b9004264eec69d25a9447bab97ebfaa9467ea77d
[deliverable/binutils-gdb.git] / gdb / f-exp.y
1 /* YACC parser for Fortran expressions, for GDB.
2 Copyright 1986, 1989, 1990, 1991, 1993, 1994, 1995, 1996, 2000, 2001
3 Free Software Foundation, Inc.
4
5 Contributed by Motorola. Adapted from the C parser by Farooq Butt
6 (fmbutt@engage.sps.mot.com).
7
8 This file is part of GDB.
9
10 This program is free software; you can redistribute it and/or modify
11 it under the terms of the GNU General Public License as published by
12 the Free Software Foundation; either version 2 of the License, or
13 (at your option) any later version.
14
15 This program is distributed in the hope that it will be useful,
16 but WITHOUT ANY WARRANTY; without even the implied warranty of
17 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
18 GNU General Public License for more details.
19
20 You should have received a copy of the GNU General Public License
21 along with this program; if not, write to the Free Software
22 Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */
23
24 /* This was blantantly ripped off the C expression parser, please
25 be aware of that as you look at its basic structure -FMB */
26
27 /* Parse a F77 expression from text in a string,
28 and return the result as a struct expression pointer.
29 That structure contains arithmetic operations in reverse polish,
30 with constants represented by operations that are followed by special data.
31 See expression.h for the details of the format.
32 What is important here is that it can be built up sequentially
33 during the process of parsing; the lower levels of the tree always
34 come first in the result.
35
36 Note that malloc's and realloc's in this file are transformed to
37 xmalloc and xrealloc respectively by the same sed command in the
38 makefile that remaps any other malloc/realloc inserted by the parser
39 generator. Doing this with #defines and trying to control the interaction
40 with include files (<malloc.h> and <stdlib.h> for example) just became
41 too messy, particularly when such includes can be inserted at random
42 times by the parser generator. */
43
44 %{
45
46 #include "defs.h"
47 #include "gdb_string.h"
48 #include "expression.h"
49 #include "value.h"
50 #include "parser-defs.h"
51 #include "language.h"
52 #include "f-lang.h"
53 #include "bfd.h" /* Required by objfiles.h. */
54 #include "symfile.h" /* Required by objfiles.h. */
55 #include "objfiles.h" /* For have_full_symbols and have_partial_symbols */
56 #include <ctype.h>
57
58 /* Remap normal yacc parser interface names (yyparse, yylex, yyerror, etc),
59 as well as gratuitiously global symbol names, so we can have multiple
60 yacc generated parsers in gdb. Note that these are only the variables
61 produced by yacc. If other parser generators (bison, byacc, etc) produce
62 additional global names that conflict at link time, then those parser
63 generators need to be fixed instead of adding those names to this list. */
64
65 #define yymaxdepth f_maxdepth
66 #define yyparse f_parse
67 #define yylex f_lex
68 #define yyerror f_error
69 #define yylval f_lval
70 #define yychar f_char
71 #define yydebug f_debug
72 #define yypact f_pact
73 #define yyr1 f_r1
74 #define yyr2 f_r2
75 #define yydef f_def
76 #define yychk f_chk
77 #define yypgo f_pgo
78 #define yyact f_act
79 #define yyexca f_exca
80 #define yyerrflag f_errflag
81 #define yynerrs f_nerrs
82 #define yyps f_ps
83 #define yypv f_pv
84 #define yys f_s
85 #define yy_yys f_yys
86 #define yystate f_state
87 #define yytmp f_tmp
88 #define yyv f_v
89 #define yy_yyv f_yyv
90 #define yyval f_val
91 #define yylloc f_lloc
92 #define yyreds f_reds /* With YYDEBUG defined */
93 #define yytoks f_toks /* With YYDEBUG defined */
94 #define yyname f_name /* With YYDEBUG defined */
95 #define yyrule f_rule /* With YYDEBUG defined */
96 #define yylhs f_yylhs
97 #define yylen f_yylen
98 #define yydefred f_yydefred
99 #define yydgoto f_yydgoto
100 #define yysindex f_yysindex
101 #define yyrindex f_yyrindex
102 #define yygindex f_yygindex
103 #define yytable f_yytable
104 #define yycheck f_yycheck
105
106 #ifndef YYDEBUG
107 #define YYDEBUG 1 /* Default to yydebug support */
108 #endif
109
110 #define YYFPRINTF parser_fprintf
111
112 int yyparse (void);
113
114 static int yylex (void);
115
116 void yyerror (char *);
117
118 static void growbuf_by_size (int);
119
120 static int match_string_literal (void);
121
122 %}
123
124 /* Although the yacc "value" of an expression is not used,
125 since the result is stored in the structure being created,
126 other node types do have values. */
127
128 %union
129 {
130 LONGEST lval;
131 struct {
132 LONGEST val;
133 struct type *type;
134 } typed_val;
135 DOUBLEST dval;
136 struct symbol *sym;
137 struct type *tval;
138 struct stoken sval;
139 struct ttype tsym;
140 struct symtoken ssym;
141 int voidval;
142 struct block *bval;
143 enum exp_opcode opcode;
144 struct internalvar *ivar;
145
146 struct type **tvec;
147 int *ivec;
148 }
149
150 %{
151 /* YYSTYPE gets defined by %union */
152 static int parse_number (char *, int, int, YYSTYPE *);
153 %}
154
155 %type <voidval> exp type_exp start variable
156 %type <tval> type typebase
157 %type <tvec> nonempty_typelist
158 /* %type <bval> block */
159
160 /* Fancy type parsing. */
161 %type <voidval> func_mod direct_abs_decl abs_decl
162 %type <tval> ptype
163
164 %token <typed_val> INT
165 %token <dval> FLOAT
166
167 /* Both NAME and TYPENAME tokens represent symbols in the input,
168 and both convey their data as strings.
169 But a TYPENAME is a string that happens to be defined as a typedef
170 or builtin type name (such as int or char)
171 and a NAME is any other symbol.
172 Contexts where this distinction is not important can use the
173 nonterminal "name", which matches either NAME or TYPENAME. */
174
175 %token <sval> STRING_LITERAL
176 %token <lval> BOOLEAN_LITERAL
177 %token <ssym> NAME
178 %token <tsym> TYPENAME
179 %type <sval> name
180 %type <ssym> name_not_typename
181 %type <tsym> typename
182
183 /* A NAME_OR_INT is a symbol which is not known in the symbol table,
184 but which would parse as a valid number in the current input radix.
185 E.g. "c" when input_radix==16. Depending on the parse, it will be
186 turned into a name or into a number. */
187
188 %token <ssym> NAME_OR_INT
189
190 %token SIZEOF
191 %token ERROR
192
193 /* Special type cases, put in to allow the parser to distinguish different
194 legal basetypes. */
195 %token INT_KEYWORD INT_S2_KEYWORD LOGICAL_S1_KEYWORD LOGICAL_S2_KEYWORD
196 %token LOGICAL_KEYWORD REAL_KEYWORD REAL_S8_KEYWORD REAL_S16_KEYWORD
197 %token COMPLEX_S8_KEYWORD COMPLEX_S16_KEYWORD COMPLEX_S32_KEYWORD
198 %token BOOL_AND BOOL_OR BOOL_NOT
199 %token <lval> CHARACTER
200
201 %token <voidval> VARIABLE
202
203 %token <opcode> ASSIGN_MODIFY
204
205 %left ','
206 %left ABOVE_COMMA
207 %right '=' ASSIGN_MODIFY
208 %right '?'
209 %left BOOL_OR
210 %right BOOL_NOT
211 %left BOOL_AND
212 %left '|'
213 %left '^'
214 %left '&'
215 %left EQUAL NOTEQUAL
216 %left LESSTHAN GREATERTHAN LEQ GEQ
217 %left LSH RSH
218 %left '@'
219 %left '+' '-'
220 %left '*' '/' '%'
221 %right UNARY
222 %right '('
223
224 \f
225 %%
226
227 start : exp
228 | type_exp
229 ;
230
231 type_exp: type
232 { write_exp_elt_opcode(OP_TYPE);
233 write_exp_elt_type($1);
234 write_exp_elt_opcode(OP_TYPE); }
235 ;
236
237 exp : '(' exp ')'
238 { }
239 ;
240
241 /* Expressions, not including the comma operator. */
242 exp : '*' exp %prec UNARY
243 { write_exp_elt_opcode (UNOP_IND); }
244
245 exp : '&' exp %prec UNARY
246 { write_exp_elt_opcode (UNOP_ADDR); }
247
248 exp : '-' exp %prec UNARY
249 { write_exp_elt_opcode (UNOP_NEG); }
250 ;
251
252 exp : BOOL_NOT exp %prec UNARY
253 { write_exp_elt_opcode (UNOP_LOGICAL_NOT); }
254 ;
255
256 exp : '~' exp %prec UNARY
257 { write_exp_elt_opcode (UNOP_COMPLEMENT); }
258 ;
259
260 exp : SIZEOF exp %prec UNARY
261 { write_exp_elt_opcode (UNOP_SIZEOF); }
262 ;
263
264 /* No more explicit array operators, we treat everything in F77 as
265 a function call. The disambiguation as to whether we are
266 doing a subscript operation or a function call is done
267 later in eval.c. */
268
269 exp : exp '('
270 { start_arglist (); }
271 arglist ')'
272 { write_exp_elt_opcode (OP_F77_UNDETERMINED_ARGLIST);
273 write_exp_elt_longcst ((LONGEST) end_arglist ());
274 write_exp_elt_opcode (OP_F77_UNDETERMINED_ARGLIST); }
275 ;
276
277 arglist :
278 ;
279
280 arglist : exp
281 { arglist_len = 1; }
282 ;
283
284 arglist : substring
285 { arglist_len = 2;}
286
287 arglist : arglist ',' exp %prec ABOVE_COMMA
288 { arglist_len++; }
289 ;
290
291 substring: exp ':' exp %prec ABOVE_COMMA
292 { }
293 ;
294
295
296 complexnum: exp ',' exp
297 { }
298 ;
299
300 exp : '(' complexnum ')'
301 { write_exp_elt_opcode(OP_COMPLEX); }
302 ;
303
304 exp : '(' type ')' exp %prec UNARY
305 { write_exp_elt_opcode (UNOP_CAST);
306 write_exp_elt_type ($2);
307 write_exp_elt_opcode (UNOP_CAST); }
308 ;
309
310 /* Binary operators in order of decreasing precedence. */
311
312 exp : exp '@' exp
313 { write_exp_elt_opcode (BINOP_REPEAT); }
314 ;
315
316 exp : exp '*' exp
317 { write_exp_elt_opcode (BINOP_MUL); }
318 ;
319
320 exp : exp '/' exp
321 { write_exp_elt_opcode (BINOP_DIV); }
322 ;
323
324 exp : exp '%' exp
325 { write_exp_elt_opcode (BINOP_REM); }
326 ;
327
328 exp : exp '+' exp
329 { write_exp_elt_opcode (BINOP_ADD); }
330 ;
331
332 exp : exp '-' exp
333 { write_exp_elt_opcode (BINOP_SUB); }
334 ;
335
336 exp : exp LSH exp
337 { write_exp_elt_opcode (BINOP_LSH); }
338 ;
339
340 exp : exp RSH exp
341 { write_exp_elt_opcode (BINOP_RSH); }
342 ;
343
344 exp : exp EQUAL exp
345 { write_exp_elt_opcode (BINOP_EQUAL); }
346 ;
347
348 exp : exp NOTEQUAL exp
349 { write_exp_elt_opcode (BINOP_NOTEQUAL); }
350 ;
351
352 exp : exp LEQ exp
353 { write_exp_elt_opcode (BINOP_LEQ); }
354 ;
355
356 exp : exp GEQ exp
357 { write_exp_elt_opcode (BINOP_GEQ); }
358 ;
359
360 exp : exp LESSTHAN exp
361 { write_exp_elt_opcode (BINOP_LESS); }
362 ;
363
364 exp : exp GREATERTHAN exp
365 { write_exp_elt_opcode (BINOP_GTR); }
366 ;
367
368 exp : exp '&' exp
369 { write_exp_elt_opcode (BINOP_BITWISE_AND); }
370 ;
371
372 exp : exp '^' exp
373 { write_exp_elt_opcode (BINOP_BITWISE_XOR); }
374 ;
375
376 exp : exp '|' exp
377 { write_exp_elt_opcode (BINOP_BITWISE_IOR); }
378 ;
379
380 exp : exp BOOL_AND exp
381 { write_exp_elt_opcode (BINOP_LOGICAL_AND); }
382 ;
383
384
385 exp : exp BOOL_OR exp
386 { write_exp_elt_opcode (BINOP_LOGICAL_OR); }
387 ;
388
389 exp : exp '=' exp
390 { write_exp_elt_opcode (BINOP_ASSIGN); }
391 ;
392
393 exp : exp ASSIGN_MODIFY exp
394 { write_exp_elt_opcode (BINOP_ASSIGN_MODIFY);
395 write_exp_elt_opcode ($2);
396 write_exp_elt_opcode (BINOP_ASSIGN_MODIFY); }
397 ;
398
399 exp : INT
400 { write_exp_elt_opcode (OP_LONG);
401 write_exp_elt_type ($1.type);
402 write_exp_elt_longcst ((LONGEST)($1.val));
403 write_exp_elt_opcode (OP_LONG); }
404 ;
405
406 exp : NAME_OR_INT
407 { YYSTYPE val;
408 parse_number ($1.stoken.ptr, $1.stoken.length, 0, &val);
409 write_exp_elt_opcode (OP_LONG);
410 write_exp_elt_type (val.typed_val.type);
411 write_exp_elt_longcst ((LONGEST)val.typed_val.val);
412 write_exp_elt_opcode (OP_LONG); }
413 ;
414
415 exp : FLOAT
416 { write_exp_elt_opcode (OP_DOUBLE);
417 write_exp_elt_type (builtin_type_f_real_s8);
418 write_exp_elt_dblcst ($1);
419 write_exp_elt_opcode (OP_DOUBLE); }
420 ;
421
422 exp : variable
423 ;
424
425 exp : VARIABLE
426 ;
427
428 exp : SIZEOF '(' type ')' %prec UNARY
429 { write_exp_elt_opcode (OP_LONG);
430 write_exp_elt_type (builtin_type_f_integer);
431 CHECK_TYPEDEF ($3);
432 write_exp_elt_longcst ((LONGEST) TYPE_LENGTH ($3));
433 write_exp_elt_opcode (OP_LONG); }
434 ;
435
436 exp : BOOLEAN_LITERAL
437 { write_exp_elt_opcode (OP_BOOL);
438 write_exp_elt_longcst ((LONGEST) $1);
439 write_exp_elt_opcode (OP_BOOL);
440 }
441 ;
442
443 exp : STRING_LITERAL
444 {
445 write_exp_elt_opcode (OP_STRING);
446 write_exp_string ($1);
447 write_exp_elt_opcode (OP_STRING);
448 }
449 ;
450
451 variable: name_not_typename
452 { struct symbol *sym = $1.sym;
453
454 if (sym)
455 {
456 if (symbol_read_needs_frame (sym))
457 {
458 if (innermost_block == 0 ||
459 contained_in (block_found,
460 innermost_block))
461 innermost_block = block_found;
462 }
463 write_exp_elt_opcode (OP_VAR_VALUE);
464 /* We want to use the selected frame, not
465 another more inner frame which happens to
466 be in the same block. */
467 write_exp_elt_block (NULL);
468 write_exp_elt_sym (sym);
469 write_exp_elt_opcode (OP_VAR_VALUE);
470 break;
471 }
472 else
473 {
474 struct minimal_symbol *msymbol;
475 register char *arg = copy_name ($1.stoken);
476
477 msymbol =
478 lookup_minimal_symbol (arg, NULL, NULL);
479 if (msymbol != NULL)
480 {
481 write_exp_msymbol (msymbol,
482 lookup_function_type (builtin_type_int),
483 builtin_type_int);
484 }
485 else if (!have_full_symbols () && !have_partial_symbols ())
486 error ("No symbol table is loaded. Use the \"file\" command.");
487 else
488 error ("No symbol \"%s\" in current context.",
489 copy_name ($1.stoken));
490 }
491 }
492 ;
493
494
495 type : ptype
496 ;
497
498 ptype : typebase
499 | typebase abs_decl
500 {
501 /* This is where the interesting stuff happens. */
502 int done = 0;
503 int array_size;
504 struct type *follow_type = $1;
505 struct type *range_type;
506
507 while (!done)
508 switch (pop_type ())
509 {
510 case tp_end:
511 done = 1;
512 break;
513 case tp_pointer:
514 follow_type = lookup_pointer_type (follow_type);
515 break;
516 case tp_reference:
517 follow_type = lookup_reference_type (follow_type);
518 break;
519 case tp_array:
520 array_size = pop_type_int ();
521 if (array_size != -1)
522 {
523 range_type =
524 create_range_type ((struct type *) NULL,
525 builtin_type_f_integer, 0,
526 array_size - 1);
527 follow_type =
528 create_array_type ((struct type *) NULL,
529 follow_type, range_type);
530 }
531 else
532 follow_type = lookup_pointer_type (follow_type);
533 break;
534 case tp_function:
535 follow_type = lookup_function_type (follow_type);
536 break;
537 }
538 $$ = follow_type;
539 }
540 ;
541
542 abs_decl: '*'
543 { push_type (tp_pointer); $$ = 0; }
544 | '*' abs_decl
545 { push_type (tp_pointer); $$ = $2; }
546 | '&'
547 { push_type (tp_reference); $$ = 0; }
548 | '&' abs_decl
549 { push_type (tp_reference); $$ = $2; }
550 | direct_abs_decl
551 ;
552
553 direct_abs_decl: '(' abs_decl ')'
554 { $$ = $2; }
555 | direct_abs_decl func_mod
556 { push_type (tp_function); }
557 | func_mod
558 { push_type (tp_function); }
559 ;
560
561 func_mod: '(' ')'
562 { $$ = 0; }
563 | '(' nonempty_typelist ')'
564 { free ((PTR)$2); $$ = 0; }
565 ;
566
567 typebase /* Implements (approximately): (type-qualifier)* type-specifier */
568 : TYPENAME
569 { $$ = $1.type; }
570 | INT_KEYWORD
571 { $$ = builtin_type_f_integer; }
572 | INT_S2_KEYWORD
573 { $$ = builtin_type_f_integer_s2; }
574 | CHARACTER
575 { $$ = builtin_type_f_character; }
576 | LOGICAL_KEYWORD
577 { $$ = builtin_type_f_logical;}
578 | LOGICAL_S2_KEYWORD
579 { $$ = builtin_type_f_logical_s2;}
580 | LOGICAL_S1_KEYWORD
581 { $$ = builtin_type_f_logical_s1;}
582 | REAL_KEYWORD
583 { $$ = builtin_type_f_real;}
584 | REAL_S8_KEYWORD
585 { $$ = builtin_type_f_real_s8;}
586 | REAL_S16_KEYWORD
587 { $$ = builtin_type_f_real_s16;}
588 | COMPLEX_S8_KEYWORD
589 { $$ = builtin_type_f_complex_s8;}
590 | COMPLEX_S16_KEYWORD
591 { $$ = builtin_type_f_complex_s16;}
592 | COMPLEX_S32_KEYWORD
593 { $$ = builtin_type_f_complex_s32;}
594 ;
595
596 typename: TYPENAME
597 ;
598
599 nonempty_typelist
600 : type
601 { $$ = (struct type **) malloc (sizeof (struct type *) * 2);
602 $<ivec>$[0] = 1; /* Number of types in vector */
603 $$[1] = $1;
604 }
605 | nonempty_typelist ',' type
606 { int len = sizeof (struct type *) * (++($<ivec>1[0]) + 1);
607 $$ = (struct type **) realloc ((char *) $1, len);
608 $$[$<ivec>$[0]] = $3;
609 }
610 ;
611
612 name : NAME
613 { $$ = $1.stoken; }
614 | TYPENAME
615 { $$ = $1.stoken; }
616 | NAME_OR_INT
617 { $$ = $1.stoken; }
618 ;
619
620 name_not_typename : NAME
621 /* These would be useful if name_not_typename was useful, but it is just
622 a fake for "variable", so these cause reduce/reduce conflicts because
623 the parser can't tell whether NAME_OR_INT is a name_not_typename (=variable,
624 =exp) or just an exp. If name_not_typename was ever used in an lvalue
625 context where only a name could occur, this might be useful.
626 | NAME_OR_INT
627 */
628 ;
629
630 %%
631
632 /* Take care of parsing a number (anything that starts with a digit).
633 Set yylval and return the token type; update lexptr.
634 LEN is the number of characters in it. */
635
636 /*** Needs some error checking for the float case ***/
637
638 static int
639 parse_number (p, len, parsed_float, putithere)
640 register char *p;
641 register int len;
642 int parsed_float;
643 YYSTYPE *putithere;
644 {
645 register LONGEST n = 0;
646 register LONGEST prevn = 0;
647 register int c;
648 register int base = input_radix;
649 int unsigned_p = 0;
650 int long_p = 0;
651 ULONGEST high_bit;
652 struct type *signed_type;
653 struct type *unsigned_type;
654
655 if (parsed_float)
656 {
657 /* It's a float since it contains a point or an exponent. */
658 /* [dD] is not understood as an exponent by atof, change it to 'e'. */
659 char *tmp, *tmp2;
660
661 tmp = xstrdup (p);
662 for (tmp2 = tmp; *tmp2; ++tmp2)
663 if (*tmp2 == 'd' || *tmp2 == 'D')
664 *tmp2 = 'e';
665 putithere->dval = atof (tmp);
666 free (tmp);
667 return FLOAT;
668 }
669
670 /* Handle base-switching prefixes 0x, 0t, 0d, 0 */
671 if (p[0] == '0')
672 switch (p[1])
673 {
674 case 'x':
675 case 'X':
676 if (len >= 3)
677 {
678 p += 2;
679 base = 16;
680 len -= 2;
681 }
682 break;
683
684 case 't':
685 case 'T':
686 case 'd':
687 case 'D':
688 if (len >= 3)
689 {
690 p += 2;
691 base = 10;
692 len -= 2;
693 }
694 break;
695
696 default:
697 base = 8;
698 break;
699 }
700
701 while (len-- > 0)
702 {
703 c = *p++;
704 if (isupper (c))
705 c = tolower (c);
706 if (len == 0 && c == 'l')
707 long_p = 1;
708 else if (len == 0 && c == 'u')
709 unsigned_p = 1;
710 else
711 {
712 int i;
713 if (c >= '0' && c <= '9')
714 i = c - '0';
715 else if (c >= 'a' && c <= 'f')
716 i = c - 'a' + 10;
717 else
718 return ERROR; /* Char not a digit */
719 if (i >= base)
720 return ERROR; /* Invalid digit in this base */
721 n *= base;
722 n += i;
723 }
724 /* Portably test for overflow (only works for nonzero values, so make
725 a second check for zero). */
726 if ((prevn >= n) && n != 0)
727 unsigned_p=1; /* Try something unsigned */
728 /* If range checking enabled, portably test for unsigned overflow. */
729 if (RANGE_CHECK && n != 0)
730 {
731 if ((unsigned_p && (unsigned)prevn >= (unsigned)n))
732 range_error("Overflow on numeric constant.");
733 }
734 prevn = n;
735 }
736
737 /* If the number is too big to be an int, or it's got an l suffix
738 then it's a long. Work out if this has to be a long by
739 shifting right and and seeing if anything remains, and the
740 target int size is different to the target long size.
741
742 In the expression below, we could have tested
743 (n >> TARGET_INT_BIT)
744 to see if it was zero,
745 but too many compilers warn about that, when ints and longs
746 are the same size. So we shift it twice, with fewer bits
747 each time, for the same result. */
748
749 if ((TARGET_INT_BIT != TARGET_LONG_BIT
750 && ((n >> 2) >> (TARGET_INT_BIT-2))) /* Avoid shift warning */
751 || long_p)
752 {
753 high_bit = ((ULONGEST)1) << (TARGET_LONG_BIT-1);
754 unsigned_type = builtin_type_unsigned_long;
755 signed_type = builtin_type_long;
756 }
757 else
758 {
759 high_bit = ((ULONGEST)1) << (TARGET_INT_BIT-1);
760 unsigned_type = builtin_type_unsigned_int;
761 signed_type = builtin_type_int;
762 }
763
764 putithere->typed_val.val = n;
765
766 /* If the high bit of the worked out type is set then this number
767 has to be unsigned. */
768
769 if (unsigned_p || (n & high_bit))
770 putithere->typed_val.type = unsigned_type;
771 else
772 putithere->typed_val.type = signed_type;
773
774 return INT;
775 }
776
777 struct token
778 {
779 char *operator;
780 int token;
781 enum exp_opcode opcode;
782 };
783
784 static const struct token dot_ops[] =
785 {
786 { ".and.", BOOL_AND, BINOP_END },
787 { ".AND.", BOOL_AND, BINOP_END },
788 { ".or.", BOOL_OR, BINOP_END },
789 { ".OR.", BOOL_OR, BINOP_END },
790 { ".not.", BOOL_NOT, BINOP_END },
791 { ".NOT.", BOOL_NOT, BINOP_END },
792 { ".eq.", EQUAL, BINOP_END },
793 { ".EQ.", EQUAL, BINOP_END },
794 { ".eqv.", EQUAL, BINOP_END },
795 { ".NEQV.", NOTEQUAL, BINOP_END },
796 { ".neqv.", NOTEQUAL, BINOP_END },
797 { ".EQV.", EQUAL, BINOP_END },
798 { ".ne.", NOTEQUAL, BINOP_END },
799 { ".NE.", NOTEQUAL, BINOP_END },
800 { ".le.", LEQ, BINOP_END },
801 { ".LE.", LEQ, BINOP_END },
802 { ".ge.", GEQ, BINOP_END },
803 { ".GE.", GEQ, BINOP_END },
804 { ".gt.", GREATERTHAN, BINOP_END },
805 { ".GT.", GREATERTHAN, BINOP_END },
806 { ".lt.", LESSTHAN, BINOP_END },
807 { ".LT.", LESSTHAN, BINOP_END },
808 { NULL, 0, 0 }
809 };
810
811 struct f77_boolean_val
812 {
813 char *name;
814 int value;
815 };
816
817 static const struct f77_boolean_val boolean_values[] =
818 {
819 { ".true.", 1 },
820 { ".TRUE.", 1 },
821 { ".false.", 0 },
822 { ".FALSE.", 0 },
823 { NULL, 0 }
824 };
825
826 static const struct token f77_keywords[] =
827 {
828 { "complex_16", COMPLEX_S16_KEYWORD, BINOP_END },
829 { "complex_32", COMPLEX_S32_KEYWORD, BINOP_END },
830 { "character", CHARACTER, BINOP_END },
831 { "integer_2", INT_S2_KEYWORD, BINOP_END },
832 { "logical_1", LOGICAL_S1_KEYWORD, BINOP_END },
833 { "logical_2", LOGICAL_S2_KEYWORD, BINOP_END },
834 { "complex_8", COMPLEX_S8_KEYWORD, BINOP_END },
835 { "integer", INT_KEYWORD, BINOP_END },
836 { "logical", LOGICAL_KEYWORD, BINOP_END },
837 { "real_16", REAL_S16_KEYWORD, BINOP_END },
838 { "complex", COMPLEX_S8_KEYWORD, BINOP_END },
839 { "sizeof", SIZEOF, BINOP_END },
840 { "real_8", REAL_S8_KEYWORD, BINOP_END },
841 { "real", REAL_KEYWORD, BINOP_END },
842 { NULL, 0, 0 }
843 };
844
845 /* Implementation of a dynamically expandable buffer for processing input
846 characters acquired through lexptr and building a value to return in
847 yylval. Ripped off from ch-exp.y */
848
849 static char *tempbuf; /* Current buffer contents */
850 static int tempbufsize; /* Size of allocated buffer */
851 static int tempbufindex; /* Current index into buffer */
852
853 #define GROWBY_MIN_SIZE 64 /* Minimum amount to grow buffer by */
854
855 #define CHECKBUF(size) \
856 do { \
857 if (tempbufindex + (size) >= tempbufsize) \
858 { \
859 growbuf_by_size (size); \
860 } \
861 } while (0);
862
863
864 /* Grow the static temp buffer if necessary, including allocating the first one
865 on demand. */
866
867 static void
868 growbuf_by_size (count)
869 int count;
870 {
871 int growby;
872
873 growby = max (count, GROWBY_MIN_SIZE);
874 tempbufsize += growby;
875 if (tempbuf == NULL)
876 tempbuf = (char *) malloc (tempbufsize);
877 else
878 tempbuf = (char *) realloc (tempbuf, tempbufsize);
879 }
880
881 /* Blatantly ripped off from ch-exp.y. This routine recognizes F77
882 string-literals.
883
884 Recognize a string literal. A string literal is a nonzero sequence
885 of characters enclosed in matching single quotes, except that
886 a single character inside single quotes is a character literal, which
887 we reject as a string literal. To embed the terminator character inside
888 a string, it is simply doubled (I.E. 'this''is''one''string') */
889
890 static int
891 match_string_literal ()
892 {
893 char *tokptr = lexptr;
894
895 for (tempbufindex = 0, tokptr++; *tokptr != '\0'; tokptr++)
896 {
897 CHECKBUF (1);
898 if (*tokptr == *lexptr)
899 {
900 if (*(tokptr + 1) == *lexptr)
901 tokptr++;
902 else
903 break;
904 }
905 tempbuf[tempbufindex++] = *tokptr;
906 }
907 if (*tokptr == '\0' /* no terminator */
908 || tempbufindex == 0) /* no string */
909 return 0;
910 else
911 {
912 tempbuf[tempbufindex] = '\0';
913 yylval.sval.ptr = tempbuf;
914 yylval.sval.length = tempbufindex;
915 lexptr = ++tokptr;
916 return STRING_LITERAL;
917 }
918 }
919
920 /* Read one token, getting characters through lexptr. */
921
922 static int
923 yylex ()
924 {
925 int c;
926 int namelen;
927 unsigned int i,token;
928 char *tokstart;
929
930 retry:
931
932 prev_lexptr = lexptr;
933
934 tokstart = lexptr;
935
936 /* First of all, let us make sure we are not dealing with the
937 special tokens .true. and .false. which evaluate to 1 and 0. */
938
939 if (*lexptr == '.')
940 {
941 for (i = 0; boolean_values[i].name != NULL; i++)
942 {
943 if STREQN (tokstart, boolean_values[i].name,
944 strlen (boolean_values[i].name))
945 {
946 lexptr += strlen (boolean_values[i].name);
947 yylval.lval = boolean_values[i].value;
948 return BOOLEAN_LITERAL;
949 }
950 }
951 }
952
953 /* See if it is a special .foo. operator */
954
955 for (i = 0; dot_ops[i].operator != NULL; i++)
956 if (STREQN (tokstart, dot_ops[i].operator, strlen (dot_ops[i].operator)))
957 {
958 lexptr += strlen (dot_ops[i].operator);
959 yylval.opcode = dot_ops[i].opcode;
960 return dot_ops[i].token;
961 }
962
963 switch (c = *tokstart)
964 {
965 case 0:
966 return 0;
967
968 case ' ':
969 case '\t':
970 case '\n':
971 lexptr++;
972 goto retry;
973
974 case '\'':
975 token = match_string_literal ();
976 if (token != 0)
977 return (token);
978 break;
979
980 case '(':
981 paren_depth++;
982 lexptr++;
983 return c;
984
985 case ')':
986 if (paren_depth == 0)
987 return 0;
988 paren_depth--;
989 lexptr++;
990 return c;
991
992 case ',':
993 if (comma_terminates && paren_depth == 0)
994 return 0;
995 lexptr++;
996 return c;
997
998 case '.':
999 /* Might be a floating point number. */
1000 if (lexptr[1] < '0' || lexptr[1] > '9')
1001 goto symbol; /* Nope, must be a symbol. */
1002 /* FALL THRU into number case. */
1003
1004 case '0':
1005 case '1':
1006 case '2':
1007 case '3':
1008 case '4':
1009 case '5':
1010 case '6':
1011 case '7':
1012 case '8':
1013 case '9':
1014 {
1015 /* It's a number. */
1016 int got_dot = 0, got_e = 0, got_d = 0, toktype;
1017 register char *p = tokstart;
1018 int hex = input_radix > 10;
1019
1020 if (c == '0' && (p[1] == 'x' || p[1] == 'X'))
1021 {
1022 p += 2;
1023 hex = 1;
1024 }
1025 else if (c == '0' && (p[1]=='t' || p[1]=='T' || p[1]=='d' || p[1]=='D'))
1026 {
1027 p += 2;
1028 hex = 0;
1029 }
1030
1031 for (;; ++p)
1032 {
1033 if (!hex && !got_e && (*p == 'e' || *p == 'E'))
1034 got_dot = got_e = 1;
1035 else if (!hex && !got_d && (*p == 'd' || *p == 'D'))
1036 got_dot = got_d = 1;
1037 else if (!hex && !got_dot && *p == '.')
1038 got_dot = 1;
1039 else if (((got_e && (p[-1] == 'e' || p[-1] == 'E'))
1040 || (got_d && (p[-1] == 'd' || p[-1] == 'D')))
1041 && (*p == '-' || *p == '+'))
1042 /* This is the sign of the exponent, not the end of the
1043 number. */
1044 continue;
1045 /* We will take any letters or digits. parse_number will
1046 complain if past the radix, or if L or U are not final. */
1047 else if ((*p < '0' || *p > '9')
1048 && ((*p < 'a' || *p > 'z')
1049 && (*p < 'A' || *p > 'Z')))
1050 break;
1051 }
1052 toktype = parse_number (tokstart, p - tokstart, got_dot|got_e|got_d,
1053 &yylval);
1054 if (toktype == ERROR)
1055 {
1056 char *err_copy = (char *) alloca (p - tokstart + 1);
1057
1058 memcpy (err_copy, tokstart, p - tokstart);
1059 err_copy[p - tokstart] = 0;
1060 error ("Invalid number \"%s\".", err_copy);
1061 }
1062 lexptr = p;
1063 return toktype;
1064 }
1065
1066 case '+':
1067 case '-':
1068 case '*':
1069 case '/':
1070 case '%':
1071 case '|':
1072 case '&':
1073 case '^':
1074 case '~':
1075 case '!':
1076 case '@':
1077 case '<':
1078 case '>':
1079 case '[':
1080 case ']':
1081 case '?':
1082 case ':':
1083 case '=':
1084 case '{':
1085 case '}':
1086 symbol:
1087 lexptr++;
1088 return c;
1089 }
1090
1091 if (!(c == '_' || c == '$'
1092 || (c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z')))
1093 /* We must have come across a bad character (e.g. ';'). */
1094 error ("Invalid character '%c' in expression.", c);
1095
1096 namelen = 0;
1097 for (c = tokstart[namelen];
1098 (c == '_' || c == '$' || (c >= '0' && c <= '9')
1099 || (c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z'));
1100 c = tokstart[++namelen]);
1101
1102 /* The token "if" terminates the expression and is NOT
1103 removed from the input stream. */
1104
1105 if (namelen == 2 && tokstart[0] == 'i' && tokstart[1] == 'f')
1106 return 0;
1107
1108 lexptr += namelen;
1109
1110 /* Catch specific keywords. */
1111
1112 for (i = 0; f77_keywords[i].operator != NULL; i++)
1113 if (STREQN(tokstart, f77_keywords[i].operator,
1114 strlen(f77_keywords[i].operator)))
1115 {
1116 /* lexptr += strlen(f77_keywords[i].operator); */
1117 yylval.opcode = f77_keywords[i].opcode;
1118 return f77_keywords[i].token;
1119 }
1120
1121 yylval.sval.ptr = tokstart;
1122 yylval.sval.length = namelen;
1123
1124 if (*tokstart == '$')
1125 {
1126 write_dollar_variable (yylval.sval);
1127 return VARIABLE;
1128 }
1129
1130 /* Use token-type TYPENAME for symbols that happen to be defined
1131 currently as names of types; NAME for other symbols.
1132 The caller is not constrained to care about the distinction. */
1133 {
1134 char *tmp = copy_name (yylval.sval);
1135 struct symbol *sym;
1136 int is_a_field_of_this = 0;
1137 int hextype;
1138
1139 sym = lookup_symbol (tmp, expression_context_block,
1140 VAR_NAMESPACE,
1141 current_language->la_language == language_cplus
1142 ? &is_a_field_of_this : NULL,
1143 NULL);
1144 if (sym && SYMBOL_CLASS (sym) == LOC_TYPEDEF)
1145 {
1146 yylval.tsym.type = SYMBOL_TYPE (sym);
1147 return TYPENAME;
1148 }
1149 if ((yylval.tsym.type = lookup_primitive_typename (tmp)) != 0)
1150 return TYPENAME;
1151
1152 /* Input names that aren't symbols but ARE valid hex numbers,
1153 when the input radix permits them, can be names or numbers
1154 depending on the parse. Note we support radixes > 16 here. */
1155 if (!sym
1156 && ((tokstart[0] >= 'a' && tokstart[0] < 'a' + input_radix - 10)
1157 || (tokstart[0] >= 'A' && tokstart[0] < 'A' + input_radix - 10)))
1158 {
1159 YYSTYPE newlval; /* Its value is ignored. */
1160 hextype = parse_number (tokstart, namelen, 0, &newlval);
1161 if (hextype == INT)
1162 {
1163 yylval.ssym.sym = sym;
1164 yylval.ssym.is_a_field_of_this = is_a_field_of_this;
1165 return NAME_OR_INT;
1166 }
1167 }
1168
1169 /* Any other kind of symbol */
1170 yylval.ssym.sym = sym;
1171 yylval.ssym.is_a_field_of_this = is_a_field_of_this;
1172 return NAME;
1173 }
1174 }
1175
1176 void
1177 yyerror (msg)
1178 char *msg;
1179 {
1180 if (prev_lexptr)
1181 lexptr = prev_lexptr;
1182
1183 error ("A %s in expression, near `%s'.", (msg ? msg : "error"), lexptr);
1184 }
This page took 0.054889 seconds and 3 git commands to generate.