Current Path : /compat/linux/proc/self/root/usr/src/usr.bin/tr/ |
FreeBSD hs32.drive.ne.jp 9.1-RELEASE FreeBSD 9.1-RELEASE #1: Wed Jan 14 12:18:08 JST 2015 root@hs32.drive.ne.jp:/sys/amd64/compile/hs32 amd64 |
Current File : //compat/linux/proc/self/root/usr/src/usr.bin/tr/str.c |
/*- * Copyright (c) 1991, 1993 * The Regents of the University of California. All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * 4. Neither the name of the University nor the names of its contributors * may be used to endorse or promote products derived from this software * without specific prior written permission. * * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF * SUCH DAMAGE. */ #include <sys/cdefs.h> __FBSDID("$FreeBSD: release/9.1.0/usr.bin/tr/str.c 216370 2010-12-11 08:32:16Z joel $"); #ifndef lint static const char sccsid[] = "@(#)str.c 8.2 (Berkeley) 4/28/95"; #endif #include <sys/types.h> #include <ctype.h> #include <err.h> #include <errno.h> #include <stddef.h> #include <stdio.h> #include <stdlib.h> #include <string.h> #include <wchar.h> #include <wctype.h> #include "extern.h" static int backslash(STR *, int *); static int bracket(STR *); static void genclass(STR *); static void genequiv(STR *); static int genrange(STR *, int); static void genseq(STR *); wint_t next(s) STR *s; { int is_octal; wint_t ch; wchar_t wch; size_t clen; switch (s->state) { case EOS: return (0); case INFINITE: return (1); case NORMAL: switch (*s->str) { case '\0': s->state = EOS; return (0); case '\\': s->lastch = backslash(s, &is_octal); break; case '[': if (bracket(s)) return (next(s)); /* FALLTHROUGH */ default: clen = mbrtowc(&wch, s->str, MB_LEN_MAX, NULL); if (clen == (size_t)-1 || clen == (size_t)-2 || clen == 0) errc(1, EILSEQ, NULL); is_octal = 0; s->lastch = wch; s->str += clen; break; } /* We can start a range at any time. */ if (s->str[0] == '-' && genrange(s, is_octal)) return (next(s)); return (1); case RANGE: if (s->cnt-- == 0) { s->state = NORMAL; return (next(s)); } ++s->lastch; return (1); case SEQUENCE: if (s->cnt-- == 0) { s->state = NORMAL; return (next(s)); } return (1); case CCLASS: case CCLASS_UPPER: case CCLASS_LOWER: s->cnt++; ch = nextwctype(s->lastch, s->cclass); if (ch == -1) { s->state = NORMAL; return (next(s)); } s->lastch = ch; return (1); case SET: if ((ch = s->set[s->cnt++]) == OOBCH) { s->state = NORMAL; return (next(s)); } s->lastch = ch; return (1); default: return (0); } /* NOTREACHED */ } static int bracket(s) STR *s; { char *p; switch (s->str[1]) { case ':': /* "[:class:]" */ if ((p = strchr(s->str + 2, ']')) == NULL) return (0); if (*(p - 1) != ':' || p - s->str < 4) goto repeat; *(p - 1) = '\0'; s->str += 2; genclass(s); s->str = p + 1; return (1); case '=': /* "[=equiv=]" */ if (s->str[2] == '\0' || (p = strchr(s->str + 3, ']')) == NULL) return (0); if (*(p - 1) != '=' || p - s->str < 4) goto repeat; s->str += 2; genequiv(s); return (1); default: /* "[\###*n]" or "[#*n]" */ repeat: if ((p = strpbrk(s->str + 2, "*]")) == NULL) return (0); if (p[0] != '*' || index(p, ']') == NULL) return (0); s->str += 1; genseq(s); return (1); } /* NOTREACHED */ } static void genclass(s) STR *s; { if ((s->cclass = wctype(s->str)) == 0) errx(1, "unknown class %s", s->str); s->cnt = 0; s->lastch = -1; /* incremented before check in next() */ if (strcmp(s->str, "upper") == 0) s->state = CCLASS_UPPER; else if (strcmp(s->str, "lower") == 0) s->state = CCLASS_LOWER; else s->state = CCLASS; } static void genequiv(s) STR *s; { int i, p, pri; char src[2], dst[3]; size_t clen; wchar_t wc; if (*s->str == '\\') { s->equiv[0] = backslash(s, NULL); if (*s->str != '=') errx(1, "misplaced equivalence equals sign"); s->str += 2; } else { clen = mbrtowc(&wc, s->str, MB_LEN_MAX, NULL); if (clen == (size_t)-1 || clen == (size_t)-2 || clen == 0) errc(1, EILSEQ, NULL); s->equiv[0] = wc; if (s->str[clen] != '=') errx(1, "misplaced equivalence equals sign"); s->str += clen + 2; } /* * Calculate the set of all characters in the same equivalence class * as the specified character (they will have the same primary * collation weights). * XXX Knows too much about how strxfrm() is implemented. Assumes * it fills the string with primary collation weight bytes. Only one- * to-one mappings are supported. * XXX Equivalence classes not supported in multibyte locales. */ src[0] = (char)s->equiv[0]; src[1] = '\0'; if (MB_CUR_MAX == 1 && strxfrm(dst, src, sizeof(dst)) == 1) { pri = (unsigned char)*dst; for (p = 1, i = 1; i < NCHARS_SB; i++) { *src = i; if (strxfrm(dst, src, sizeof(dst)) == 1 && pri && pri == (unsigned char)*dst) s->equiv[p++] = i; } s->equiv[p] = OOBCH; } s->cnt = 0; s->state = SET; s->set = s->equiv; } static int genrange(STR *s, int was_octal) { int stopval, octal; char *savestart; int n, cnt, *p; size_t clen; wchar_t wc; octal = 0; savestart = s->str; if (*++s->str == '\\') stopval = backslash(s, &octal); else { clen = mbrtowc(&wc, s->str, MB_LEN_MAX, NULL); if (clen == (size_t)-1 || clen == (size_t)-2) errc(1, EILSEQ, NULL); stopval = wc; s->str += clen; } /* * XXX Characters are not ordered according to collating sequence in * multibyte locales. */ if (octal || was_octal || MB_CUR_MAX > 1) { if (stopval < s->lastch) { s->str = savestart; return (0); } s->cnt = stopval - s->lastch + 1; s->state = RANGE; --s->lastch; return (1); } if (charcoll((const void *)&stopval, (const void *)&(s->lastch)) < 0) { s->str = savestart; return (0); } if ((s->set = p = malloc((NCHARS_SB + 1) * sizeof(int))) == NULL) err(1, "genrange() malloc"); for (cnt = 0; cnt < NCHARS_SB; cnt++) if (charcoll((const void *)&cnt, (const void *)&(s->lastch)) >= 0 && charcoll((const void *)&cnt, (const void *)&stopval) <= 0) *p++ = cnt; *p = OOBCH; n = p - s->set; s->cnt = 0; s->state = SET; if (n > 1) mergesort(s->set, n, sizeof(*(s->set)), charcoll); return (1); } static void genseq(s) STR *s; { char *ep; wchar_t wc; size_t clen; if (s->which == STRING1) errx(1, "sequences only valid in string2"); if (*s->str == '\\') s->lastch = backslash(s, NULL); else { clen = mbrtowc(&wc, s->str, MB_LEN_MAX, NULL); if (clen == (size_t)-1 || clen == (size_t)-2) errc(1, EILSEQ, NULL); s->lastch = wc; s->str += clen; } if (*s->str != '*') errx(1, "misplaced sequence asterisk"); switch (*++s->str) { case '\\': s->cnt = backslash(s, NULL); break; case ']': s->cnt = 0; ++s->str; break; default: if (isdigit((u_char)*s->str)) { s->cnt = strtol(s->str, &ep, 0); if (*ep == ']') { s->str = ep + 1; break; } } errx(1, "illegal sequence count"); /* NOTREACHED */ } s->state = s->cnt ? SEQUENCE : INFINITE; } /* * Translate \??? into a character. Up to 3 octal digits, if no digits either * an escape code or a literal character. */ static int backslash(STR *s, int *is_octal) { int ch, cnt, val; if (is_octal != NULL) *is_octal = 0; for (cnt = val = 0;;) { ch = (u_char)*++s->str; if (!isdigit(ch) || ch > '7') break; val = val * 8 + ch - '0'; if (++cnt == 3) { ++s->str; break; } } if (cnt) { if (is_octal != NULL) *is_octal = 1; return (val); } if (ch != '\0') ++s->str; switch (ch) { case 'a': /* escape characters */ return ('\7'); case 'b': return ('\b'); case 'f': return ('\f'); case 'n': return ('\n'); case 'r': return ('\r'); case 't': return ('\t'); case 'v': return ('\13'); case '\0': /* \" -> \ */ s->state = EOS; return ('\\'); default: /* \x" -> x */ return (ch); } }