4 * The Regents of the University of California. All rights reserved.
6 * This code is derived from software contributed to Berkeley by
7 * Paul Borman at Krystal Technologies.
9 * Redistribution and use in source and binary forms, with or without
10 * modification, are permitted provided that the following conditions
12 * 1. Redistributions of source code must retain the above copyright
13 * notice, this list of conditions and the following disclaimer.
14 * 2. Redistributions in binary form must reproduce the above copyright
15 * notice, this list of conditions and the following disclaimer in the
16 * documentation and/or other materials provided with the distribution.
17 * 3. All advertising materials mentioning features or use of this software
18 * must display the following acknowledgement:
19 * This product includes software developed by the University of
20 * California, Berkeley and its contributors.
21 * 4. Neither the name of the University nor the names of its contributors
22 * may be used to endorse or promote products derived from this software
23 * without specific prior written permission.
25 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
26 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
27 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
28 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
29 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
30 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
31 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
32 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
33 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
34 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
37 * @(#)yacc.y 8.1 (Berkeley) 6/6/93
38 * $FreeBSD: src/usr.bin/mklocale/yacc.y,v 1.6.2.3 2003/06/03 21:15:48 ache Exp $
39 * $DragonFly: src/usr.bin/mklocale/yacc.y,v 1.2 2003/06/17 04:29:29 dillon Exp $
42 #include <arpa/inet.h>
56 static void *xmalloc(unsigned int sz);
57 static unsigned long *xlalloc(unsigned int sz);
58 void yyerror(const char *s);
59 static unsigned long *xrelalloc(unsigned long *old, unsigned int sz);
60 static void dump_tables(void);
61 static void cleanout(void);
63 const char *locale_file = "<stdout>";
65 rune_map maplower = { { 0 }, NULL };
66 rune_map mapupper = { { 0 }, NULL };
67 rune_map types = { { 0 }, NULL };
69 _RuneLocale new_locale = { "", "", NULL, NULL, 0, {}, {}, {},
70 {0, NULL}, {0, NULL}, {0, NULL}, NULL, 0 };
72 void set_map(rune_map *, rune_list *, unsigned long);
73 void set_digitmap(rune_map *, rune_list *);
74 void add_map(rune_map *, rune_list *, unsigned long);
113 entry : ENCODING STRING
114 { if (strcmp($2, "NONE") &&
115 strcmp($2, "UTF2") &&
116 strcmp($2, "UTF-8") &&
118 strcmp($2, "BIG5") &&
119 strcmp($2, "MSKanji"))
120 warnx("ENCODING %s is not supported by libc", $2);
121 strncpy(new_locale.encoding, $2, sizeof(new_locale.encoding)); }
123 { new_locale.variable_len = strlen($1) + 1;
124 new_locale.variable = malloc(new_locale.variable_len);
125 strcpy((char *)new_locale.variable, $1);
128 { warnx("the INVALID keyword is deprecated");
129 new_locale.invalid_rune = $2;
132 { set_map(&types, $2, $1); }
134 { set_map(&maplower, $2, 0); }
136 { set_map(&mapupper, $2, 0); }
138 { set_digitmap(&types, $2); }
143 $$ = (rune_list *)malloc(sizeof(rune_list));
150 $$ = (rune_list *)malloc(sizeof(rune_list));
157 $$ = (rune_list *)malloc(sizeof(rune_list));
162 | list RUNE THRU RUNE
164 $$ = (rune_list *)malloc(sizeof(rune_list));
171 map : LBRK RUNE RUNE RBRK
173 $$ = (rune_list *)malloc(sizeof(rune_list));
179 | map LBRK RUNE RUNE RBRK
181 $$ = (rune_list *)malloc(sizeof(rune_list));
187 | LBRK RUNE THRU RUNE ':' RUNE RBRK
189 $$ = (rune_list *)malloc(sizeof(rune_list));
195 | map LBRK RUNE THRU RUNE ':' RUNE RBRK
197 $$ = (rune_list *)malloc(sizeof(rune_list));
217 main(int ac, char *av[])
225 while ((x = getopt(ac, av, "do:")) != EOF) {
231 locale_file = optarg;
232 if ((fp = fopen(locale_file, "w")) == 0) {
240 fprintf(stderr, "usage: mklocale [-d] [-o output] [source]\n");
245 switch (ac - optind) {
249 if (freopen(av[optind], "r", stdin) == 0) {
257 for (x = 0; x < _CACHED_RUNES; ++x) {
261 new_locale.invalid_rune = _INVALID_RUNE;
262 memcpy(new_locale.magic, _RUNE_MAGIC_1, sizeof(new_locale.magic));
273 fprintf(stderr, "%s\n", s);
280 void *r = malloc(sz);
288 static unsigned long *
292 unsigned long *r = (unsigned long *)malloc(sz * sizeof(unsigned long));
300 static unsigned long *
305 unsigned long *r = (unsigned long *)realloc((char *)old,
306 sz * sizeof(unsigned long));
315 set_map(map, list, flag)
321 rune_list *nlist = list->next;
322 add_map(map, list, flag);
328 set_digitmap(map, list)
335 rune_list *nlist = list->next;
336 for (i = list->min; i <= list->max; ++i) {
337 if (list->map + (i - list->min)) {
338 rune_list *tmp = (rune_list *)xmalloc(sizeof(rune_list));
341 add_map(map, tmp, list->map + (i - list->min));
350 add_map(map, list, flag)
360 while (list->min < _CACHED_RUNES && list->min <= list->max) {
362 map->map[list->min++] |= flag;
364 map->map[list->min++] = list->map++;
367 if (list->min > list->max) {
372 run = list->max - list->min + 1;
374 if (!(r = map->root) || (list->max < r->min - 1)
375 || (!flag && list->max == r->min - 1)) {
377 list->types = xlalloc(run);
378 for (i = 0; i < run; ++i)
379 list->types[i] = flag;
381 list->next = map->root;
386 for (r = map->root; r && r->max + 1 < list->min; r = r->next)
391 * We are off the end.
394 list->types = xlalloc(run);
395 for (i = 0; i < run; ++i)
396 list->types[i] = flag;
403 if (list->max < r->min - 1) {
405 * We come before this range and we do not intersect it.
406 * We are not before the root node, it was checked before the loop
409 list->types = xlalloc(run);
410 for (i = 0; i < run; ++i)
411 list->types[i] = flag;
413 list->next = lr->next;
419 * At this point we have found that we at least intersect with
420 * the range pointed to by `r', we might intersect with one or
421 * more ranges beyond `r' as well.
424 if (!flag && list->map - list->min != r->map - r->min) {
426 * There are only two cases when we are doing case maps and
427 * our maps needn't have the same offset. When we are adjoining
428 * but not intersecting.
430 if (list->max + 1 == r->min) {
435 if (list->min - 1 == r->max) {
436 list->next = r->next;
440 fprintf(stderr, "Error: conflicting map entries\n");
444 if (list->min >= r->min && list->max <= r->max) {
450 for (i = list->min; i <= list->max; ++i)
451 r->types[i - r->min] |= flag;
456 if (list->min <= r->min && list->max >= r->max) {
458 * Superset case. Make him big enough to hold us.
459 * We might need to merge with the guy after him.
462 list->types = xlalloc(list->max - list->min + 1);
464 for (i = list->min; i <= list->max; ++i)
465 list->types[i - list->min] = flag;
467 for (i = r->min; i <= r->max; ++i)
468 list->types[i - list->min] |= r->types[i - r->min];
471 r->types = list->types;
478 } else if (list->min < r->min) {
480 * Our tail intersects his head.
483 list->types = xlalloc(r->max - list->min + 1);
485 for (i = r->min; i <= r->max; ++i)
486 list->types[i - list->min] = r->types[i - r->min];
488 for (i = list->min; i < r->min; ++i)
489 list->types[i - list->min] = flag;
491 for (i = r->min; i <= list->max; ++i)
492 list->types[i - list->min] |= flag;
495 r->types = list->types;
504 * Our head intersects his tail.
505 * We might need to merge with the guy after him.
508 r->types = xrelalloc(r->types, list->max - r->min + 1);
510 for (i = list->min; i <= r->max; ++i)
511 r->types[i - r->min] |= flag;
513 for (i = r->max+1; i <= list->max; ++i)
514 r->types[i - r->min] = flag;
521 * Okay, check to see if we grew into the next guy(s)
523 while ((lr = r->next) && r->max >= lr->min) {
525 if (r->max >= lr->max) {
527 * Good, we consumed all of him.
529 for (i = lr->min; i <= lr->max; ++i)
530 r->types[i - r->min] |= lr->types[i - lr->min];
533 * "append" him on to the end of us.
535 r->types = xrelalloc(r->types, lr->max - r->min + 1);
537 for (i = lr->min; i <= r->max; ++i)
538 r->types[i - r->min] |= lr->types[i - lr->min];
540 for (i = r->max+1; i <= lr->max; ++i)
541 r->types[i - r->min] = lr->types[i - lr->min];
546 if (lr->max > r->max)
561 int x, first_d, curr_d;
565 * See if we can compress some of the istype arrays
567 for(list = types.root; list; list = list->next) {
568 list->map = list->types[0];
569 for (x = 1; x < list->max - list->min + 1; ++x) {
570 if ((rune_t)list->types[x] != list->map) {
578 for (x = 0; x < _CACHED_RUNES; ++x) {
579 unsigned long r = types.map[x];
583 first_d = curr_d = x;
584 else if (x != curr_d + 1) {
585 fprintf(stderr, "Error: DIGIT range is not contiguous\n");
587 } else if (x - first_d > 9) {
588 fprintf(stderr, "Error: DIGIT range is too big\n");
592 if (!(r & _CTYPE_X)) {
593 fprintf(stderr, "Error: DIGIT range is not a subset of XDIGIT range\n");
599 fprintf(stderr, "Error: no DIGIT range defined in the single byte area\n");
601 } else if (curr_d - first_d < 9) {
602 fprintf(stderr, "Error: DIGIT range is too small in the single byte area\n");
606 new_locale.invalid_rune = htonl(new_locale.invalid_rune);
609 * Fill in our tables. Do this in network order so that
610 * diverse machines have a chance of sharing data.
611 * (Machines like Crays cannot share with little machines due to
612 * word size. Sigh. We tried.)
614 for (x = 0; x < _CACHED_RUNES; ++x) {
615 new_locale.runetype[x] = htonl(types.map[x]);
616 new_locale.maplower[x] = htonl(maplower.map[x]);
617 new_locale.mapupper[x] = htonl(mapupper.map[x]);
621 * Count up how many ranges we will need for each of the extents.
626 new_locale.runetype_ext.nranges++;
629 new_locale.runetype_ext.nranges = htonl(new_locale.runetype_ext.nranges);
631 list = maplower.root;
634 new_locale.maplower_ext.nranges++;
637 new_locale.maplower_ext.nranges = htonl(new_locale.maplower_ext.nranges);
639 list = mapupper.root;
642 new_locale.mapupper_ext.nranges++;
645 new_locale.mapupper_ext.nranges = htonl(new_locale.mapupper_ext.nranges);
647 new_locale.variable_len = htonl(new_locale.variable_len);
650 * Okay, we are now ready to write the new locale file.
654 * PART 1: The _RuneLocale structure
656 if (fwrite((char *)&new_locale, sizeof(new_locale), 1, fp) != 1) {
661 * PART 2: The runetype_ext structures (not the actual tables)
668 re.min = htonl(list->min);
669 re.max = htonl(list->max);
670 re.map = htonl(list->map);
672 if (fwrite((char *)&re, sizeof(re), 1, fp) != 1) {
680 * PART 3: The maplower_ext structures
682 list = maplower.root;
687 re.min = htonl(list->min);
688 re.max = htonl(list->max);
689 re.map = htonl(list->map);
691 if (fwrite((char *)&re, sizeof(re), 1, fp) != 1) {
699 * PART 4: The mapupper_ext structures
701 list = mapupper.root;
706 re.min = htonl(list->min);
707 re.max = htonl(list->max);
708 re.map = htonl(list->map);
710 if (fwrite((char *)&re, sizeof(re), 1, fp) != 1) {
718 * PART 5: The runetype_ext tables
723 for (x = 0; x < list->max - list->min + 1; ++x)
724 list->types[x] = htonl(list->types[x]);
727 if (fwrite((char *)list->types,
728 (list->max - list->min + 1) * sizeof(unsigned long),
737 * PART 5: And finally the variable data
739 if (fwrite((char *)new_locale.variable,
740 ntohl(new_locale.variable_len), 1, fp) != 1) {
744 if (fclose(fp) != 0) {
753 if (new_locale.encoding[0])
754 fprintf(stderr, "ENCODING %s\n", new_locale.encoding);
755 if (new_locale.variable)
756 fprintf(stderr, "VARIABLE %s\n", (char *)new_locale.variable);
758 fprintf(stderr, "\nMAPLOWER:\n\n");
760 for (x = 0; x < _CACHED_RUNES; ++x) {
761 if (isprint(maplower.map[x]))
762 fprintf(stderr, " '%c'", (int)maplower.map[x]);
763 else if (maplower.map[x])
764 fprintf(stderr, "%04lx", maplower.map[x]);
766 fprintf(stderr, "%4x", 0);
767 if ((x & 0xf) == 0xf)
768 fprintf(stderr, "\n");
770 fprintf(stderr, " ");
772 fprintf(stderr, "\n");
774 for (list = maplower.root; list; list = list->next)
775 fprintf(stderr, "\t%04x - %04x : %04x\n", list->min, list->max, list->map);
777 fprintf(stderr, "\nMAPUPPER:\n\n");
779 for (x = 0; x < _CACHED_RUNES; ++x) {
780 if (isprint(mapupper.map[x]))
781 fprintf(stderr, " '%c'", (int)mapupper.map[x]);
782 else if (mapupper.map[x])
783 fprintf(stderr, "%04lx", mapupper.map[x]);
785 fprintf(stderr, "%4x", 0);
786 if ((x & 0xf) == 0xf)
787 fprintf(stderr, "\n");
789 fprintf(stderr, " ");
791 fprintf(stderr, "\n");
793 for (list = mapupper.root; list; list = list->next)
794 fprintf(stderr, "\t%04x - %04x : %04x\n", list->min, list->max, list->map);
797 fprintf(stderr, "\nTYPES:\n\n");
799 for (x = 0; x < _CACHED_RUNES; ++x) {
800 unsigned long r = types.map[x];
804 fprintf(stderr, " '%c': %2d", x, (int)(r & 0xff));
806 fprintf(stderr, "%04x: %2d", x, (int)(r & 0xff));
808 fprintf(stderr, " %4s", (r & _CTYPE_A) ? "alph" : "");
809 fprintf(stderr, " %4s", (r & _CTYPE_C) ? "ctrl" : "");
810 fprintf(stderr, " %4s", (r & _CTYPE_D) ? "dig" : "");
811 fprintf(stderr, " %4s", (r & _CTYPE_G) ? "graf" : "");
812 fprintf(stderr, " %4s", (r & _CTYPE_L) ? "low" : "");
813 fprintf(stderr, " %4s", (r & _CTYPE_P) ? "punc" : "");
814 fprintf(stderr, " %4s", (r & _CTYPE_S) ? "spac" : "");
815 fprintf(stderr, " %4s", (r & _CTYPE_U) ? "upp" : "");
816 fprintf(stderr, " %4s", (r & _CTYPE_X) ? "xdig" : "");
817 fprintf(stderr, " %4s", (r & _CTYPE_B) ? "blnk" : "");
818 fprintf(stderr, " %4s", (r & _CTYPE_R) ? "prnt" : "");
819 fprintf(stderr, " %4s", (r & _CTYPE_I) ? "ideo" : "");
820 fprintf(stderr, " %4s", (r & _CTYPE_T) ? "spec" : "");
821 fprintf(stderr, " %4s", (r & _CTYPE_Q) ? "phon" : "");
822 fprintf(stderr, "\n");
826 for (list = types.root; list; list = list->next) {
827 if (list->map && list->min + 3 < list->max) {
828 unsigned long r = list->map;
830 fprintf(stderr, "%04lx: %2d",
831 (unsigned long)list->min, (int)(r & 0xff));
833 fprintf(stderr, " %4s", (r & _CTYPE_A) ? "alph" : "");
834 fprintf(stderr, " %4s", (r & _CTYPE_C) ? "ctrl" : "");
835 fprintf(stderr, " %4s", (r & _CTYPE_D) ? "dig" : "");
836 fprintf(stderr, " %4s", (r & _CTYPE_G) ? "graf" : "");
837 fprintf(stderr, " %4s", (r & _CTYPE_L) ? "low" : "");
838 fprintf(stderr, " %4s", (r & _CTYPE_P) ? "punc" : "");
839 fprintf(stderr, " %4s", (r & _CTYPE_S) ? "spac" : "");
840 fprintf(stderr, " %4s", (r & _CTYPE_U) ? "upp" : "");
841 fprintf(stderr, " %4s", (r & _CTYPE_X) ? "xdig" : "");
842 fprintf(stderr, " %4s", (r & _CTYPE_B) ? "blnk" : "");
843 fprintf(stderr, " %4s", (r & _CTYPE_R) ? "prnt" : "");
844 fprintf(stderr, " %4s", (r & _CTYPE_I) ? "ideo" : "");
845 fprintf(stderr, " %4s", (r & _CTYPE_T) ? "spec" : "");
846 fprintf(stderr, " %4s", (r & _CTYPE_Q) ? "phon" : "");
847 fprintf(stderr, "\n...\n");
849 fprintf(stderr, "%04lx: %2d",
850 (unsigned long)list->max, (int)(r & 0xff));
852 fprintf(stderr, " %4s", (r & _CTYPE_A) ? "alph" : "");
853 fprintf(stderr, " %4s", (r & _CTYPE_C) ? "ctrl" : "");
854 fprintf(stderr, " %4s", (r & _CTYPE_D) ? "dig" : "");
855 fprintf(stderr, " %4s", (r & _CTYPE_G) ? "graf" : "");
856 fprintf(stderr, " %4s", (r & _CTYPE_L) ? "low" : "");
857 fprintf(stderr, " %4s", (r & _CTYPE_P) ? "punc" : "");
858 fprintf(stderr, " %4s", (r & _CTYPE_S) ? "spac" : "");
859 fprintf(stderr, " %4s", (r & _CTYPE_U) ? "upp" : "");
860 fprintf(stderr, " %4s", (r & _CTYPE_X) ? "xdig" : "");
861 fprintf(stderr, " %4s", (r & _CTYPE_B) ? "blnk" : "");
862 fprintf(stderr, " %4s", (r & _CTYPE_R) ? "prnt" : "");
863 fprintf(stderr, " %4s", (r & _CTYPE_I) ? "ideo" : "");
864 fprintf(stderr, " %4s", (r & _CTYPE_T) ? "spec" : "");
865 fprintf(stderr, " %4s", (r & _CTYPE_Q) ? "phon" : "");
866 fprintf(stderr, "\n");
868 for (x = list->min; x <= list->max; ++x) {
869 unsigned long r = ntohl(list->types[x - list->min]);
872 fprintf(stderr, "%04x: %2d", x, (int)(r & 0xff));
874 fprintf(stderr, " %4s", (r & _CTYPE_A) ? "alph" : "");
875 fprintf(stderr, " %4s", (r & _CTYPE_C) ? "ctrl" : "");
876 fprintf(stderr, " %4s", (r & _CTYPE_D) ? "dig" : "");
877 fprintf(stderr, " %4s", (r & _CTYPE_G) ? "graf" : "");
878 fprintf(stderr, " %4s", (r & _CTYPE_L) ? "low" : "");
879 fprintf(stderr, " %4s", (r & _CTYPE_P) ? "punc" : "");
880 fprintf(stderr, " %4s", (r & _CTYPE_S) ? "spac" : "");
881 fprintf(stderr, " %4s", (r & _CTYPE_U) ? "upp" : "");
882 fprintf(stderr, " %4s", (r & _CTYPE_X) ? "xdig" : "");
883 fprintf(stderr, " %4s", (r & _CTYPE_B) ? "blnk" : "");
884 fprintf(stderr, " %4s", (r & _CTYPE_R) ? "prnt" : "");
885 fprintf(stderr, " %4s", (r & _CTYPE_I) ? "ideo" : "");
886 fprintf(stderr, " %4s", (r & _CTYPE_T) ? "spec" : "");
887 fprintf(stderr, " %4s", (r & _CTYPE_Q) ? "phon" : "");
888 fprintf(stderr, "\n");