the debugger */
save_argv(argc, argv);
/* initialize global (main) execution context */
push_context(new_context());
/* option processing. ready, set, go! */
for (optopt = 0, old_optind = 1;
(c = getopt_long(argc, argv, optlist, optab, NULL)) != EOF;
optopt = 0, old_optind = optind) {
if (do_posix)
opterr = true;
switch (c) {
case 'F':
add_preassign(PRE_ASSIGN_FS, optarg);
break;
case 'E':
disallow_var_assigns = true;
/* fall through */
case 'f':
/*
* Allow multiple -f options.
* This makes function libraries real easy.
* Most of the magic is in the scanner.
*
* The following is to allow for whitespace at the end
* of a #! /bin/gawk line in an executable file
*/
scan = optarg;
if (argv[optind-1] != optarg)
while (isspace((unsigned char) *scan))
scan++;
src = (*scan == '\0' ? argv[optind++] : optarg);
(void) add_srcfile((src && src[0] == '-' && src[1] == '\0') ?
SRC_STDIN : SRC_FILE,
src, srcfiles, NULL, NULL);
break;
case 'v':
add_preassign(PRE_ASSIGN, optarg);
break;
case 'm':
/*
* BWK awk extension.
* -mf nnn set # fields, gawk ignores
* -mr nnn set record length, ditto
*
* As of at least 10/2007, BWK awk also ignores it.
*/
if (do_lint)
lintwarn(_("`-m[fr]' option irrelevant in gawk"));
if (optarg[0] != 'r' && optarg[0] != 'f')
warning(_("-m option usage: `-m[fr] nnn'"));
break;
case 'b':
do_binary = true;
break;
case 'c':
do_flags |= DO_TRADITIONAL;
break;
case 'C':
copyleft();
break;
case 'd':
do_flags |= DO_DUMP_VARS;
if (optarg != NULL && optarg[0] != '\0')
varfile = optarg;
break;
case 'D':
do_flags |= DO_DEBUG;
if (optarg != NULL && optarg[0] != '\0')
command_file = optarg;
break;
case 'e':
if (optarg[0] == '\0')
warning(_("empty argument to `-e/--source' ignored"));
else
(void) add_srcfile(SRC_CMDLINE, optarg, srcfiles, NULL, NULL);
break;
case 'g':
do_flags |= DO_INTL;
break;
case 'h':
/* write usage to stdout, per GNU coding stds */
usage(EXIT_SUCCESS, stdout);
break;
case 'i':
(void) add_srcfile(SRC_INC, optarg, srcfiles, NULL, NULL);
break;
case 'l':
(void) add_srcfile(SRC_EXTLIB, optarg, srcfiles, NULL, NULL);
break;
case 'L':
#ifndef NO_LINT
do_flags |= DO_LINT_ALL;
if (optarg != NULL) {
if (strcmp(optarg, "fatal") == 0)
lintfunc = r_fatal;
else if (strcmp(optarg, "invalid") == 0) {
do_flags &= ~DO_LINT_ALL;
do_flags |= DO_LINT_INVALID;
}
}
break;
case 't':
do_flags |= DO_LINT_OLD;
break;
#else
case 'L':
case 't':
break;
#endif
case 'n':
do_flags |= DO_NON_DEC_DATA;
break;
case 'N':
use_lc_numeric = true;
break;
case 'O':
do_optimize++;
break;
case 'p':
do_flags |= DO_PROFILE;
/* fall through */
case 'o':
do_flags |= DO_PRETTY_PRINT;
if (optarg != NULL)
set_prof_file(optarg);
else
set_prof_file(DEFAULT_PROFILE);
break;
case 'M':
#ifdef HAVE_MPFR
do_flags |= DO_MPFR;
#endif
break;
case 'P':
do_flags |= DO_POSIX;
break;
case 'r':
do_flags |= DO_INTERVALS;
break;
case 'S':
do_flags |= DO_SANDBOX;
break;
case 'V':
do_version = true;
break;
case 'W': /* gawk specific options - now in getopt_long */
fprintf(stderr, _("%s: option `-W %s' unrecognized, ignored\n"),
argv[0], optarg);
break;
case 0:
/*
* getopt_long found an option that sets a variable
* instead of returning a letter. Do nothing, just
* cycle around for the next one.
*/
break;
case 'Y':
#if defined(YYDEBUG) || defined(GAWKDEBUG)
if (c == 'Y') {
yydebug = 2;
break;
}
#endif
/* if not debugging, fall through */
case '?':
default:
/*
* If not posix, an unrecognized option stops argument
* processing so that it can go into ARGV for the awk
* program to see. This makes use of ``#! /bin/gawk -f''
* easier.
*
* However, it's never simple. If optopt is set,
* an option that requires an argument didn't get the
* argument. We care because if opterr is 0, then
* getopt_long won't print the error message for us.
*/
if (! do_posix
&& (optopt == '\0' || strchr(optlist, optopt) == NULL)) {
/*
* can't just do optind--. In case of an
* option with >= 2 letters, getopt_long
* won't have incremented optind.
*/
optind = old_optind;
stopped_early = true;
goto out;
} else if (optopt != '\0') {
/* Use POSIX required message format */
fprintf(stderr,
_("%s: option requires an argument -- %c\n"),
myname, optopt);
usage(EXIT_FAILURE, stderr);
}
/* else
let getopt print error message for us */
break;
}
if (c == 'E') /* --exec ends option processing */
break;
}
out:
if (do_nostalgia)
nostalgia();
/* check for POSIXLY_CORRECT environment variable */
if (! do_posix && getenv("POSIXLY_CORRECT") != NULL) {
do_flags |= DO_POSIX;
if (do_lint)
lintwarn(
_("environment variable `POSIXLY_CORRECT' set: turning on `--posix'"));
}
if (do_posix) {
use_lc_numeric = true;
if (do_traditional) /* both on command line */
warning(_("`--posix' overrides `--traditional'"));
else
do_flags |= DO_TRADITIONAL;
/*
* POSIX compliance also implies
* no GNU extensions either.
*/
}
if (do_traditional && do_non_decimal_data) {
do_flags &= ~DO_NON_DEC_DATA;
warning(_("`--posix'/`--traditional' overrides `--non-decimal-data'"));
}
if (do_lint && os_is_setuid())
warning(_("running %s setuid root may be a security problem"), myname);
#if MBS_SUPPORT
if (do_binary) {
if (do_posix)
warning(_("`--posix' overrides `--characters-as-bytes'"));
else
gawk_mb_cur_max = 1; /* hands off my data! */
#if defined(LC_ALL)
setlocale(LC_ALL, "C");
#endif
}
#endif
if (do_debug) /* Need to register the debugger pre-exec hook before any other */
init_debug();
#ifdef HAVE_MPFR
/* Set up MPFR defaults, and register pre-exec hook to process arithmetic opcodes */
if (do_mpfr)
init_mpfr(DEFAULT_PREC, DEFAULT_ROUNDMODE);
#endif
/* load group set */
init_groupset();
#ifdef HAVE_MPFR
if (do_mpfr) {
mpz_init(Nnull_string->mpg_i);
Nnull_string->flags = (MALLOC|STRCUR|STRING|MPZN|NUMCUR|NUMBER);
} else
#endif
{
Nnull_string->numbr = 0.0;
Nnull_string->flags = (MALLOC|STRCUR|STRING|NUMCUR|NUMBER);
}
/*
* Tell the regex routines how they should work.
* Do this before initializing variables, since
* they could want to do a regexp compile.
*/
resetup();
/* Set up the special variables */
init_vars();
/* Set up the field variables */
init_fields();
/* Now process the pre-assignments */
for (i = 0; i <= numassigns; i++) {
if (preassigns[i].type == PRE_ASSIGN)
(void) arg_assign(preassigns[i].val, true);
else /* PRE_ASSIGN_FS */
cmdline_fs(preassigns[i].val);
efree(preassigns[i].val);
}
if (preassigns != NULL)
efree(preassigns);
if ((BINMODE & 1) != 0)
if (os_setbinmode(fileno(stdin), O_BINARY) == -1)
fatal(_("can't set binary mode on stdin (%s)"), strerror(errno));
if ((BINMODE & 2) != 0) {
if (os_setbinmode(fileno(stdout), O_BINARY) == -1)
fatal(_("can't set binary mode on stdout (%s)"), strerror(errno));
if (os_setbinmode(fileno(stderr), O_BINARY) == -1)
fatal(_("can't set binary mode on stderr (%s)"), strerror(errno));
}
#ifdef GAWKDEBUG
setbuf(stdout, (char *) NULL); /* make debugging easier */
#endif
if (os_isatty(fileno(stdout)))
output_is_tty = true;
/* initialize API before loading extension libraries */
init_ext_api();
/* load extension libs */
for (s = srcfiles->next; s != srcfiles; s = s->next) {
if (s->stype == SRC_EXTLIB)
load_ext(s->fullpath);
else if (s->stype != SRC_INC)
have_srcfile++;
}
/* do version check after extensions are loaded to get extension info */
if (do_version)
version();
/* No -f or --source options, use next arg */
if (! have_srcfile) {
if (optind > argc - 1 || stopped_early) /* no args left or no program */
usage(EXIT_FAILURE, stderr);
(void) add_srcfile(SRC_CMDLINE, argv[optind], srcfiles, NULL, NULL);
optind++;
}
/* Select the interpreter routine */
init_interpret();
init_args(optind, argc,
do_posix ? argv[0] : myname,
argv);
#if defined(LC_NUMERIC)
/*
* FRAGILE! CAREFUL!
* Pre-initing the variables with arg_assign() can change the
* locale. Force it to C before parsing the program.
*/
setlocale(LC_NUMERIC, "C");
#endif
/* Read in the program */
if (parse_program(& code_block) != 0)
exit(EXIT_FAILURE);
if (do_intl)
exit(EXIT_SUCCESS);
if (do_lint)
shadow_funcs();
if (do_lint && code_block->nexti->opcode == Op_atexit)
lintwarn(_("no program text at all!"));
load_symbols();
if (do_profile)
init_profiling_signals();
#if defined(LC_NUMERIC)
/*
* See comment above about using locale's decimal point.
*
* 10/2005:
* Bitter experience teaches us that most people the world over
* use period as the decimal point, not whatever their locale
* uses. Thus, only use the locale's decimal point if being
* posixly anal-retentive.
*
* 7/2007:
* Be a little bit kinder. Allow the --use-lc-numeric option
* to also use the local decimal point. This avoids the draconian
* strictness of POSIX mode if someone just wants to parse their
* data using the local decimal point.
*/
if (use_lc_numeric)
setlocale(LC_NUMERIC, "");
#endif
init_io();
output_fp = stdout;
if (do_debug)
debug_prog(code_block);
else
interpret(code_block);
if (do_pretty_print) {
dump_prog(code_block);
dump_funcs();
}
if (do_dump_vars)
dump_vars(varfile);
if (do_tidy_mem)
release_all_vars();
/* keep valgrind happier */
if (extra_stack)
efree(extra_stack);
final_exit(exit_val);
return exit_val; /* to suppress warnings */
}