More updates towards 2.0.
[mpdm.git] / mpdm_s.c
blob4421b120f2f105530235b69ee3eb9e592500ece9
1 /*
3 MPDM - Minimum Profit Data Manager
4 Copyright (C) 2003/2010 Angel Ortega <angel@triptico.com>
6 mpdm_s.c - String management
8 This program is free software; you can redistribute it and/or
9 modify it under the terms of the GNU General Public License
10 as published by the Free Software Foundation; either version 2
11 of the License, or (at your option) any later version.
13 This program is distributed in the hope that it will be useful,
14 but WITHOUT ANY WARRANTY; without even the implied warranty of
15 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
16 GNU General Public License for more details.
18 You should have received a copy of the GNU General Public License
19 along with this program; if not, write to the Free Software
20 Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
22 http://www.triptico.com
26 #include "config.h"
28 #include <stdio.h>
29 #include <stdlib.h>
30 #include <string.h>
31 #include <wchar.h>
32 #include <locale.h>
33 #include <wctype.h>
35 #ifdef CONFOPT_GETTEXT
36 #include <libintl.h>
37 #endif
39 #ifdef CONFOPT_WIN32
40 #include <windows.h>
41 #endif
43 #include "mpdm.h"
46 /** code **/
48 void *mpdm_poke_o(void *dst, int *dsize, int *offset, const void *org, int osize, int esize)
50 if (org != NULL && osize) {
51 /* enough room? */
52 if (*offset + osize > *dsize) {
53 /* no; enlarge */
54 *dsize += osize;
56 dst = realloc(dst, *dsize * esize);
59 memcpy((char *)dst + (*offset * esize), org, osize * esize);
60 *offset += osize;
63 return dst;
67 void *mpdm_poke(void *dst, int *dsize, const void *org, int osize, int esize)
68 /* pokes (adds) org into dst, which is a dynamic string, making it grow */
70 int offset = *dsize;
72 return mpdm_poke_o(dst, dsize, &offset, org, osize, esize);
76 wchar_t *mpdm_pokewsn(wchar_t *dst, int *dsize, const wchar_t *str, int slen)
77 /* adds a wide string to dst using mpdm_poke() with size */
79 if (str)
80 dst = mpdm_poke(dst, dsize, str, slen, sizeof(wchar_t));
82 return dst;
86 wchar_t *mpdm_pokews(wchar_t *dst, int *dsize, const wchar_t *str)
87 /* adds a wide string to dst using mpdm_poke() */
89 if (str)
90 dst = mpdm_pokewsn(dst, dsize, str, wcslen(str));
92 return dst;
96 wchar_t *mpdm_pokev(wchar_t * dst, int *dsize, const mpdm_t v)
97 /* adds the string in v to dst using mpdm_poke() */
99 if (v != NULL) {
100 const wchar_t *ptr = mpdm_string(v);
102 dst = mpdm_pokews(dst, dsize, ptr);
105 return dst;
109 wchar_t *mpdm_mbstowcs(const char *str, int *s, int l)
110 /* converts an mbs to a wcs, but filling invalid chars
111 with question marks instead of just failing */
113 wchar_t *ptr = NULL;
114 char tmp[64]; /* really MB_CUR_MAX + 1 */
115 wchar_t wc;
116 int n, i, c, t = 0;
117 char *cstr;
119 /* allow NULL values for s */
120 if (s == NULL)
121 s = &t;
123 /* if there is a limit, duplicate and break the string */
124 if (l >= 0) {
125 cstr = strdup(str);
126 cstr[l] = '\0';
128 else
129 cstr = (char *) str;
131 /* try first a direct conversion with mbstowcs */
132 if ((*s = mbstowcs(NULL, cstr, 0)) != -1) {
133 /* direct conversion is possible; do it */
134 if ((ptr = malloc((*s + 1) * sizeof(wchar_t))) != NULL) {
135 mbstowcs(ptr, cstr, *s);
136 ptr[*s] = L'\0';
139 else {
140 /* zero everything */
141 *s = n = i = 0;
143 for (;;) {
144 /* no more characters to process? */
145 if ((c = cstr[n + i]) == '\0' && i == 0)
146 break;
148 tmp[i++] = c;
149 tmp[i] = '\0';
151 /* try to convert */
152 if (mbstowcs(&wc, tmp, 1) == (size_t) -1) {
153 /* can still be an incomplete multibyte char? */
154 if (c != '\0' && i <= (int) MB_CUR_MAX)
155 continue;
156 else {
157 /* too many failing bytes; skip 1 byte */
158 wc = L'?';
159 i = 1;
163 /* skip used bytes and back again */
164 n += i;
165 i = 0;
167 /* store new char */
168 if ((ptr = mpdm_poke(ptr, s, &wc, 1, sizeof(wchar_t))) == NULL)
169 break;
172 /* null terminate and count one less */
173 if (ptr != NULL) {
174 ptr = mpdm_poke(ptr, s, L"", 1, sizeof(wchar_t));
175 (*s)--;
179 /* free the duplicate */
180 if (cstr != str)
181 free(cstr);
183 return ptr;
187 char *mpdm_wcstombs(const wchar_t * str, int *s)
188 /* converts a wcs to an mbs, but filling invalid chars
189 with question marks instead of just failing */
191 char *ptr = NULL;
192 char tmp[64]; /* really MB_CUR_MAX + 1 */
193 int l, t = 0;
195 /* allow NULL values for s */
196 if (s == NULL)
197 s = &t;
199 /* try first a direct conversion with wcstombs */
200 if ((*s = wcstombs(NULL, str, 0)) != -1) {
201 /* direct conversion is possible; do it and return */
202 if ((ptr = malloc(*s + 1)) != NULL) {
203 wcstombs(ptr, str, *s);
204 ptr[*s] = '\0';
207 return ptr;
210 /* invalid encoding? convert characters one by one */
211 *s = 0;
213 while (*str) {
214 if ((l = wctomb(tmp, *str)) <= 0) {
215 /* if char couldn't be converted,
216 write a question mark instead */
217 l = wctomb(tmp, L'?');
220 tmp[l] = '\0';
221 if ((ptr = mpdm_poke(ptr, s, tmp, l, 1)) == NULL)
222 break;
224 str++;
227 /* null terminate and count one less */
228 if (ptr != NULL) {
229 ptr = mpdm_poke(ptr, s, "", 1, 1);
230 (*s)--;
233 return ptr;
237 mpdm_t mpdm_new_wcs(int flags, const wchar_t * str, int size, int cpy)
238 /* creates a new string value from a wcs */
240 wchar_t *ptr;
242 /* a size of -1 means 'calculate it' */
243 if (size == -1 && str != NULL)
244 size = wcslen(str);
246 /* create a copy? */
247 if (cpy) {
248 /* free() on destruction */
249 flags |= MPDM_FREE;
251 /* allocs */
252 if ((ptr = malloc((size + 1) * sizeof(wchar_t))) == NULL)
253 return NULL;
255 /* if no source, reset to zeroes; otherwise, copy */
256 if (str == NULL)
257 memset(ptr, '\0', size * sizeof(wchar_t));
258 else {
259 wcsncpy(ptr, str, size);
260 ptr[size] = L'\0';
263 else
264 ptr = (wchar_t *)str;
266 /* it's a string */
267 flags |= MPDM_STRING;
269 return mpdm_new(flags, ptr, size);
273 mpdm_t mpdm_new_mbstowcs(int flags, const char *str, int l)
274 /* creates a new string value from an mbs */
276 wchar_t *ptr;
277 int size;
279 if ((ptr = mpdm_mbstowcs(str, &size, l)) == NULL)
280 return NULL;
282 /* it's a string */
283 flags |= (MPDM_STRING | MPDM_FREE);
285 return mpdm_new(flags, ptr, size);
289 mpdm_t mpdm_new_wcstombs(int flags, const wchar_t * str)
290 /* creates a new mbs value from a wbs */
292 char *ptr;
293 int size;
295 ptr = mpdm_wcstombs(str, &size);
297 flags |= MPDM_FREE;
299 /* unset the string flag; mbs,s are not 'strings' */
300 flags &= ~MPDM_STRING;
302 return mpdm_new(flags, ptr, size);
306 mpdm_t mpdm_new_i(int ival)
307 /* creates a new string value from an integer */
309 mpdm_t v;
310 char tmp[32];
312 /* creates the visual representation */
313 snprintf(tmp, sizeof(tmp), "%d", ival);
315 v = MPDM_MBS(tmp);
317 return mpdm_set_ival(v, ival);
321 mpdm_t mpdm_new_r(double rval)
322 /* creates a new string value from a real number */
324 mpdm_t v;
325 char tmp[128];
327 /* creates the visual representation */
328 snprintf(tmp, sizeof(tmp), "%lf", rval);
330 /* manually strip useless zeroes */
331 if (strchr(tmp, '.') != NULL) {
332 char *ptr;
334 for (ptr = tmp + strlen(tmp) - 1; *ptr == '0'; ptr--);
336 /* if it's over the ., strip it also */
337 if (*ptr != '.')
338 ptr++;
340 *ptr = '\0';
343 v = MPDM_MBS(tmp);
345 return mpdm_set_rval(v, rval);
349 /* interface */
352 * mpdm_string - Returns a printable representation of a value.
353 * @v: the value
355 * Returns a printable representation of a value. For strings, it's
356 * the value data itself; for any other type, a conversion to string
357 * is returned instead. This value should be used immediately, as it
358 * can be a pointer to a static buffer.
359 * [Strings]
361 wchar_t *mpdm_string(const mpdm_t v)
363 static wchar_t wtmp[32];
364 char tmp[32];
366 /* if it's NULL, return a constant */
367 if (v == NULL)
368 return L"[NULL]";
370 /* if it's a string, return it */
371 if (v->flags & MPDM_STRING)
372 return (wchar_t *) v->data;
374 /* otherwise, return a visual representation */
375 snprintf(tmp, sizeof(tmp), "%p", v);
376 mbstowcs(wtmp, tmp, sizeof(wtmp));
377 wtmp[(sizeof(wtmp) / sizeof(wchar_t)) - 1] = L'\0';
379 return wtmp;
384 * mpdm_cmp - Compares two values.
385 * @v1: the first value
386 * @v2: the second value
388 * Compares two values. If both has the MPDM_STRING flag set,
389 * a comparison using wcscoll() is returned; if both are arrays,
390 * the size is compared first and, if they have the same number
391 * elements, each one is compared; otherwise, a simple pointer
392 * comparison is done.
393 * [Strings]
395 int mpdm_cmp(const mpdm_t v1, const mpdm_t v2)
397 int r;
399 /* same values? */
400 if (v1 == v2)
401 return 0;
403 /* is any value NULL? */
404 if (v1 == NULL)
405 return -1;
406 if (v2 == NULL)
407 return 1;
409 /* different values, but same content? (unlikely) */
410 if (v1->data == v2->data)
411 return 0;
413 if (MPDM_IS_STRING(v1) && MPDM_IS_STRING(v2))
414 r = wcscoll((wchar_t *) v1->data, (wchar_t *) v2->data);
415 else
416 if (MPDM_IS_ARRAY(v1) && MPDM_IS_ARRAY(v2)) {
417 /* compare first the sizes */
418 if ((r = mpdm_size(v1) - mpdm_size(v2)) == 0) {
419 int n;
421 /* they have the same size;
422 compare each pair of elements */
423 for (n = 0; n < mpdm_size(v1); n++) {
424 if ((r = mpdm_cmp(mpdm_aget(v1, n),
425 mpdm_aget(v2, n))) != 0)
426 break;
430 else
431 /* in any other case, compare just pointers */
432 r = (int) ((char *)v1->data - (char *)v2->data);
434 return r;
439 * mpdm_cmp_s - Compares two values (string version).
440 * @v1: the first value
441 * @v2: the second value
443 * Compares two values. Compares both values using wcscoll()
444 * if the first one is a string, or returns 1 otherwise.
446 int mpdm_cmp_s(const mpdm_t v1, const wchar_t *v2)
448 int r = 1;
450 if (MPDM_IS_STRING(v1))
451 r = wcscoll((wchar_t *) v1->data, v2);
453 return r;
458 * mpdm_splice - Creates a new string value from another.
459 * @v: the original value
460 * @i: the value to be inserted
461 * @offset: offset where the substring is to be inserted
462 * @del: number of characters to delete
464 * Creates a new string value from @v, deleting @del chars at @offset
465 * and substituting them by @i. If @del is 0, no deletion is done.
466 * both @offset and @del can be negative; if this is the case, it's
467 * assumed as counting from the end of @v. If @v is NULL, @i will become
468 * the new string, and both @offset and @del will be ignored. If @v is
469 * not NULL and @i is, no insertion process is done (only deletion, if
470 * applicable).
472 * Returns a two element array, with the new string in the first
473 * element and the deleted string in the second (with a NULL value
474 * if @del is 0).
475 * [Strings]
477 mpdm_t mpdm_splice(const mpdm_t v, const mpdm_t i, int offset, int del)
479 mpdm_t w;
480 mpdm_t n = NULL;
481 mpdm_t d = NULL;
482 int os, ns, r;
483 int ins = 0;
484 wchar_t *ptr;
486 if (v != NULL) {
487 os = mpdm_size(v);
489 /* negative offsets start from the end */
490 if (offset < 0)
491 offset = os + 1 - offset;
493 /* never add further the end */
494 if (offset > os)
495 offset = os;
497 /* negative del counts as 'characters left' */
498 if (del < 0)
499 del = os + 1 - offset + del;
501 /* something to delete? */
502 if (del > 0) {
503 /* never delete further the end */
504 if (offset + del > os)
505 del = os - offset;
507 /* deleted string */
508 d = MPDM_NS(((wchar_t *) v->data) + offset, del);
510 else
511 del = 0;
513 /* something to insert? */
514 ins = mpdm_size(i);
516 /* new size and remainder */
517 ns = os + ins - del;
518 r = offset + del;
520 if ((n = MPDM_NS(NULL, ns)) == NULL)
521 return NULL;
523 ptr = (wchar_t *)n->data;
525 /* copy the beginning */
526 if (offset > 0) {
527 wcsncpy(ptr, v->data, offset);
528 ptr += offset;
531 /* copy the text to be inserted */
532 if (ins > 0) {
533 wcsncpy(ptr, i->data, ins);
534 ptr += ins;
537 /* copy the remaining */
538 os -= r;
539 if (os > 0) {
540 wcsncpy(ptr, ((wchar_t *) v->data) + r, os);
541 ptr += os;
544 /* null terminate */
545 *ptr = L'\0';
547 else
548 n = i;
550 /* creates the output array */
551 w = MPDM_A(2);
553 mpdm_aset(w, n, 0);
554 mpdm_aset(w, d, 1);
556 return w;
561 * mpdm_strcat_sn - Concatenates two strings (string with size version).
562 * @s1: the first string
563 * @s2: the second string
564 * @size: the size of the second string
566 * Returns a new string formed by the concatenation of @s1 and @s2.
567 * [Strings]
569 mpdm_t mpdm_strcat_sn(const mpdm_t s1, const wchar_t *s2, int size)
571 wchar_t *ptr = NULL;
572 int s = 0;
574 if (s1 == NULL && s2 == NULL)
575 return NULL;
577 ptr = mpdm_pokev(ptr, &s, s1);
578 ptr = mpdm_pokewsn(ptr, &s, s2, size);
580 /* if no characters were added, returns an empty string */
581 if (ptr == NULL)
582 return MPDM_LS(L"");
584 ptr = mpdm_poke(ptr, &s, L"", 1, sizeof(wchar_t));
585 return MPDM_ENS(ptr, s - 1);
590 * mpdm_strcat_s - Concatenates two strings (string version).
591 * @s1: the first string
592 * @s2: the second string
594 * Returns a new string formed by the concatenation of @s1 and @s2.
595 * [Strings]
597 mpdm_t mpdm_strcat_s(const mpdm_t s1, const wchar_t *s2)
599 return mpdm_strcat_sn(s1, s2, s2 ? wcslen(s2) : 0);
604 * mpdm_strcat - Concatenates two strings.
605 * @s1: the first string
606 * @s2: the second string
608 * Returns a new string formed by the concatenation of @s1 and @s2.
609 * [Strings]
611 mpdm_t mpdm_strcat(const mpdm_t s1, const mpdm_t s2)
613 return mpdm_strcat_s(s1, s2 ? mpdm_string(s2) : NULL);
618 * mpdm_ival - Returns a value's data as an integer.
619 * @v: the value
621 * Returns a value's data as an integer. If the value is a string,
622 * it's converted via sscanf and returned; non-string values have all
623 * an ival of 0. The converted integer is cached, so costly string
624 * conversions are only done once. Values created with the MPDM_IVAL
625 * flag set have its ival cached from the beginning.
626 * [Strings]
627 * [Value Management]
629 int mpdm_ival(mpdm_t v)
631 if (v == NULL)
632 return 0;
634 /* if there is no cached integer, calculate it */
635 if (!(v->flags & MPDM_IVAL)) {
636 int i = 0;
638 /* if it's a string, calculate it; other
639 values will have an ival of 0 */
640 if (v->flags & MPDM_STRING) {
641 char tmp[32];
642 char *fmt = "%i";
644 wcstombs(tmp, (wchar_t *) v->data, sizeof(tmp));
645 tmp[sizeof(tmp) - 1] = '\0';
647 /* workaround for mingw32: as it doesn't
648 correctly parse octal and hexadecimal
649 numbers, they are tried as special cases */
650 if (tmp[0] == '0') {
651 if (tmp[1] == 'b' || tmp[1] == 'B') {
652 /* binary number */
653 fmt = NULL;
654 char *ptr = &tmp[2];
656 while (*ptr == '0' || *ptr == '1') {
657 i <<= 1;
659 if (*ptr == '1')
660 i |= 1;
662 ptr++;
665 else
666 if (tmp[1] == 'x' || tmp[1] == 'X')
667 fmt = "%x";
668 else
669 fmt = "%o";
672 if (fmt != NULL)
673 sscanf(tmp, fmt, &i);
676 mpdm_set_ival(v, i);
679 return v->ival;
684 * mpdm_rval - Returns a value's data as a real number (double).
685 * @v: the value
687 * Returns a value's data as a real number (double float). If the value
688 * is a string, it's converted via sscanf and returned; non-string values
689 * have all an rval of 0. The converted double is cached, so costly string
690 * conversions are only done once. Values created with the MPDM_RVAL
691 * flag set have its rval cached from the beginning.
692 * [Strings]
693 * [Value Management]
695 double mpdm_rval(mpdm_t v)
697 if (v == NULL)
698 return 0;
700 /* if there is no cached double, calculate it */
701 if (!(v->flags & MPDM_RVAL)) {
702 double r = 0.0;
704 /* if it's a string, calculate it; other
705 values will have an rval of 0.0 */
706 if (v->flags & MPDM_STRING) {
707 char tmp[128];
708 char *prev_locale;
710 wcstombs(tmp, (wchar_t *) v->data, sizeof(tmp));
711 tmp[sizeof(tmp) - 1] = '\0';
713 /* if the number starts with 0, it's
714 an octal or hexadecimal number; just
715 take the integer value and cast it */
716 if (tmp[0] == '0' && tmp[1] != '.')
717 r = (double) mpdm_ival(v);
718 else {
719 /* set locale to C for non locale-dependent
720 floating point conversion */
721 prev_locale = setlocale(LC_NUMERIC, "C");
723 /* read */
724 sscanf(tmp, "%lf", &r);
726 /* set previous locale */
727 setlocale(LC_NUMERIC, prev_locale);
731 mpdm_set_rval(v, r);
734 return v->rval;
739 * mpdm_gettext - Translates a string to the current language.
740 * @str: the string
742 * Translates the @str string to the current language.
744 * This function can still be used even if there is no real gettext
745 * support() by manually filling the __I18N__ hash.
747 * If the string is found in the current table, the translation is
748 * returned; otherwise, the same @str value is returned.
749 * [Strings]
750 * [Localization]
752 mpdm_t mpdm_gettext(const mpdm_t str)
754 mpdm_t v;
755 mpdm_t i18n = NULL;
757 /* gets the cache, if any */
758 if ((i18n = mpdm_hget_s(mpdm_root(), L"__I18N__")) == NULL)
759 return str;
761 /* try first the cache */
762 if ((v = mpdm_hget(i18n, str)) == NULL) {
763 #ifdef CONFOPT_GETTEXT
764 char *s;
765 mpdm_t t;
767 /* convert to mbs */
768 t = mpdm_ref(MPDM_2MBS(str->data));
770 /* ask gettext for it */
771 s = gettext((char *) t->data);
773 /* create new value only if it's different */
774 if (s != t->data) {
775 v = MPDM_MBS(s);
777 /* store in the cache */
778 mpdm_hset(i18n, str, v);
780 else
781 v = str;
783 mpdm_unref(t);
785 #else /* CONFOPT_GETTEXT */
787 v = str;
789 #endif /* CONFOPT_GETTEXT */
792 return v;
797 * mpdm_gettext_domain - Sets domain and data directory for translations.
798 * @dom: the domain (application name)
799 * @data: directory contaning the .mo files
801 * Sets the domain (application name) and translation data for translating
802 * strings that will be returned by mpdm_gettext().@data must point to a
803 * directory containing the .mo (compiled .po) files.
805 * If there is no gettext support, returns 0, or 1 otherwise.
806 * [Strings]
807 * [Localization]
809 int mpdm_gettext_domain(const mpdm_t dom, const mpdm_t data)
811 int ret = 0;
813 #ifdef CONFOPT_GETTEXT
815 mpdm_t dm;
816 mpdm_t dt;
818 /* convert both to mbs,s */
819 dm = mpdm_ref(MPDM_2MBS(dom->data));
820 dt = mpdm_ref(MPDM_2MBS(data->data));
822 /* bind and set domain */
823 bindtextdomain((char *) dm->data, (char *) dt->data);
824 textdomain((char *) dm->data);
826 mpdm_hset_s(mpdm_root(), L"__I18N__", MPDM_H(0));
828 mpdm_unref(dt);
829 mpdm_unref(dm);
831 ret = 1;
833 #endif /* CONFOPT_GETTEXT */
835 #ifdef CONFOPT_WIN32
837 mpdm_t v;
839 if ((v = mpdm_hget_s(mpdm_root(), L"ENV")) != NULL &&
840 mpdm_hget_s(v, L"LANG") == NULL) {
841 wchar_t *wptr = L"en";
843 /* MS Windows crappy language constants... */
845 switch((GetSystemDefaultLangID() & 0x00ff)) {
846 case 0x01: wptr = L"ar"; break; /* arabic */
847 case 0x02: wptr = L"bg"; break; /* bulgarian */
848 case 0x03: wptr = L"ca"; break; /* catalan */
849 case 0x04: wptr = L"zh"; break; /* chinese */
850 case 0x05: wptr = L"cz"; break; /* czech */
851 case 0x06: wptr = L"da"; break; /* danish */
852 case 0x07: wptr = L"de"; break; /* german */
853 case 0x08: wptr = L"el"; break; /* greek */
854 case 0x09: wptr = L"en"; break; /* english */
855 case 0x0a: wptr = L"es"; break; /* spanish */
856 case 0x0b: wptr = L"fi"; break; /* finnish */
857 case 0x0c: wptr = L"fr"; break; /* french */
858 case 0x0d: wptr = L"he"; break; /* hebrew */
859 case 0x0e: wptr = L"hu"; break; /* hungarian */
860 case 0x0f: wptr = L"is"; break; /* icelandic */
861 case 0x10: wptr = L"it"; break; /* italian */
862 case 0x11: wptr = L"jp"; break; /* japanese */
863 case 0x12: wptr = L"ko"; break; /* korean */
864 case 0x13: wptr = L"nl"; break; /* dutch */
865 case 0x14: wptr = L"no"; break; /* norwegian */
866 case 0x15: wptr = L"po"; break; /* polish */
867 case 0x16: wptr = L"pt"; break; /* portuguese */
868 case 0x17: wptr = L"rm"; break; /* romansh (switzerland) */
869 case 0x18: wptr = L"ro"; break; /* romanian */
870 case 0x19: wptr = L"ru"; break; /* russian */
871 case 0x1a: wptr = L"sr"; break; /* serbian */
872 case 0x1b: wptr = L"sk"; break; /* slovak */
873 case 0x1c: wptr = L"sq"; break; /* albanian */
874 case 0x1d: wptr = L"sv"; break; /* swedish */
877 mpdm_hset_s(v, L"LANG", MPDM_S(wptr));
880 #endif /* CONFOPT_WIN32 */
882 return ret;
886 #ifdef CONFOPT_WCWIDTH
888 int wcwidth(wchar_t);
890 int mpdm_wcwidth(wchar_t c)
892 return wcwidth(c);
895 #else /* CONFOPT_WCWIDTH */
897 #include "wcwidth.c"
899 int mpdm_wcwidth(wchar_t c)
901 return mk_wcwidth(c);
904 #endif /* CONFOPT_WCWIDTH */
908 * mpdm_sprintf - Formats a sprintf()-like string.
909 * @fmt: the string format
910 * @args: an array of values
912 * Formats a string using the sprintf() format taking the values from @args.
913 * [Strings]
915 mpdm_t mpdm_sprintf(const mpdm_t fmt, const mpdm_t args)
917 const wchar_t *i = fmt->data;
918 wchar_t *o = NULL;
919 int l = 0, n = 0;
920 wchar_t c;
922 /* loop all characters */
923 while ((c = *i++) != L'\0') {
924 int m = 0;
925 wchar_t *tptr = NULL;
926 wchar_t *wptr = NULL;
928 if (c == L'%') {
929 /* format directive */
930 char t_fmt[128];
931 char tmp[1024];
932 mpdm_t v;
933 char *ptr = NULL;
935 /* transfer the % */
936 t_fmt[m++] = '%';
938 /* transform the format to mbs */
939 while (*i != L'\0' &&
940 m < (int)(sizeof(t_fmt) - MB_CUR_MAX - 1) &&
941 wcschr(L"-.0123456789", *i) != NULL)
942 m += wctomb(&t_fmt[m], *i++);
944 /* transfer the directive */
945 m += wctomb(&t_fmt[m], *i++);
947 t_fmt[m] = '\0';
949 /* by default, copies the format */
950 strcpy(tmp, t_fmt);
952 /* pick next value */
953 v = mpdm_aget(args, n++);
955 switch (t_fmt[m - 1]) {
956 case 'd':
957 case 'i':
958 case 'u':
959 case 'x':
960 case 'X':
961 case 'o':
963 /* integer value */
964 snprintf(tmp, sizeof(tmp) - 1,
965 t_fmt, mpdm_ival(v));
966 break;
968 case 'f':
970 /* float (real) value */
971 snprintf(tmp, sizeof(tmp) - 1,
972 t_fmt, mpdm_rval(v));
973 break;
975 case 's':
977 /* string value */
978 ptr = mpdm_wcstombs(mpdm_string(v), NULL);
979 snprintf(tmp, sizeof(tmp) - 1, t_fmt, ptr);
980 free(ptr);
982 break;
984 case 'c':
986 /* char */
987 m = 1;
988 wptr = &c;
989 c = mpdm_ival(v);
990 break;
992 case 'b':
994 ptr = tmp;
995 unsigned int mask;
996 int p = 0;
998 mask = 1 << ((sizeof(int) * 8) - 1);
999 while (mask) {
1000 if (mask & (unsigned int) mpdm_ival(v)) {
1001 *ptr++ = '1';
1002 p = 1;
1004 else
1005 if (p)
1006 *ptr++ = '0';
1008 mask >>= 1;
1011 if (ptr == tmp)
1012 *ptr++ = '0';
1014 *ptr = '\0';
1015 break;
1017 case '%':
1019 /* percent sign */
1020 m = 1;
1021 wptr = &c;
1022 break;
1025 /* transfer */
1026 if (wptr == NULL)
1027 wptr = tptr = mpdm_mbstowcs(tmp, &m, -1);
1029 else {
1030 /* raw character */
1031 m = 1;
1032 wptr = &c;
1035 /* transfer */
1036 o = mpdm_poke(o, &l, wptr, m, sizeof(wchar_t));
1038 /* free the temporary buffer, if any */
1039 if (tptr != NULL)
1040 free(tptr);
1043 if (o == NULL)
1044 return NULL;
1046 /* null-terminate */
1047 o = mpdm_poke(o, &l, L"", 1, sizeof(wchar_t));
1049 return MPDM_ENS(o, l - 1);
1054 * mpdm_ulc - Converts a string to uppercase or lowecase.
1055 * @s: the string
1056 * @u: convert to uppercase (1) or to lowercase (0).
1058 * Converts @s to uppercase (for @u == 1) or to lowercase (@u == 0).
1059 * [Strings]
1061 mpdm_t mpdm_ulc(const mpdm_t s, int u)
1063 mpdm_t r = NULL;
1064 wchar_t *optr;
1065 int i = mpdm_size(s);
1067 if ((optr = malloc((i + 1) * sizeof(wchar_t))) != NULL) {
1068 wchar_t *iptr = mpdm_string(s);
1069 int n;
1071 for (n = 0; n < i; n++)
1072 optr[n] = u ? towupper(iptr[n]) : towlower(iptr[n]);
1074 optr[n] = L'\0';
1075 r = MPDM_ENS(optr, i);
1078 return r;
1082 /* scanf working buffers */
1083 #define SCANF_BUF_SIZE 1024
1084 static wchar_t scanf_yset[SCANF_BUF_SIZE];
1085 static wchar_t scanf_nset[SCANF_BUF_SIZE];
1086 static wchar_t scanf_mark[SCANF_BUF_SIZE];
1088 struct {
1089 wchar_t cmd;
1090 wchar_t *yset;
1091 wchar_t *nset;
1092 } scanf_sets[] = {
1093 { L's', L"", L" \t" },
1094 { L'u', L"0123456789", L"" },
1095 { L'd', L"-0123456789", L"" },
1096 { L'i', L"-0123456789", L"" },
1097 { L'f', L"-0123456789.", L"" },
1098 { L'x', L"-0123456789xabcdefABCDEF", L"" },
1099 { L'\0', NULL, NULL },
1103 * mpdm_sscanf - Extracts data like sscanf().
1104 * @fmt: the string format
1105 * @str: the string to be parsed
1106 * @offset: the character offset to start scanning
1108 * Extracts data from a string using a special format pattern, very
1109 * much like the scanf() series of functions in the C library. Apart
1110 * from the standard percent-sign-commands (s, u, d, i, f, x,
1111 * n, [, with optional size and * to ignore), it implements S,
1112 * to match a string of characters upto what follows in the format
1113 * string. Also, the [ set of characters can include other % formats.
1115 * Returns an array with the extracted values. If %n is used, the
1116 * position in the scanned string is returned as the value.
1117 * [Strings]
1119 mpdm_t mpdm_sscanf(const mpdm_t fmt, const mpdm_t str, int offset)
1121 wchar_t *i = (wchar_t *)str->data;
1122 wchar_t *f = (wchar_t *)fmt->data;
1123 mpdm_t r;
1125 i += offset;
1126 r = MPDM_A(0);
1128 while (*f) {
1129 if (*f == L'%') {
1130 wchar_t *ptr = NULL;
1131 int size = 0;
1132 wchar_t cmd;
1133 int vsize = 0;
1134 int ignore = 0;
1135 int msize = 0;
1137 /* empty all buffers */
1138 scanf_yset[0] = scanf_nset[0] = scanf_mark[0] = L'\0';
1140 f++;
1142 /* an asterisk? don't return next value */
1143 if (*f == L'*') {
1144 ignore = 1;
1145 f++;
1148 /* does it have a size? */
1149 while (wcschr(L"0123456789", *f)) {
1150 vsize *= 10;
1151 vsize += *f - L'0';
1152 f++;
1155 /* if no size, set it to an arbitrary big limit */
1156 if (!vsize)
1157 vsize = 0xfffffff;
1159 /* now *f should contain a command */
1160 cmd = *f;
1161 f++;
1163 /* is it a verbatim percent sign? */
1164 if (cmd == L'%') {
1165 vsize = 1;
1166 ignore = 1;
1167 wcscpy(scanf_yset, L"%");
1169 else
1170 /* a position? */
1171 if (cmd == L'n') {
1172 vsize = 0;
1173 ignore = 1;
1174 mpdm_push(r, MPDM_I(i - (wchar_t *)str->data));
1176 else
1177 /* string upto a mark */
1178 if (cmd == L'S') {
1179 wchar_t *tmp = f;
1181 /* fill the mark upto another command */
1182 while (*tmp) {
1183 if (*tmp == L'%') {
1184 tmp++;
1186 /* is it an 'n'? ignore and go on */
1187 if (*tmp == L'n') {
1188 tmp++;
1189 continue;
1191 else
1192 if (*tmp == L'%')
1193 scanf_mark[msize++] = *tmp;
1194 else
1195 break;
1197 else
1198 scanf_mark[msize++] = *tmp;
1200 tmp++;
1203 scanf_mark[msize] = L'\0';
1205 else
1206 /* raw set */
1207 if (cmd == L'[') {
1208 int n = 0;
1209 wchar_t *set = scanf_yset;
1211 /* is it an inverse set? */
1212 if (*f == L'^') {
1213 set = scanf_nset;
1214 f++;
1217 /* first one is a ]? add it */
1218 if (*f == L']') {
1219 set[n++] = *f;
1220 f++;
1223 /* now build the set */
1224 for (; n < SCANF_BUF_SIZE - 1 && *f && *f != L']'; f++) {
1225 /* is it a range? */
1226 if (*f == L'-') {
1227 f++;
1229 /* start or end? hyphen itself */
1230 if (n == 0 || *f == L']')
1231 set[n++] = L'-';
1232 else {
1233 /* pick previous char */
1234 wchar_t c = set[n - 1];
1236 /* fill */
1237 while (n < SCANF_BUF_SIZE - 1 && c < *f)
1238 set[n++] = ++c;
1241 else
1242 /* is it another command? */
1243 if (*f == L'%') {
1244 int i;
1246 f++;
1247 for (i = 0; scanf_sets[i].cmd; i++) {
1248 if (*f == scanf_sets[i].cmd) {
1249 set[n] = L'\0';
1250 wcscat(set, scanf_sets[i].yset);
1251 n += wcslen(scanf_sets[i].yset);
1252 break;
1256 else
1257 set[n++] = *f;
1260 /* skip the ] */
1261 f++;
1263 set[n] = L'\0';
1265 else
1266 /* a standard set? */
1268 int n;
1270 for (n = 0; scanf_sets[n].cmd != L'\0'; n++) {
1271 if (cmd == scanf_sets[n].cmd) {
1272 wcscpy(scanf_yset, scanf_sets[n].yset);
1273 wcscpy(scanf_nset, scanf_sets[n].nset);
1274 break;
1279 /* now fill the dynamic string */
1280 while (vsize &&
1281 !wcschr(scanf_nset, *i) &&
1282 (scanf_yset[0] == L'\0' || wcschr(scanf_yset, *i)) &&
1283 (msize == 0 || wcsncmp(i, scanf_mark, msize) != 0)) {
1285 /* only add if not being ignored */
1286 if (!ignore)
1287 ptr = mpdm_poke(ptr, &size, i, 1, sizeof(wchar_t));
1289 i++;
1290 vsize--;
1293 if (!ignore && size) {
1294 /* null terminate and push */
1295 ptr = mpdm_poke(ptr, &size, L"", 1, sizeof(wchar_t));
1296 mpdm_push(r, MPDM_ENS(ptr, size));
1299 else
1300 if (*f == L' ' || *f == L'\t') {
1301 /* if it's a blank, sync to next non-blank */
1302 f++;
1304 while (*i == L' ' || *i == L'\t')
1305 i++;
1307 else
1308 /* test for literals in the format string */
1309 if (*i == *f) {
1310 i++;
1311 f++;
1313 else
1314 break;
1317 return r;