1 /* expr -- evaluate expressions.
2 Copyright (C) 1986-2012 Free Software Foundation, Inc.
4 This program is free software: you can redistribute it and/or modify
5 it under the terms of the GNU General Public License as published by
6 the Free Software Foundation, either version 3 of the License, or
7 (at your option) any later version.
9 This program is distributed in the hope that it will be useful,
10 but WITHOUT ANY WARRANTY; without even the implied warranty of
11 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
12 GNU General Public License for more details.
14 You should have received a copy of the GNU General Public License
15 along with this program. If not, see <http://www.gnu.org/licenses/>. */
17 /* Author: Mike Parker.
18 Modified for arbitrary-precision calculation by James Youngman.
20 This program evaluates expressions. Each token (operator, operand,
21 parenthesis) of the expression must be a seperate argument. The
22 parser used is a reasonably general one, though any incarnation of
23 it is language-specific. It is especially nice for expressions.
25 No parse tree is needed; a new node is evaluated immediately.
26 One function can handle multiple operators all of equal precedence,
27 provided they all associate ((x op x) op x).
29 Define EVAL_TRACE to print an evaluation trace. */
33 #include <sys/types.h>
38 #include "long-options.h"
40 #include "strnumcmp.h"
43 /* Various parts of this code assume size_t fits into unsigned long
44 int, the widest unsigned type that GMP supports. */
45 verify (SIZE_MAX
<= ULONG_MAX
);
47 static void integer_overflow (char) ATTRIBUTE_NORETURN
;
56 /* Approximate gmp.h well enough for expr.c's purposes. */
57 typedef intmax_t mpz_t
[1];
58 static void mpz_clear (mpz_t z
) { (void) z
; }
59 static void mpz_init_set_ui (mpz_t z
, unsigned long int i
) { z
[0] = i
; }
61 mpz_init_set_str (mpz_t z
, char *s
, int base
)
63 return xstrtoimax (s
, NULL
, base
, z
, NULL
) == LONGINT_OK
? 0 : -1;
66 mpz_add (mpz_t r
, mpz_t a0
, mpz_t b0
)
71 if ((val
< a
) != (b
< 0))
72 integer_overflow ('+');
76 mpz_sub (mpz_t r
, mpz_t a0
, mpz_t b0
)
81 if ((a
< val
) != (b
< 0))
82 integer_overflow ('-');
86 mpz_mul (mpz_t r
, mpz_t a0
, mpz_t b0
)
91 if (! (a
== 0 || b
== 0
92 || ((val
< 0) == ((a
< 0) ^ (b
< 0)) && val
/ a
== b
)))
93 integer_overflow ('*');
97 mpz_tdiv_q (mpz_t r
, mpz_t a0
, mpz_t b0
)
102 /* Some x86-style hosts raise an exception for INT_MIN / -1. */
103 if (a
< - INTMAX_MAX
&& b
== -1)
104 integer_overflow ('/');
108 mpz_tdiv_r (mpz_t r
, mpz_t a0
, mpz_t b0
)
113 /* Some x86-style hosts raise an exception for INT_MIN % -1. */
114 r
[0] = a
< - INTMAX_MAX
&& b
== -1 ? 0 : a
% b
;
117 mpz_get_str (char const *str
, int base
, mpz_t z
)
119 (void) str
; (void) base
;
120 char buf
[INT_BUFSIZE_BOUND (intmax_t)];
121 return xstrdup (imaxtostr (z
[0], buf
));
126 return z
[0] < 0 ? -1 : 0 < z
[0];
129 mpz_fits_ulong_p (mpz_t z
)
131 return 0 <= z
[0] && z
[0] <= ULONG_MAX
;
133 static unsigned long int
139 mpz_out_str (FILE *stream
, int base
, mpz_t z
)
142 char buf
[INT_BUFSIZE_BOUND (intmax_t)];
143 return fputs (imaxtostr (z
[0], buf
), stream
) != EOF
;
147 /* The official name of this program (e.g., no 'g' prefix). */
148 #define PROGRAM_NAME "expr"
151 proper_name ("Mike Parker"), \
152 proper_name ("James Youngman"), \
153 proper_name ("Paul Eggert")
158 /* Invalid expression: e.g., its form does not conform to the
159 grammar for expressions. Our grammar is an extension of the
163 /* An internal error occurred, e.g., arithmetic overflow, storage
168 /* The kinds of value we can have. */
174 typedef enum valtype TYPE
;
179 TYPE type
; /* Which kind. */
181 { /* The value itself. */
186 typedef struct valinfo VALUE
;
188 /* The arguments given to the program, minus the program name. */
191 static VALUE
*eval (bool);
192 static bool nomoreargs (void);
193 static bool null (VALUE
*v
);
194 static void printv (VALUE
*v
);
199 if (status
!= EXIT_SUCCESS
)
204 Usage: %s EXPRESSION\n\
207 program_name
, program_name
);
209 fputs (HELP_OPTION_DESCRIPTION
, stdout
);
210 fputs (VERSION_OPTION_DESCRIPTION
, stdout
);
213 Print the value of EXPRESSION to standard output. A blank line below\n\
214 separates increasing precedence groups. EXPRESSION may be:\n\
216 ARG1 | ARG2 ARG1 if it is neither null nor 0, otherwise ARG2\n\
218 ARG1 & ARG2 ARG1 if neither argument is null or 0, otherwise 0\n\
222 ARG1 < ARG2 ARG1 is less than ARG2\n\
223 ARG1 <= ARG2 ARG1 is less than or equal to ARG2\n\
224 ARG1 = ARG2 ARG1 is equal to ARG2\n\
225 ARG1 != ARG2 ARG1 is unequal to ARG2\n\
226 ARG1 >= ARG2 ARG1 is greater than or equal to ARG2\n\
227 ARG1 > ARG2 ARG1 is greater than ARG2\n\
231 ARG1 + ARG2 arithmetic sum of ARG1 and ARG2\n\
232 ARG1 - ARG2 arithmetic difference of ARG1 and ARG2\n\
234 /* Tell xgettext that the "% A" below is not a printf-style
235 format string: xgettext:no-c-format */
238 ARG1 * ARG2 arithmetic product of ARG1 and ARG2\n\
239 ARG1 / ARG2 arithmetic quotient of ARG1 divided by ARG2\n\
240 ARG1 % ARG2 arithmetic remainder of ARG1 divided by ARG2\n\
244 STRING : REGEXP anchored pattern match of REGEXP in STRING\n\
246 match STRING REGEXP same as STRING : REGEXP\n\
247 substr STRING POS LENGTH substring of STRING, POS counted from 1\n\
248 index STRING CHARS index in STRING where any CHARS is found, or 0\n\
249 length STRING length of STRING\n\
252 + TOKEN interpret TOKEN as a string, even if it is a\n\
253 keyword like 'match' or an operator like '/'\n\
255 ( EXPRESSION ) value of EXPRESSION\n\
259 Beware that many operators need to be escaped or quoted for shells.\n\
260 Comparisons are arithmetic if both ARGs are numbers, else lexicographical.\n\
261 Pattern matches return the string matched between \\( and \\) or null; if\n\
262 \\( and \\) are not used, they return the number of characters matched or 0.\n\
266 Exit status is 0 if EXPRESSION is neither null nor 0, 1 if EXPRESSION is null\n\
267 or 0, 2 if EXPRESSION is syntactically invalid, and 3 if an error occurred.\n\
269 emit_ancillary_info ();
274 /* Report a syntax error and exit. */
278 error (EXPR_INVALID
, 0, _("syntax error"));
281 /* Report an integer overflow for operation OP and exit. */
283 integer_overflow (char op
)
285 error (EXPR_FAILURE
, ERANGE
, "%c", op
);
286 abort (); /* notreached */
289 static void die (int errno_val
, char const *msg
)
292 die (int errno_val
, char const *msg
)
294 error (EXPR_FAILURE
, errno_val
, "%s", msg
);
295 abort (); /* notreached */
299 main (int argc
, char **argv
)
303 initialize_main (&argc
, &argv
);
304 set_program_name (argv
[0]);
305 setlocale (LC_ALL
, "");
306 bindtextdomain (PACKAGE
, LOCALEDIR
);
307 textdomain (PACKAGE
);
309 initialize_exit_failure (EXPR_FAILURE
);
310 atexit (close_stdout
);
312 parse_long_options (argc
, argv
, PROGRAM_NAME
, PACKAGE_NAME
, VERSION
,
313 usage
, AUTHORS
, (char const *) NULL
);
315 /* The above handles --help and --version.
316 Since there is no other invocation of getopt, handle '--' here. */
317 unsigned int u_argc
= argc
;
318 if (1 < u_argc
&& STREQ (argv
[1], "--"))
326 error (0, 0, _("missing operand"));
327 usage (EXPR_INVALID
);
340 /* Return a VALUE for I. */
343 int_value (unsigned long int i
)
345 VALUE
*v
= xmalloc (sizeof *v
);
347 mpz_init_set_ui (v
->u
.i
, i
);
351 /* Return a VALUE for S. */
354 str_value (char const *s
)
356 VALUE
*v
= xmalloc (sizeof *v
);
358 v
->u
.s
= xstrdup (s
);
362 /* Free VALUE V, including structure components. */
367 if (v
->type
== string
)
382 mpz_out_str (stdout
, 10, v
->u
.i
);
393 /* Return true if V is a null-string or zero-number. */
395 static bool _GL_ATTRIBUTE_PURE
401 return mpz_sgn (v
->u
.i
) == 0;
404 char const *cp
= v
->u
.s
;
424 /* Return true if CP takes the form of an integer. */
426 static bool _GL_ATTRIBUTE_PURE
427 looks_like_integer (char const *cp
)
439 /* Coerce V to a string value (can't fail). */
448 char *s
= mpz_get_str (NULL
, 10, v
->u
.i
);
461 /* Coerce V to an integer value. Return true on success, false on failure. */
474 if (! looks_like_integer (s
))
476 if (mpz_init_set_str (v
->u
.i
, s
, 10) != 0 && !HAVE_GMP
)
477 error (EXPR_FAILURE
, ERANGE
, "%s", s
);
487 /* Extract a size_t value from an integer value I.
488 If the value is negative, return SIZE_MAX.
489 If the value is too large, return SIZE_MAX - 1. */
495 if (mpz_fits_ulong_p (i
))
497 unsigned long int ul
= mpz_get_ui (i
);
504 /* Return true and advance if the next token matches STR exactly.
505 STR must not be NULL. */
508 nextarg (char const *str
)
514 bool r
= STREQ (*args
, str
);
520 /* Return true if there no more tokens. */
529 /* Print evaluation trace and args remaining. */
538 for (a
= args
; *a
; a
++)
544 /* Do the : operator.
545 SV is the VALUE for the lhs (the string),
546 PV is the VALUE for the rhs (the pattern). */
549 docolon (VALUE
*sv
, VALUE
*pv
)
551 VALUE
*v
IF_LINT ( = NULL
);
553 struct re_pattern_buffer re_buffer
;
554 char fastmap
[UCHAR_MAX
+ 1];
555 struct re_registers re_regs
;
561 re_regs
.num_regs
= 0;
562 re_regs
.start
= NULL
;
565 re_buffer
.buffer
= NULL
;
566 re_buffer
.allocated
= 0;
567 re_buffer
.fastmap
= fastmap
;
568 re_buffer
.translate
= NULL
;
570 RE_SYNTAX_POSIX_BASIC
& ~RE_CONTEXT_INVALID_DUP
& ~RE_NO_EMPTY_RANGES
;
571 errmsg
= re_compile_pattern (pv
->u
.s
, strlen (pv
->u
.s
), &re_buffer
);
573 error (EXPR_INVALID
, 0, "%s", errmsg
);
574 re_buffer
.newline_anchor
= 0;
576 matchlen
= re_match (&re_buffer
, sv
->u
.s
, strlen (sv
->u
.s
), 0, &re_regs
);
579 /* Were \(...\) used? */
580 if (re_buffer
.re_nsub
> 0)
582 sv
->u
.s
[re_regs
.end
[1]] = '\0';
583 v
= str_value (sv
->u
.s
+ re_regs
.start
[1]);
586 v
= int_value (matchlen
);
588 else if (matchlen
== -1)
590 /* Match failed -- return the right kind of null. */
591 if (re_buffer
.re_nsub
> 0)
598 (matchlen
== -2 ? errno
: EOVERFLOW
),
599 _("error in regular expression matcher"));
601 if (0 < re_regs
.num_regs
)
603 free (re_regs
.start
);
606 re_buffer
.fastmap
= NULL
;
607 regfree (&re_buffer
);
611 /* Handle bare operands and ( expr ) syntax. */
614 eval7 (bool evaluate
)
635 return str_value (*args
++);
638 /* Handle match, substr, index, and length keywords, and quoting "+". */
641 eval6 (bool evaluate
)
656 return str_value (*args
++);
658 else if (nextarg ("length"))
660 r
= eval6 (evaluate
);
662 v
= int_value (strlen (r
->u
.s
));
666 else if (nextarg ("match"))
668 l
= eval6 (evaluate
);
669 r
= eval6 (evaluate
);
680 else if (nextarg ("index"))
684 l
= eval6 (evaluate
);
685 r
= eval6 (evaluate
);
688 pos
= strcspn (l
->u
.s
, r
->u
.s
);
689 v
= int_value (l
->u
.s
[pos
] ? pos
+ 1 : 0);
694 else if (nextarg ("substr"))
697 l
= eval6 (evaluate
);
698 i1
= eval6 (evaluate
);
699 i2
= eval6 (evaluate
);
701 llen
= strlen (l
->u
.s
);
703 if (!toarith (i1
) || !toarith (i2
))
707 size_t pos
= getsize (i1
->u
.i
);
708 size_t len
= getsize (i2
->u
.i
);
710 if (llen
< pos
|| pos
== 0 || len
== 0 || len
== SIZE_MAX
)
714 size_t vlen
= MIN (len
, llen
- pos
+ 1);
716 v
= xmalloc (sizeof *v
);
718 v
->u
.s
= xmalloc (vlen
+ 1);
719 vlim
= mempcpy (v
->u
.s
, l
->u
.s
+ pos
- 1, vlen
);
729 return eval7 (evaluate
);
732 /* Handle : operator (pattern matching).
733 Calls docolon to do the real work. */
736 eval5 (bool evaluate
)
745 l
= eval6 (evaluate
);
750 r
= eval6 (evaluate
);
764 /* Handle *, /, % operators. */
767 eval4 (bool evaluate
)
771 enum { multiply
, divide
, mod
} fxn
;
776 l
= eval5 (evaluate
);
781 else if (nextarg ("/"))
783 else if (nextarg ("%"))
787 r
= eval5 (evaluate
);
790 if (!toarith (l
) || !toarith (r
))
791 error (EXPR_INVALID
, 0, _("non-integer argument"));
792 if (fxn
!= multiply
&& mpz_sgn (r
->u
.i
) == 0)
793 error (EXPR_INVALID
, 0, _("division by zero"));
794 ((fxn
== multiply
? mpz_mul
795 : fxn
== divide
? mpz_tdiv_q
797 (l
->u
.i
, l
->u
.i
, r
->u
.i
));
803 /* Handle +, - operators. */
806 eval3 (bool evaluate
)
810 enum { plus
, minus
} fxn
;
815 l
= eval4 (evaluate
);
820 else if (nextarg ("-"))
824 r
= eval4 (evaluate
);
827 if (!toarith (l
) || !toarith (r
))
828 error (EXPR_INVALID
, 0, _("non-integer argument"));
829 (fxn
== plus
? mpz_add
: mpz_sub
) (l
->u
.i
, l
->u
.i
, r
->u
.i
);
835 /* Handle comparisons. */
838 eval2 (bool evaluate
)
845 l
= eval3 (evaluate
);
851 less_than
, less_equal
, equal
, not_equal
, greater_equal
, greater_than
857 else if (nextarg ("<="))
859 else if (nextarg ("=") || nextarg ("=="))
861 else if (nextarg ("!="))
863 else if (nextarg (">="))
865 else if (nextarg (">"))
869 r
= eval3 (evaluate
);
877 if (looks_like_integer (l
->u
.s
) && looks_like_integer (r
->u
.s
))
878 cmp
= strintcmp (l
->u
.s
, r
->u
.s
);
882 cmp
= strcoll (l
->u
.s
, r
->u
.s
);
886 error (0, errno
, _("string comparison failed"));
887 error (0, 0, _("set LC_ALL='C' to work around the problem"));
888 error (EXPR_INVALID
, 0,
889 _("the strings compared were %s and %s"),
890 quotearg_n_style (0, locale_quoting_style
, l
->u
.s
),
891 quotearg_n_style (1, locale_quoting_style
, r
->u
.s
));
897 case less_than
: val
= (cmp
< 0); break;
898 case less_equal
: val
= (cmp
<= 0); break;
899 case equal
: val
= (cmp
== 0); break;
900 case not_equal
: val
= (cmp
!= 0); break;
901 case greater_equal
: val
= (cmp
>= 0); break;
902 case greater_than
: val
= (cmp
> 0); break;
916 eval1 (bool evaluate
)
924 l
= eval2 (evaluate
);
929 r
= eval2 (evaluate
&& !null (l
));
930 if (null (l
) || null (r
))
955 l
= eval1 (evaluate
);
960 r
= eval1 (evaluate
&& null (l
));