xref: /openbsd-src/lib/libedit/chartype.c (revision 4c1e55dc91edd6e69ccc60ce855900fbc12cf34f)
1 /*	$OpenBSD: chartype.c,v 1.4 2011/11/17 20:14:24 nicm Exp $	*/
2 /*	$NetBSD: chartype.c,v 1.4 2010/04/15 00:55:57 christos Exp $	*/
3 
4 /*-
5  * Copyright (c) 2009 The NetBSD Foundation, Inc.
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  * 3. All advertising materials mentioning features or use of this software
17  *    must display the following acknowledgement:
18  *        This product includes software developed by the NetBSD
19  *        Foundation, Inc. and its contributors.
20  * 4. Neither the name of The NetBSD Foundation nor the names of its
21  *    contributors may be used to endorse or promote products derived
22  *    from this software without specific prior written permission.
23  *
24  * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
25  * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
26  * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
27  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
28  * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
29  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
30  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
31  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
32  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
33  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
34  * POSSIBILITY OF SUCH DAMAGE.
35  */
36 
37 /*
38  * chartype.c: character classification and meta information
39  */
40 #include "config.h"
41 #include "el.h"
42 #include <stdlib.h>
43 
44 #define CT_BUFSIZ 1024
45 
46 #ifdef WIDECHAR
47 protected void
48 ct_conv_buff_resize(ct_buffer_t *conv, size_t mincsize, size_t minwsize)
49 {
50 	void *p;
51 	if (mincsize > conv->csize) {
52 		conv->csize = mincsize;
53 		p = el_realloc(conv->cbuff, conv->csize);
54 		if (p == NULL) {
55 			conv->csize = 0;
56 			el_free(conv->cbuff);
57 			conv->cbuff = NULL;
58 		} else
59 			conv->cbuff = p;
60 	}
61 
62 	if (minwsize > conv->wsize) {
63 		conv->wsize = minwsize;
64 		p = el_realloc(conv->wbuff, conv->wsize);
65 		if (p == NULL) {
66 			conv->wsize = 0;
67 			el_free(conv->wbuff);
68 			conv->wbuff = NULL;
69 		} else
70 			conv->wbuff = p;
71 	}
72 }
73 
74 
75 public char *
76 ct_encode_string(const Char *s, ct_buffer_t *conv)
77 {
78 	char *dst;
79 	ssize_t used = 0;
80 
81 	if (!s)
82 		return NULL;
83 	if (!conv->cbuff)
84 		ct_conv_buff_resize(conv, CT_BUFSIZ, 0);
85 	if (!conv->cbuff)
86 		return NULL;
87 
88 	dst = conv->cbuff;
89 	while (*s) {
90 		used = ct_encode_char(dst, (int)(conv->csize -
91 		    (dst - conv->cbuff)), *s);
92 		if (used == -1) { /* failed to encode, need more buffer space */
93 			used = dst - conv->cbuff;
94 			ct_conv_buff_resize(conv, conv->csize + CT_BUFSIZ, 0);
95 			if (!conv->cbuff)
96 				return NULL;
97 			dst = conv->cbuff + used;
98 			/* don't increment s here - we want to retry it! */
99 		}
100 		else
101 			++s;
102 		dst += used;
103 	}
104 	if (dst >= (conv->cbuff + conv->csize)) {
105 		used = dst - conv->cbuff;
106 		ct_conv_buff_resize(conv, conv->csize + 1, 0);
107 		if (!conv->cbuff)
108 			return NULL;
109 		dst = conv->cbuff + used;
110 	}
111 	*dst = '\0';
112 	return conv->cbuff;
113 }
114 
115 public Char *
116 ct_decode_string(const char *s, ct_buffer_t *conv)
117 {
118 	size_t len = 0;
119 
120 	if (!s)
121 		return NULL;
122 	if (!conv->wbuff)
123 		ct_conv_buff_resize(conv, 0, CT_BUFSIZ);
124 	if (!conv->wbuff)
125 		return NULL;
126 
127 	len = ct_mbstowcs(0, s, 0);
128 	if (len > conv->wsize)
129 		ct_conv_buff_resize(conv, 0, len + 1);
130 	if (!conv->wbuff)
131 		return NULL;
132 	ct_mbstowcs(conv->wbuff, s, conv->wsize);
133 	return conv->wbuff;
134 }
135 
136 
137 protected Char **
138 ct_decode_argv(int argc, const char *argv[], ct_buffer_t *conv)
139 {
140 	size_t bufspace;
141 	int i;
142 	Char *p;
143 	Char **wargv;
144 	size_t wlen;
145 
146 	/* Make sure we have enough space in the conversion buffer to store all
147 	 * the argv strings. */
148 	for (i = 0, bufspace = 0; i < argc; ++i)
149 		bufspace += argv[i] ? strlen(argv[i]) + 1 : 0;
150 	ct_conv_buff_resize(conv, 0, bufspace * sizeof(*p));
151 	if (!conv->wsize)
152 		return NULL;
153 
154 	wargv = el_malloc(argc * sizeof(*wargv));
155 
156 	for (i = 0, p = conv->wbuff; i < argc; ++i) {
157 		if (!argv[i]) {   /* don't pass null pointers to mbstowcs */
158 			wargv[i] = NULL;
159 			continue;
160 		} else {
161 			wargv[i] = p;
162 			wlen = mbstowcs(p, argv[i], bufspace);
163 		}
164 		if (wlen == (size_t)-1 || wlen == bufspace) {
165 			/* Encoding error or not enough room for NUL. */
166 			el_free(wargv);
167 			return NULL;
168 		} else
169 			wlen++; /* include NUL in the count */
170 		bufspace -= wlen;
171 		p += wlen;
172 	}
173 
174 	return wargv;
175 }
176 
177 
178 protected size_t
179 ct_enc_width(Char c)
180 {
181 	/* UTF-8 encoding specific values */
182 	if (c < 0x80)
183 		return 1;
184 	else if (c < 0x0800)
185 		return 2;
186 	else if (c < 0x10000)
187 		return 3;
188 	else if (c < 0x110000)
189 		return 4;
190 	else
191 		return 0; /* not a valid codepoint */
192 }
193 
194 protected ssize_t
195 ct_encode_char(char *dst, size_t len, Char c)
196 {
197 	ssize_t l = 0;
198 	if (len < ct_enc_width(c))
199 		return -1;
200 	l = ct_wctomb(dst, c);
201 
202 	if (l < 0) {
203 		ct_wctomb_reset;
204 		l = 0;
205 	}
206 	return l;
207 }
208 #endif
209 
210 protected const Char *
211 ct_visual_string(const Char *s)
212 {
213 	static Char *buff = NULL;
214 	static size_t buffsize = 0;
215 	void *p;
216 	Char *dst;
217 	ssize_t used = 0;
218 
219 	if (!s)
220 		return NULL;
221 	if (!buff) {
222 	    buffsize = CT_BUFSIZ;
223 	    buff = el_malloc(buffsize * sizeof(*buff));
224 	}
225 	dst = buff;
226 	while (*s) {
227 		used = ct_visual_char(dst, buffsize - (dst - buff), *s);
228 		if (used == -1) { /* failed to encode, need more buffer space */
229 			used = dst - buff;
230 			buffsize += CT_BUFSIZ;
231 			p = el_realloc(buff, buffsize * sizeof(*buff));
232 			if (p == NULL)
233 				goto out;
234 			buff = p;
235 			dst = buff + used;
236 			/* don't increment s here - we want to retry it! */
237 		}
238 		else
239 		    ++s;
240 		dst += used;
241 	}
242 	if (dst >= (buff + buffsize)) { /* sigh */
243 		buffsize += 1;
244 		p = el_realloc(buff, buffsize * sizeof(*buff));
245 		if (p == NULL)
246 			goto out;
247 		buff = p;
248 		dst = buff + buffsize - 1;
249 	}
250 	*dst = 0;
251 	return buff;
252 out:
253 	el_free(buff);
254 	buffsize = 0;
255 	return NULL;
256 }
257 
258 
259 
260 protected int
261 ct_visual_width(Char c)
262 {
263 	int t = ct_chr_class(c);
264 #ifdef WIDECHAR
265 	int w;
266 #endif
267 	switch (t) {
268 	case CHTYPE_ASCIICTL:
269 		return 2; /* ^@ ^? etc. */
270 	case CHTYPE_TAB:
271 		return 1; /* Hmm, this really need to be handled outside! */
272 	case CHTYPE_NL:
273 		return 0; /* Should this be 1 instead? */
274 #ifdef WIDECHAR
275 	case CHTYPE_PRINT:
276 		w = wcwidth(c);
277 		return (w == -1 ? 0 : w);
278 	case CHTYPE_NONPRINT:
279 		if (c > 0xffff) /* prefer standard 4-byte display over 5-byte */
280 			return 8; /* \U+12345 */
281 		else
282 			return 7; /* \U+1234 */
283 #else
284 	case CHTYPE_PRINT:
285 		return 1;
286 	case CHTYPE_NONPRINT:
287 		return 4; /* \123 */
288 #endif
289 	default:
290 		return 0; /* should not happen */
291 	}
292 }
293 
294 
295 protected ssize_t
296 ct_visual_char(Char *dst, size_t len, Char c)
297 {
298 	int t = ct_chr_class(c);
299 	switch (t) {
300 	case CHTYPE_TAB:
301 	case CHTYPE_NL:
302 	case CHTYPE_ASCIICTL:
303 		if (len < 2)
304 			return -1;   /* insufficient space */
305 		*dst++ = '^';
306 		if (c == '\177')
307 			*dst = '?'; /* DEL -> ^? */
308 		else
309 			*dst = c | 0100;    /* uncontrolify it */
310 		return 2;
311 	case CHTYPE_PRINT:
312 		if (len < 1)
313 			return -1;  /* insufficient space */
314 		*dst = c;
315 		return 1;
316 	case CHTYPE_NONPRINT:
317 		/* we only use single-width glyphs for display,
318 		 * so this is right */
319 		if ((ssize_t)len < ct_visual_width(c))
320 			return -1;   /* insufficient space */
321 #ifdef WIDECHAR
322 		*dst++ = '\\';
323 		*dst++ = 'U';
324 		*dst++ = '+';
325 #define tohexdigit(v) "0123456789ABCDEF"[v]
326 		if (c > 0xffff) /* prefer standard 4-byte display over 5-byte */
327 			*dst++ = tohexdigit(((unsigned int) c >> 16) & 0xf);
328 		*dst++ = tohexdigit(((unsigned int) c >> 12) & 0xf);
329 		*dst++ = tohexdigit(((unsigned int) c >>  8) & 0xf);
330 		*dst++ = tohexdigit(((unsigned int) c >>  4) & 0xf);
331 		*dst   = tohexdigit(((unsigned int) c      ) & 0xf);
332 		return (c > 0xffff) ? 8 : 7;
333 #else
334 		*dst++ = '\\';
335 #define tooctaldigit(v) ((v) + '0')
336 		*dst++ = tooctaldigit(((unsigned int) c >> 6) & 0x7);
337 		*dst++ = tooctaldigit(((unsigned int) c >> 3) & 0x7);
338 		*dst++ = tooctaldigit(((unsigned int) c     ) & 0x7);
339 #endif
340 		/*FALLTHROUGH*/
341 	/* these two should be handled outside this function */
342 	default:            /* we should never hit the default */
343 		return 0;
344 	}
345 }
346 
347 
348 
349 
350 protected int
351 ct_chr_class(Char c)
352 {
353 	if (c == '\t')
354 		return CHTYPE_TAB;
355 	else if (c == '\n')
356 		return CHTYPE_NL;
357 	else if (IsASCII(c) && Iscntrl(c))
358 		return CHTYPE_ASCIICTL;
359 	else if (Isprint(c))
360 		return CHTYPE_PRINT;
361 	else
362 		return CHTYPE_NONPRINT;
363 }
364