xref: /freebsd/usr.bin/sort/bwstring.h (revision fcb560670601b2a4d87bb31d7531c8dcc37ee71b)
1 /*	$FreeBSD$	*/
2 
3 /*-
4  * Copyright (C) 2009 Gabor Kovesdan <gabor@FreeBSD.org>
5  * Copyright (C) 2012 Oleg Moskalenko <mom040267@gmail.com>
6  * All rights reserved.
7  *
8  * Redistribution and use in source and binary forms, with or without
9  * modification, are permitted provided that the following conditions
10  * are met:
11  * 1. Redistributions of source code must retain the above copyright
12  *    notice, this list of conditions and the following disclaimer.
13  * 2. Redistributions in binary form must reproduce the above copyright
14  *    notice, this list of conditions and the following disclaimer in the
15  *    documentation and/or other materials provided with the distribution.
16  *
17  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
18  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20  * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
21  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
22  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
23  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
24  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
25  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
26  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
27  * SUCH DAMAGE.
28  */
29 
30 #if !defined(__BWSTRING_H__)
31 #define	__BWSTRING_H__
32 
33 #include <stdbool.h>
34 #include <stdio.h>
35 #include <errno.h>
36 #include <sysexits.h>
37 #include <wchar.h>
38 
39 #include "mem.h"
40 
41 extern bool byte_sort;
42 
43 /* wchar_t is of 4 bytes: */
44 #define	SIZEOF_WCHAR_STRING(LEN) ((LEN)*sizeof(wchar_t))
45 
46 /*
47  * Binary "wide" string
48  */
49 struct bwstring
50 {
51 	size_t				len;
52 	union
53 	{
54 		wchar_t		wstr[0];
55 		unsigned char	cstr[0];
56 	}				data;
57 };
58 
59 struct reader_buffer
60 {
61 	wchar_t			*fgetwln_z_buffer;
62 	size_t			 fgetwln_z_buffer_size;
63 };
64 
65 typedef void *bwstring_iterator;
66 
67 #define	BWSLEN(s) ((s)->len)
68 
69 struct bwstring *bwsalloc(size_t sz);
70 
71 size_t bwsrawlen(const struct bwstring *bws);
72 const void* bwsrawdata(const struct bwstring *bws);
73 void bws_setlen(struct bwstring *bws, size_t newlen);
74 size_t bws_memsize(const struct bwstring *bws);
75 double bwstod(struct bwstring *s0, bool *empty);
76 int bws_month_score(const struct bwstring *s0);
77 
78 struct bwstring *ignore_leading_blanks(struct bwstring *str);
79 struct bwstring *ignore_nonprinting(struct bwstring *str);
80 struct bwstring *dictionary_order(struct bwstring *str);
81 struct bwstring *ignore_case(struct bwstring *str);
82 
83 void bwsprintf(FILE*, struct bwstring*, const char *prefix, const char *suffix);
84 void bws_disorder_warnx(struct bwstring *s, const char *fn, size_t pos);
85 
86 struct bwstring *bwsdup(const struct bwstring *s);
87 struct bwstring *bwssbdup(const wchar_t *str, size_t size);
88 struct bwstring *bwscsbdup(const unsigned char *str, size_t size);
89 void bwsfree(const struct bwstring *s);
90 size_t bwscpy(struct bwstring *dst, const struct bwstring *src);
91 struct bwstring *bwsncpy(struct bwstring *dst, const struct bwstring *src, size_t size);
92 struct bwstring *bwsnocpy(struct bwstring *dst, const struct bwstring *src, size_t offset, size_t size);
93 int bwscmp(const struct bwstring *bws1, const struct bwstring *bws2, size_t offset);
94 int bwsncmp(const struct bwstring *bws1, const struct bwstring *bws2, size_t offset, size_t len);
95 int bwscoll(const struct bwstring *bws1, const struct bwstring *bws2, size_t offset);
96 size_t bwsfwrite(struct bwstring *bws, FILE *f, bool zero_ended);
97 struct bwstring *bwsfgetln(FILE *file, size_t *len, bool zero_ended, struct reader_buffer *rb);
98 
99 static inline bwstring_iterator
100 bws_begin(struct bwstring *bws)
101 {
102 
103 	return (bwstring_iterator) (&(bws->data));
104 }
105 
106 static inline bwstring_iterator
107 bws_end(struct bwstring *bws)
108 {
109 
110 	return ((MB_CUR_MAX == 1) ?
111 	    (bwstring_iterator) (bws->data.cstr + bws->len) :
112 	    (bwstring_iterator) (bws->data.wstr + bws->len));
113 }
114 
115 static inline bwstring_iterator
116 bws_iterator_inc(bwstring_iterator iter, size_t pos)
117 {
118 
119 	if (MB_CUR_MAX == 1)
120 		return ((unsigned char *) iter) + pos;
121 	else
122 		return ((wchar_t*) iter) + pos;
123 }
124 
125 static inline wchar_t
126 bws_get_iter_value(bwstring_iterator iter)
127 {
128 
129 	if (MB_CUR_MAX == 1)
130 		return *((unsigned char *) iter);
131 	else
132 		return *((wchar_t*) iter);
133 }
134 
135 int
136 bws_iterator_cmp(bwstring_iterator iter1, bwstring_iterator iter2, size_t len);
137 
138 #define	BWS_GET(bws, pos) ((MB_CUR_MAX == 1) ? ((bws)->data.cstr[(pos)]) : (bws)->data.wstr[(pos)])
139 
140 void initialise_months(void);
141 
142 #endif /* __BWSTRING_H__ */
143