tc.str.c revision 167465 
1/* $Header: /p/tcsh/cvsroot/tcsh/tc.str.c,v 3.26 2006/03/02 18:46:45 christos Exp $ */
2/*
3 * tc.str.c: Short string package
4 * 	     This has been a lesson of how to write buggy code!
5 */
6/*-
7 * Copyright (c) 1980, 1991 The Regents of the University of California.
8 * All rights reserved.
9 *
10 * Redistribution and use in source and binary forms, with or without
11 * modification, are permitted provided that the following conditions
12 * are met:
13 * 1. Redistributions of source code must retain the above copyright
14 *    notice, this list of conditions and the following disclaimer.
15 * 2. Redistributions in binary form must reproduce the above copyright
16 *    notice, this list of conditions and the following disclaimer in the
17 *    documentation and/or other materials provided with the distribution.
18 * 3. Neither the name of the University nor the names of its contributors
19 *    may be used to endorse or promote products derived from this software
20 *    without specific prior written permission.
21 *
22 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
23 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
24 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
25 * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
26 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
27 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
28 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
29 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
30 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
31 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
32 * SUCH DAMAGE.
33 */
34#include "sh.h"
35
36#include <limits.h>
37
38RCSID("$tcsh: tc.str.c,v 3.26 2006/03/02 18:46:45 christos Exp $")
39
40#define MALLOC_INCR	128
41#ifdef WIDE_STRINGS
42#define MALLOC_SURPLUS	MB_LEN_MAX /* Space for one multibyte character */
43#else
44#define MALLOC_SURPLUS	0
45#endif
46
47#ifdef WIDE_STRINGS
48size_t
49one_mbtowc(wchar_t *pwc, const char *s, size_t n)
50{
51    int len;
52
53    len = rt_mbtowc(pwc, s, n);
54    if (len == -1) {
55        mbtowc(NULL, NULL, 0);
56	*pwc = (unsigned char)*s | INVALID_BYTE;
57    }
58    if (len <= 0)
59	len = 1;
60    return len;
61}
62
63size_t
64one_wctomb(char *s, wchar_t wchar)
65{
66    int len;
67
68    if (wchar & INVALID_BYTE) {
69	s[0] = wchar & 0xFF;
70	len = 1;
71    } else {
72	len = wctomb(s, wchar);
73	if (len == -1)
74	    s[0] = wchar;
75	if (len <= 0)
76	    len = 1;
77    }
78    return len;
79}
80
81int
82rt_mbtowc(wchar_t *pwc, const char *s, size_t n)
83{
84    int ret;
85    char back[MB_LEN_MAX];
86
87    ret = mbtowc(pwc, s, n);
88    if (ret > 0 && (wctomb(back, *pwc) != ret || memcmp(s, back, ret) != 0))
89	ret = -1;
90    return ret;
91}
92#endif
93
94#ifdef SHORT_STRINGS
95Char  **
96blk2short(char **src)
97{
98    size_t     n;
99    Char **sdst, **dst;
100
101    /*
102     * Count
103     */
104    for (n = 0; src[n] != NULL; n++)
105	continue;
106    sdst = dst = xmalloc((n + 1) * sizeof(Char *));
107
108    for (; *src != NULL; src++)
109	*dst++ = SAVE(*src);
110    *dst = NULL;
111    return (sdst);
112}
113
114char  **
115short2blk(Char **src)
116{
117    size_t     n;
118    char **sdst, **dst;
119
120    /*
121     * Count
122     */
123    for (n = 0; src[n] != NULL; n++)
124	continue;
125    sdst = dst = xmalloc((n + 1) * sizeof(char *));
126
127    for (; *src != NULL; src++)
128	*dst++ = strsave(short2str(*src));
129    *dst = NULL;
130    return (sdst);
131}
132
133Char   *
134str2short(const char *src)
135{
136    static struct Strbuf buf; /* = Strbuf_INIT; */
137
138    if (src == NULL)
139	return (NULL);
140
141    buf.len = 0;
142    while (*src) {
143	Char wc;
144
145	src += one_mbtowc(&wc, src, MB_LEN_MAX);
146	Strbuf_append1(&buf, wc);
147    }
148    Strbuf_terminate(&buf);
149    return buf.s;
150}
151
152char   *
153short2str(const Char *src)
154{
155    static char *sdst = NULL;
156    static size_t dstsize = 0;
157    char *dst, *edst;
158
159    if (src == NULL)
160	return (NULL);
161
162    if (sdst == NULL) {
163	dstsize = MALLOC_INCR;
164	sdst = xmalloc((dstsize + MALLOC_SURPLUS) * sizeof(char));
165    }
166    dst = sdst;
167    edst = &dst[dstsize];
168    while (*src) {
169	dst += one_wctomb(dst, *src & CHAR);
170	src++;
171	if (dst >= edst) {
172	    dstsize += MALLOC_INCR;
173	    sdst = xrealloc(sdst, (dstsize + MALLOC_SURPLUS) * sizeof(char));
174	    edst = &sdst[dstsize];
175	    dst = &edst[-MALLOC_INCR];
176	}
177    }
178    *dst = 0;
179    return (sdst);
180}
181
182#ifndef WIDE_STRINGS
183Char   *
184s_strcpy(Char *dst, const Char *src)
185{
186    Char *sdst;
187
188    sdst = dst;
189    while ((*dst++ = *src++) != '\0')
190	continue;
191    return (sdst);
192}
193
194Char   *
195s_strncpy(Char *dst, const Char *src, size_t n)
196{
197    Char *sdst;
198
199    if (n == 0)
200	return(dst);
201
202    sdst = dst;
203    do
204	if ((*dst++ = *src++) == '\0') {
205	    while (--n != 0)
206		*dst++ = '\0';
207	    return(sdst);
208	}
209    while (--n != 0);
210    return (sdst);
211}
212
213Char   *
214s_strcat(Char *dst, const Char *src)
215{
216    Strcpy(Strend(dst), src);
217    return dst;
218}
219
220#ifdef NOTUSED
221Char   *
222s_strncat(Char *dst, const Char *src, size_t n)
223{
224    Char *sdst;
225
226    if (n == 0)
227	return (dst);
228
229    sdst = dst;
230
231    while (*dst)
232	dst++;
233
234    do
235	if ((*dst++ = *src++) == '\0')
236	    return(sdst);
237    while (--n != 0)
238	continue;
239
240    *dst = '\0';
241    return (sdst);
242}
243
244#endif
245
246Char   *
247s_strchr(const Char *str, int ch)
248{
249    do
250	if (*str == ch)
251	    return ((Char *)(intptr_t)str);
252    while (*str++);
253    return (NULL);
254}
255
256Char   *
257s_strrchr(const Char *str, int ch)
258{
259    const Char *rstr;
260
261    rstr = NULL;
262    do
263	if (*str == ch)
264	    rstr = str;
265    while (*str++);
266    return ((Char *)(intptr_t)rstr);
267}
268
269size_t
270s_strlen(const Char *str)
271{
272    size_t n;
273
274    for (n = 0; *str++; n++)
275	continue;
276    return (n);
277}
278
279int
280s_strcmp(const Char *str1, const Char *str2)
281{
282    for (; *str1 && *str1 == *str2; str1++, str2++)
283	continue;
284    /*
285     * The following case analysis is necessary so that characters which look
286     * negative collate low against normal characters but high against the
287     * end-of-string NUL.
288     */
289    if (*str1 == '\0' && *str2 == '\0')
290	return (0);
291    else if (*str1 == '\0')
292	return (-1);
293    else if (*str2 == '\0')
294	return (1);
295    else
296	return (*str1 - *str2);
297}
298
299int
300s_strncmp(const Char *str1, const Char *str2, size_t n)
301{
302    if (n == 0)
303	return (0);
304    do {
305	if (*str1 != *str2) {
306	    /*
307	     * The following case analysis is necessary so that characters
308	     * which look negative collate low against normal characters
309	     * but high against the end-of-string NUL.
310	     */
311	    if (*str1 == '\0')
312		return (-1);
313	    else if (*str2 == '\0')
314		return (1);
315	    else
316		return (*str1 - *str2);
317	}
318        if (*str1 == '\0')
319	    return(0);
320	str1++, str2++;
321    } while (--n != 0);
322    return(0);
323}
324#endif /* not WIDE_STRINGS */
325
326int
327s_strcasecmp(const Char *str1, const Char *str2)
328{
329#ifdef WIDE_STRINGS
330    wchar_t l1 = 0, l2 = 0;
331    for (; *str1 && ((*str1 == *str2 && (l1 = l2 = 0) == 0) ||
332	(l1 = towlower(*str1)) == (l2 = towlower(*str2))); str1++, str2++)
333	continue;
334
335#else
336    unsigned char c1, c2, l1 = 0, l2 = 0;
337    for (; *str1 && ((*str1 == *str2 && (l1 = l2 = 0) == 0) ||
338	((c1 = (unsigned char)*str1) == *str1 &&
339	 (c2 = (unsigned char)*str2) == *str2 &&
340	(l1 = tolower(c1)) == (l2 = tolower(c2)))); str1++, str2++)
341	continue;
342#endif
343    /*
344     * The following case analysis is necessary so that characters which look
345     * negative collate low against normal characters but high against the
346     * end-of-string NUL.
347     */
348    if (*str1 == '\0' && *str2 == '\0')
349	return (0);
350    else if (*str1 == '\0')
351	return (-1);
352    else if (*str2 == '\0')
353	return (1);
354    else if (l1 == l2)	/* They are zero when they are equal */
355	return (*str1 - *str2);
356    else
357	return (l1 - l2);
358}
359
360Char   *
361s_strnsave(const Char *s, size_t len)
362{
363    Char *n;
364
365    n = xmalloc((len + 1) * sizeof (*n));
366    memcpy(n, s, len * sizeof (*n));
367    n[len] = '\0';
368    return n;
369}
370
371Char   *
372s_strsave(const Char *s)
373{
374    Char   *n;
375    size_t size;
376
377    if (s == NULL)
378	s = STRNULL;
379    size = (Strlen(s) + 1) * sizeof(*n);
380    n = xmalloc(size);
381    memcpy(n, s, size);
382    return (n);
383}
384
385Char   *
386s_strspl(const Char *cp, const Char *dp)
387{
388    Char *res, *ep;
389    const Char *p, *q;
390
391    if (!cp)
392	cp = STRNULL;
393    if (!dp)
394	dp = STRNULL;
395    for (p = cp; *p++;)
396	continue;
397    for (q = dp; *q++;)
398	continue;
399    res = xmalloc(((p - cp) + (q - dp) - 1) * sizeof(Char));
400    for (ep = res, q = cp; (*ep++ = *q++) != '\0';)
401	continue;
402    for (ep--, q = dp; (*ep++ = *q++) != '\0';)
403	continue;
404    return (res);
405}
406
407Char   *
408s_strend(const Char *cp)
409{
410    if (!cp)
411	return ((Char *)(intptr_t) cp);
412    while (*cp)
413	cp++;
414    return ((Char *)(intptr_t) cp);
415}
416
417Char   *
418s_strstr(const Char *s, const Char *t)
419{
420    do {
421	const Char *ss = s;
422	const Char *tt = t;
423
424	do
425	    if (*tt == '\0')
426		return ((Char *)(intptr_t) s);
427	while (*ss++ == *tt++);
428    } while (*s++ != '\0');
429    return (NULL);
430}
431
432#else /* !SHORT_STRINGS */
433char *
434caching_strip(const char *s)
435{
436    static char *buf = NULL;
437    static size_t buf_size = 0;
438    size_t size;
439
440    if (s == NULL)
441      return NULL;
442    size = strlen(s) + 1;
443    if (buf_size < size) {
444	buf = xrealloc(buf, size);
445	buf_size = size;
446    }
447    memcpy(buf, s, size);
448    strip(buf);
449    return buf;
450}
451#endif
452
453char   *
454short2qstr(const Char *src)
455{
456    static char *sdst = NULL;
457    static size_t dstsize = 0;
458    char *dst, *edst;
459
460    if (src == NULL)
461	return (NULL);
462
463    if (sdst == NULL) {
464	dstsize = MALLOC_INCR;
465	sdst = xmalloc((dstsize + MALLOC_SURPLUS) * sizeof(char));
466    }
467    dst = sdst;
468    edst = &dst[dstsize];
469    while (*src) {
470	if (*src & QUOTE) {
471	    *dst++ = '\\';
472	    if (dst == edst) {
473		dstsize += MALLOC_INCR;
474		sdst = xrealloc(sdst,
475				(dstsize + MALLOC_SURPLUS) * sizeof(char));
476		edst = &sdst[dstsize];
477		dst = &edst[-MALLOC_INCR];
478	    }
479	}
480	dst += one_wctomb(dst, *src & CHAR);
481	src++;
482	if (dst >= edst) {
483	    dstsize += MALLOC_INCR;
484	    sdst = xrealloc(sdst, (dstsize + MALLOC_SURPLUS) * sizeof(char));
485	    edst = &sdst[dstsize];
486	    dst = &edst[-MALLOC_INCR];
487	}
488    }
489    *dst = 0;
490    return (sdst);
491}
492
493static void
494bb_store(struct blk_buf *bb, Char *str)
495{
496    if (bb->len == bb->size) { /* Keep space for terminating NULL */
497	if (bb->size == 0)
498	    bb->size = 16; /* Arbitrary */
499	else
500	    bb->size *= 2;
501	bb->vec = xrealloc(bb->vec, bb->size * sizeof (*bb->vec));
502    }
503    bb->vec[bb->len] = str;
504}
505
506void
507bb_append(struct blk_buf *bb, Char *str)
508{
509    bb_store(bb, str);
510    bb->len++;
511}
512
513void
514bb_cleanup(void *xbb)
515{
516    struct blk_buf *bb;
517    size_t i;
518
519    bb = xbb;
520    for (i = 0; i < bb->len; i++)
521	xfree(bb->vec[i]);
522    xfree(bb->vec);
523}
524
525Char **
526bb_finish(struct blk_buf *bb)
527{
528    bb_store(bb, NULL);
529    return xrealloc(bb->vec, (bb->len + 1) * sizeof (*bb->vec));
530}
531
532#define DO_STRBUF(STRBUF, CHAR, STRLEN)				\
533static void							\
534STRBUF##_store1(struct STRBUF *buf, CHAR c)			\
535{								\
536    if (buf->size == buf->len) {				\
537	if (buf->size == 0)					\
538	    buf->size = 64; /* Arbitrary */			\
539	else							\
540	    buf->size *= 2;					\
541	buf->s = xrealloc(buf->s, buf->size * sizeof(*buf->s));	\
542    }								\
543    buf->s[buf->len] = c;					\
544}								\
545								\
546/* Like strbuf_append1(buf, '\0'), but don't advance len */	\
547void								\
548STRBUF##_terminate(struct STRBUF *buf)				\
549{								\
550    STRBUF##_store1(buf, '\0');					\
551}								\
552								\
553void								\
554STRBUF##_append1(struct STRBUF *buf, CHAR c)			\
555{								\
556    STRBUF##_store1(buf, c);					\
557    buf->len++;							\
558}								\
559								\
560void								\
561STRBUF##_appendn(struct STRBUF *buf, const CHAR *s, size_t len)	\
562{								\
563    if (buf->size < buf->len + len) {				\
564	if (buf->size == 0)					\
565	    buf->size = 64; /* Arbitrary */			\
566	while (buf->size < buf->len + len)			\
567	    buf->size *= 2;					\
568	buf->s = xrealloc(buf->s, buf->size * sizeof(*buf->s));	\
569    }								\
570    memcpy(buf->s + buf->len, s, len * sizeof(*buf->s));	\
571    buf->len += len;						\
572}								\
573								\
574void								\
575STRBUF##_append(struct STRBUF *buf, const CHAR *s)		\
576{								\
577    STRBUF##_appendn(buf, s, STRLEN(s));			\
578}								\
579								\
580CHAR *								\
581STRBUF##_finish(struct STRBUF *buf)				\
582{								\
583    STRBUF##_append1(buf, 0);					\
584    return xrealloc(buf->s, buf->len * sizeof(*buf->s));	\
585}								\
586								\
587void								\
588STRBUF##_cleanup(void *xbuf)					\
589{								\
590    struct STRBUF *buf;						\
591								\
592    buf = xbuf;							\
593    xfree(buf->s);						\
594}								\
595								\
596const struct STRBUF STRBUF##_init /* = STRBUF##_INIT; */
597
598DO_STRBUF(strbuf, char, strlen);
599DO_STRBUF(Strbuf, Char, Strlen);
600