2 * Copyright (C) 2009 Gabor Kovesdan <gabor@FreeBSD.org>
3 * Copyright (C) 2012 Oleg Moskalenko <mom040267@gmail.com>
6 * Redistribution and use in source and binary forms, with or without
7 * modification, are permitted provided that the following conditions
9 * 1. Redistributions of source code must retain the above copyright
10 * notice, this list of conditions and the following disclaimer.
11 * 2. Redistributions in binary form must reproduce the above copyright
12 * notice, this list of conditions and the following disclaimer in the
13 * documentation and/or other materials provided with the distribution.
15 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
16 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
17 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
18 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
19 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
20 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
21 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
22 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
23 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
24 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
28 #include <sys/cdefs.h>
29 __FBSDID("$FreeBSD$");
46 static wchar_t **wmonths;
47 static unsigned char **cmonths;
49 /* initialise months */
52 initialise_months(void)
54 const nl_item item[12] = { ABMON_1, ABMON_2, ABMON_3, ABMON_4,
55 ABMON_5, ABMON_6, ABMON_7, ABMON_8, ABMON_9, ABMON_10,
60 if (MB_CUR_MAX == 1) {
61 if (cmonths == NULL) {
64 cmonths = sort_malloc(sizeof(unsigned char*) * 12);
65 for (int i = 0; i < 12; i++) {
67 tmp = (unsigned char *) nl_langinfo(item[i]);
69 printf("month[%d]=%s\n", i, tmp);
74 for (unsigned int j = 0; j < len; j++)
81 if (wmonths == NULL) {
84 wmonths = sort_malloc(sizeof(wchar_t *) * 12);
85 for (int i = 0; i < 12; i++) {
87 tmp = (unsigned char *) nl_langinfo(item[i]);
89 printf("month[%d]=%s\n", i, tmp);
93 m = sort_malloc(SIZEOF_WCHAR_STRING(len + 1));
94 if (mbstowcs(m, (char*)tmp, len) ==
100 for (unsigned int j = 0; j < len; j++)
101 m[j] = towupper(m[j]);
109 * Compare two wide-character strings
112 wide_str_coll(const wchar_t *s1, const wchar_t *s2)
117 ret = wcscoll(s1, s2);
118 if (errno == EILSEQ) {
120 ret = wcscmp(s1, s2);
122 for (size_t i = 0; ; ++i) {
126 return ((c2 == L'\0') ? 0 : -1);
131 return ((int)(c1 - c2));
138 /* counterparts of wcs functions */
141 bwsprintf(FILE *f, struct bwstring *bws, const char *prefix, const char *suffix)
145 fprintf(f, "%s%s%s", prefix, bws->data.cstr, suffix);
147 fprintf(f, "%s%S%s", prefix, bws->data.wstr, suffix);
150 const void* bwsrawdata(const struct bwstring *bws)
153 return (&(bws->data));
156 size_t bwsrawlen(const struct bwstring *bws)
159 return ((MB_CUR_MAX == 1) ? bws->len : SIZEOF_WCHAR_STRING(bws->len));
163 bws_memsize(const struct bwstring *bws)
166 return ((MB_CUR_MAX == 1) ? (bws->len + 2 + sizeof(struct bwstring)) :
167 (SIZEOF_WCHAR_STRING(bws->len + 1) + sizeof(struct bwstring)));
171 bws_setlen(struct bwstring *bws, size_t newlen)
174 if (bws && newlen != bws->len && newlen <= bws->len) {
177 bws->data.cstr[newlen] = '\0';
179 bws->data.wstr[newlen] = L'\0';
184 * Allocate a new binary string of specified size
189 struct bwstring *ret;
192 ret = sort_malloc(sizeof(struct bwstring) + 1 + sz);
194 ret = sort_malloc(sizeof(struct bwstring) +
195 SIZEOF_WCHAR_STRING(sz + 1));
199 ret->data.cstr[ret->len] = '\0';
201 ret->data.wstr[ret->len] = L'\0';
207 * Create a copy of binary string.
208 * New string size equals the length of the old string.
211 bwsdup(const struct bwstring *s)
217 struct bwstring *ret = bwsalloc(s->len);
220 memcpy(ret->data.cstr, s->data.cstr, (s->len));
222 memcpy(ret->data.wstr, s->data.wstr,
223 SIZEOF_WCHAR_STRING(s->len));
230 * Create a new binary string from a raw binary buffer.
233 bwssbdup(const wchar_t *str, size_t len)
237 return ((len == 0) ? bwsalloc(0) : NULL);
239 struct bwstring *ret;
244 for (size_t i = 0; i < len; ++i)
245 ret->data.cstr[i] = (unsigned char) str[i];
247 memcpy(ret->data.wstr, str, SIZEOF_WCHAR_STRING(len));
254 * Create a new binary string from a raw binary buffer.
257 bwscsbdup(const unsigned char *str, size_t len)
259 struct bwstring *ret;
265 memcpy(ret->data.cstr, str, len);
269 size_t charlen, chars, cptr;
273 s = (const char *) str;
275 memset(&mbs, 0, sizeof(mbs));
278 size_t n = MB_CUR_MAX;
282 charlen = mbrlen(s + cptr, n, &mbs);
289 ret->data.wstr[chars++] =
290 (unsigned char) s[cptr];
294 n = mbrtowc(ret->data.wstr + (chars++),
295 s + cptr, charlen, &mbs);
296 if ((n == (size_t)-1) || (n == (size_t)-2))
298 err(2, "mbrtowc error");
304 ret->data.wstr[ret->len] = L'\0';
311 * De-allocate object memory
314 bwsfree(const struct bwstring *s)
322 * Copy content of src binary string to dst.
323 * If the capacity of the dst string is not sufficient,
324 * then the data is truncated.
327 bwscpy(struct bwstring *dst, const struct bwstring *src)
329 size_t nums = src->len;
335 if (MB_CUR_MAX == 1) {
336 memcpy(dst->data.cstr, src->data.cstr, nums);
337 dst->data.cstr[dst->len] = '\0';
339 memcpy(dst->data.wstr, src->data.wstr,
340 SIZEOF_WCHAR_STRING(nums + 1));
341 dst->data.wstr[dst->len] = L'\0';
348 * Copy content of src binary string to dst,
349 * with specified number of symbols to be copied.
350 * If the capacity of the dst string is not sufficient,
351 * then the data is truncated.
354 bwsncpy(struct bwstring *dst, const struct bwstring *src, size_t size)
356 size_t nums = src->len;
364 if (MB_CUR_MAX == 1) {
365 memcpy(dst->data.cstr, src->data.cstr, nums);
366 dst->data.cstr[dst->len] = '\0';
368 memcpy(dst->data.wstr, src->data.wstr,
369 SIZEOF_WCHAR_STRING(nums + 1));
370 dst->data.wstr[dst->len] = L'\0';
377 * Copy content of src binary string to dst,
378 * with specified number of symbols to be copied.
379 * An offset value can be specified, from the start of src string.
380 * If the capacity of the dst string is not sufficient,
381 * then the data is truncated.
384 bwsnocpy(struct bwstring *dst, const struct bwstring *src, size_t offset,
388 if (offset >= src->len) {
389 dst->data.wstr[0] = 0;
392 size_t nums = src->len - offset;
399 if (MB_CUR_MAX == 1) {
400 memcpy(dst->data.cstr, src->data.cstr + offset,
402 dst->data.cstr[dst->len] = '\0';
404 memcpy(dst->data.wstr, src->data.wstr + offset,
405 SIZEOF_WCHAR_STRING(nums));
406 dst->data.wstr[dst->len] = L'\0';
413 * Write binary string to the file.
414 * The output is ended either with '\n' (nl == true)
415 * or '\0' (nl == false).
418 bwsfwrite(struct bwstring *bws, FILE *f, bool zero_ended)
421 if (MB_CUR_MAX == 1) {
422 size_t len = bws->len;
425 bws->data.cstr[len] = '\n';
427 if (fwrite(bws->data.cstr, len + 1, 1, f) < 1)
430 bws->data.cstr[len] = '\0';
431 } else if (fwrite(bws->data.cstr, len + 1, 1, f) < 1)
440 eols = zero_ended ? btowc('\0') : btowc('\n');
442 while (printed < BWSLEN(bws)) {
443 const wchar_t *s = bws->data.wstr + printed;
448 nums = fwprintf(f, L"%lc", *s);
456 nums = fwprintf(f, L"%ls", s);
463 fwprintf(f, L"%lc", eols);
464 return (printed + 1);
469 * Allocate and read a binary string from file.
470 * The strings are nl-ended or zero-ended, depending on the sort setting.
473 bwsfgetln(FILE *f, size_t *len, bool zero_ended, struct reader_buffer *rb)
477 eols = zero_ended ? btowc('\0') : btowc('\n');
479 if (!zero_ended && (MB_CUR_MAX > 1)) {
482 ret = fgetwln(f, len);
490 if (ret[*len - 1] == (wchar_t)eols)
493 return (bwssbdup(ret, *len));
495 } else if (!zero_ended && (MB_CUR_MAX == 1)) {
498 ret = fgetln(f, len);
506 if (ret[*len - 1] == '\n')
509 return (bwscsbdup((unsigned char*)ret, *len));
517 if (2 >= rb->fgetwln_z_buffer_size) {
518 rb->fgetwln_z_buffer_size += 256;
519 rb->fgetwln_z_buffer = sort_realloc(rb->fgetwln_z_buffer,
520 sizeof(wchar_t) * rb->fgetwln_z_buffer_size);
522 rb->fgetwln_z_buffer[*len] = 0;
538 if (*len + 1 >= rb->fgetwln_z_buffer_size) {
539 rb->fgetwln_z_buffer_size += 256;
540 rb->fgetwln_z_buffer = sort_realloc(rb->fgetwln_z_buffer,
541 SIZEOF_WCHAR_STRING(rb->fgetwln_z_buffer_size));
544 rb->fgetwln_z_buffer[*len] = c;
545 rb->fgetwln_z_buffer[++(*len)] = 0;
561 if (*len + 1 >= rb->fgetwln_z_buffer_size) {
562 rb->fgetwln_z_buffer_size += 256;
563 rb->fgetwln_z_buffer = sort_realloc(rb->fgetwln_z_buffer,
564 SIZEOF_WCHAR_STRING(rb->fgetwln_z_buffer_size));
567 rb->fgetwln_z_buffer[*len] = c;
568 rb->fgetwln_z_buffer[++(*len)] = 0;
572 /* we do not count the last 0 */
573 return (bwssbdup(rb->fgetwln_z_buffer, *len));
578 bwsncmp(const struct bwstring *bws1, const struct bwstring *bws2,
579 size_t offset, size_t len)
581 size_t cmp_len, len1, len2;
588 if (len1 <= offset) {
589 return ((len2 <= offset) ? 0 : -1);
605 if (MB_CUR_MAX == 1) {
606 const unsigned char *s1, *s2;
608 s1 = bws1->data.cstr + offset;
609 s2 = bws2->data.cstr + offset;
611 res = memcmp(s1, s2, cmp_len);
614 const wchar_t *s1, *s2;
616 s1 = bws1->data.wstr + offset;
617 s2 = bws2->data.wstr + offset;
619 res = memcmp(s1, s2, SIZEOF_WCHAR_STRING(cmp_len));
625 if (len1 < cmp_len && len1 < len2)
627 else if (len2 < cmp_len && len2 < len1)
635 bwscmp(const struct bwstring *bws1, const struct bwstring *bws2, size_t offset)
637 size_t len1, len2, cmp_len;
651 res = bwsncmp(bws1, bws2, offset, cmp_len);
656 else if (len2 < len1)
664 bws_iterator_cmp(bwstring_iterator iter1, bwstring_iterator iter2, size_t len)
669 for (i = 0; i < len; ++i) {
670 c1 = bws_get_iter_value(iter1);
671 c2 = bws_get_iter_value(iter2);
674 iter1 = bws_iterator_inc(iter1, 1);
675 iter2 = bws_iterator_inc(iter2, 1);
682 bwscoll(const struct bwstring *bws1, const struct bwstring *bws2, size_t offset)
690 return ((len2 <= offset) ? 0 : -1);
698 if (MB_CUR_MAX == 1) {
699 const unsigned char *s1, *s2;
701 s1 = bws1->data.cstr + offset;
702 s2 = bws2->data.cstr + offset;
708 res = memcmp(s1, s2, len2);
711 } else if (len1 < len2) {
712 res = memcmp(s1, s2, len1);
716 res = memcmp(s1, s2, len1);
731 /* goto next non-zero part: */
732 while ((i < maxlen) &&
742 err(2, "bwscoll error 01");
745 } else if (s2[i] == 0)
748 res = strcoll((const char*)(s1 + i), (const char*)(s2 + i));
752 while ((i < maxlen) &&
765 } else if (s2[i] == 0)
769 err(2, "bwscoll error 02");
774 else if (len1 > len2)
780 const wchar_t *s1, *s2;
784 s1 = bws1->data.wstr + offset;
785 s2 = bws2->data.wstr + offset;
795 /* goto next non-zero part: */
796 while ((i < maxlen) &&
806 err(2, "bwscoll error 1");
809 } else if (s2[i] == 0)
812 res = wide_str_coll(s1 + i, s2 + i);
816 while ((i < maxlen) && s1[i] && s2[i])
828 } else if (s2[i] == 0)
832 err(2, "bwscoll error 2");
837 else if (len1 > len2)
847 * Correction of the system API
850 bwstod(struct bwstring *s0, bool *empty)
854 if (MB_CUR_MAX == 1) {
855 unsigned char *end, *s;
862 while (isblank(*s) && s < end)
870 ret = strtod((char*)s, &ep);
871 if ((unsigned char*) ep == s) {
876 wchar_t *end, *ep, *s;
882 while (iswblank(*s) && s < end)
890 ret = wcstod(s, &ep);
902 * A helper function for monthcoll. If a line matches
903 * a month name, it returns (number of the month - 1),
904 * while if there is no match, it just return -1.
908 bws_month_score(const struct bwstring *s0)
911 if (MB_CUR_MAX == 1) {
912 const unsigned char *end, *s;
918 while (isblank(*s) && s < end)
921 len = strlen((const char*)s);
923 for (int i = 11; i >= 0; --i) {
925 (s == (unsigned char*)strstr((const char*)s, (char*)(cmonths[i]))))
930 const wchar_t *end, *s;
936 while (iswblank(*s) && s < end)
941 for (int i = 11; i >= 0; --i) {
942 if (wmonths[i] && (s == wcsstr(s, wmonths[i])))
951 * Rips out leading blanks (-b).
954 ignore_leading_blanks(struct bwstring *str)
957 if (MB_CUR_MAX == 1) {
958 unsigned char *dst, *end, *src;
960 src = str->data.cstr;
962 end = src + str->len;
964 while (src < end && isblank(*src))
970 newlen = BWSLEN(str) - (src - dst);
977 bws_setlen(str, newlen);
980 wchar_t *dst, *end, *src;
982 src = str->data.wstr;
984 end = src + str->len;
986 while (src < end && iswblank(*src))
991 size_t newlen = BWSLEN(str) - (src - dst);
998 bws_setlen(str, newlen);
1006 * Rips out nonprinting characters (-i).
1009 ignore_nonprinting(struct bwstring *str)
1011 size_t newlen = str->len;
1013 if (MB_CUR_MAX == 1) {
1014 unsigned char *dst, *end, *src;
1017 src = str->data.cstr;
1019 end = src + str->len;
1033 wchar_t *dst, *end, *src;
1036 src = str->data.wstr;
1038 end = src + str->len;
1052 bws_setlen(str, newlen);
1058 * Rips out any characters that are not alphanumeric characters
1062 dictionary_order(struct bwstring *str)
1064 size_t newlen = str->len;
1066 if (MB_CUR_MAX == 1) {
1067 unsigned char *dst, *end, *src;
1070 src = str->data.cstr;
1072 end = src + str->len;
1076 if (isalnum(c) || isblank(c)) {
1086 wchar_t *dst, *end, *src;
1089 src = str->data.wstr;
1091 end = src + str->len;
1095 if (iswalnum(c) || iswblank(c)) {
1105 bws_setlen(str, newlen);
1111 * Converts string to lower case(-f).
1114 ignore_case(struct bwstring *str)
1117 if (MB_CUR_MAX == 1) {
1118 unsigned char *end, *s;
1142 bws_disorder_warnx(struct bwstring *s, const char *fn, size_t pos)
1145 if (MB_CUR_MAX == 1)
1146 warnx("%s:%zu: disorder: %s", fn, pos + 1, s->data.cstr);
1148 warnx("%s:%zu: disorder: %ls", fn, pos + 1, s->data.wstr);