xref: /openbsd/usr.bin/sort/bwstring.h (revision 264ca280)
1 /*	$OpenBSD: bwstring.h,v 1.2 2015/12/31 16:09:31 millert Exp $	*/
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 <wchar.h>
37 
38 #include "mem.h"
39 
40 extern bool byte_sort;
41 extern size_t sort_mb_cur_max;
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 	size_t				len;
51 	union
52 	{
53 		wchar_t		wstr[0];
54 		unsigned char	cstr[0];
55 	}				data;
56 };
57 
58 struct reader_buffer {
59 	wchar_t			*fgetwln_z_buffer;
60 	size_t			 fgetwln_z_buffer_size;
61 };
62 
63 typedef void *bwstring_iterator;
64 
65 #define	BWSLEN(s) ((s)->len)
66 
67 struct bwstring *bwsalloc(size_t sz);
68 
69 size_t bwsrawlen(const struct bwstring *bws);
70 const void* bwsrawdata(const struct bwstring *bws);
71 void bws_setlen(struct bwstring *bws, size_t newlen);
72 size_t bws_memsize(const struct bwstring *bws);
73 double bwstod(struct bwstring *s0, bool *empty);
74 int bws_month_score(const struct bwstring *s0);
75 
76 struct bwstring *ignore_leading_blanks(struct bwstring *str);
77 struct bwstring *ignore_nonprinting(struct bwstring *str);
78 struct bwstring *dictionary_order(struct bwstring *str);
79 struct bwstring *ignore_case(struct bwstring *str);
80 
81 void bwsprintf(FILE*, struct bwstring*, const char *prefix, const char *suffix);
82 void bws_disorder_warnx(struct bwstring *s, const char *fn, size_t pos);
83 
84 struct bwstring *bwsdup(const struct bwstring *s);
85 struct bwstring *bwssbdup(const wchar_t *str, size_t size);
86 struct bwstring *bwscsbdup(const unsigned char *str, size_t size);
87 void bwsfree(struct bwstring *s);
88 size_t bwscpy(struct bwstring *dst, const struct bwstring *src);
89 struct bwstring *bwsncpy(struct bwstring *dst, const struct bwstring *src, size_t size);
90 struct bwstring *bwsnocpy(struct bwstring *dst, const struct bwstring *src, size_t offset, size_t size);
91 int bwscmp(const struct bwstring *bws1, const struct bwstring *bws2, size_t offset);
92 int bwsncmp(const struct bwstring *bws1, const struct bwstring *bws2, size_t offset, size_t len);
93 int bwscoll(const struct bwstring *bws1, const struct bwstring *bws2, size_t offset);
94 size_t bwsfwrite(struct bwstring *bws, FILE *f, bool zero_ended);
95 struct bwstring *bwsfgetln(FILE *file, size_t *len, bool zero_ended, struct reader_buffer *rb);
96 
97 static inline bwstring_iterator
98 bws_begin(struct bwstring *bws)
99 {
100 
101 	return (bwstring_iterator) (&(bws->data));
102 }
103 
104 static inline bwstring_iterator
105 bws_end(struct bwstring *bws)
106 {
107 
108 	return ((sort_mb_cur_max == 1) ?
109 	    (bwstring_iterator) (bws->data.cstr + bws->len) :
110 	    (bwstring_iterator) (bws->data.wstr + bws->len));
111 }
112 
113 static inline bwstring_iterator
114 bws_iterator_inc(bwstring_iterator iter, size_t pos)
115 {
116 
117 	if (sort_mb_cur_max == 1)
118 		return ((unsigned char *) iter) + pos;
119 	else
120 		return ((wchar_t*) iter) + pos;
121 }
122 
123 static inline wchar_t
124 bws_get_iter_value(bwstring_iterator iter)
125 {
126 
127 	if (sort_mb_cur_max == 1)
128 		return *((unsigned char *) iter);
129 	else
130 		return *((wchar_t*) iter);
131 }
132 
133 int
134 bws_iterator_cmp(bwstring_iterator iter1, bwstring_iterator iter2, size_t len);
135 
136 #define	BWS_GET(bws, pos) ((sort_mb_cur_max == 1) ? ((bws)->data.cstr[(pos)]) : (bws)->data.wstr[(pos)])
137 
138 void initialise_months(void);
139 
140 #endif /* __BWSTRING_H__ */
141