xref: /freebsd/contrib/bmake/str.c (revision a7623790fb345e6dc986dfd31df0ace115e6f2e4)
1 /*	$NetBSD: str.c,v 1.74 2020/11/16 18:28:27 rillig Exp $	*/
2 
3 /*-
4  * Copyright (c) 1988, 1989, 1990, 1993
5  *	The Regents of the University of California.  All rights reserved.
6  *
7  * This code is derived from software contributed to Berkeley by
8  * Adam de Boor.
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 
35 /*-
36  * Copyright (c) 1989 by Berkeley Softworks
37  * All rights reserved.
38  *
39  * This code is derived from software contributed to Berkeley by
40  * Adam de Boor.
41  *
42  * Redistribution and use in source and binary forms, with or without
43  * modification, are permitted provided that the following conditions
44  * are met:
45  * 1. Redistributions of source code must retain the above copyright
46  *    notice, this list of conditions and the following disclaimer.
47  * 2. Redistributions in binary form must reproduce the above copyright
48  *    notice, this list of conditions and the following disclaimer in the
49  *    documentation and/or other materials provided with the distribution.
50  * 3. All advertising materials mentioning features or use of this software
51  *    must display the following acknowledgement:
52  *	This product includes software developed by the University of
53  *	California, Berkeley and its contributors.
54  * 4. Neither the name of the University nor the names of its contributors
55  *    may be used to endorse or promote products derived from this software
56  *    without specific prior written permission.
57  *
58  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
59  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
60  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
61  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
62  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
63  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
64  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
65  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
66  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
67  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
68  * SUCH DAMAGE.
69  */
70 
71 #include "make.h"
72 
73 /*	"@(#)str.c	5.8 (Berkeley) 6/1/90"	*/
74 MAKE_RCSID("$NetBSD: str.c,v 1.74 2020/11/16 18:28:27 rillig Exp $");
75 
76 /* Return the concatenation of s1 and s2, freshly allocated. */
77 char *
78 str_concat2(const char *s1, const char *s2)
79 {
80 	size_t len1 = strlen(s1);
81 	size_t len2 = strlen(s2);
82 	char *result = bmake_malloc(len1 + len2 + 1);
83 	memcpy(result, s1, len1);
84 	memcpy(result + len1, s2, len2 + 1);
85 	return result;
86 }
87 
88 /* Return the concatenation of s1, s2 and s3, freshly allocated. */
89 char *
90 str_concat3(const char *s1, const char *s2, const char *s3)
91 {
92 	size_t len1 = strlen(s1);
93 	size_t len2 = strlen(s2);
94 	size_t len3 = strlen(s3);
95 	char *result = bmake_malloc(len1 + len2 + len3 + 1);
96 	memcpy(result, s1, len1);
97 	memcpy(result + len1, s2, len2);
98 	memcpy(result + len1 + len2, s3, len3 + 1);
99 	return result;
100 }
101 
102 /* Return the concatenation of s1, s2, s3 and s4, freshly allocated. */
103 char *
104 str_concat4(const char *s1, const char *s2, const char *s3, const char *s4)
105 {
106 	size_t len1 = strlen(s1);
107 	size_t len2 = strlen(s2);
108 	size_t len3 = strlen(s3);
109 	size_t len4 = strlen(s4);
110 	char *result = bmake_malloc(len1 + len2 + len3 + len4 + 1);
111 	memcpy(result, s1, len1);
112 	memcpy(result + len1, s2, len2);
113 	memcpy(result + len1 + len2, s3, len3);
114 	memcpy(result + len1 + len2 + len3, s4, len4 + 1);
115 	return result;
116 }
117 
118 /* Fracture a string into an array of words (as delineated by tabs or spaces)
119  * taking quotation marks into account.
120  *
121  * If expand is TRUE, quotes are removed and escape sequences such as \r, \t,
122  * etc... are expanded. In this case, return NULL on parse errors.
123  *
124  * Returns the fractured words, which must be freed later using Words_Free,
125  * unless the returned Words.words was NULL.
126  */
127 Words
128 Str_Words(const char *str, Boolean expand)
129 {
130 	size_t str_len;
131 	char *words_buf;
132 	size_t words_cap;
133 	char **words;
134 	size_t words_len;
135 	char inquote;
136 	char *word_start;
137 	char *word_end;
138 	const char *str_p;
139 
140 	/* XXX: why only hspace, not whitespace? */
141 	cpp_skip_hspace(&str);	/* skip leading space chars. */
142 
143 	/* words_buf holds the words, separated by '\0'. */
144 	str_len = strlen(str);
145 	words_buf = bmake_malloc(strlen(str) + 1);
146 
147 	words_cap = str_len / 5 > 50 ? str_len / 5 : 50;
148 	words = bmake_malloc((words_cap + 1) * sizeof(char *));
149 
150 	/*
151 	 * copy the string; at the same time, parse backslashes,
152 	 * quotes and build the word list.
153 	 */
154 	words_len = 0;
155 	inquote = '\0';
156 	word_start = words_buf;
157 	word_end = words_buf;
158 	for (str_p = str;; ++str_p) {
159 		char ch = *str_p;
160 		switch (ch) {
161 		case '"':
162 		case '\'':
163 			if (inquote) {
164 				if (inquote == ch)
165 					inquote = '\0';
166 				else
167 					break;
168 			} else {
169 				inquote = ch;
170 				/* Don't miss "" or '' */
171 				if (word_start == NULL && str_p[1] == inquote) {
172 					if (!expand) {
173 						word_start = word_end;
174 						*word_end++ = ch;
175 					} else
176 						word_start = word_end + 1;
177 					str_p++;
178 					inquote = '\0';
179 					break;
180 				}
181 			}
182 			if (!expand) {
183 				if (word_start == NULL)
184 					word_start = word_end;
185 				*word_end++ = ch;
186 			}
187 			continue;
188 		case ' ':
189 		case '\t':
190 		case '\n':
191 			if (inquote)
192 				break;
193 			if (word_start == NULL)
194 				continue;
195 			/* FALLTHROUGH */
196 		case '\0':
197 			/*
198 			 * end of a token -- make sure there's enough words
199 			 * space and save off a pointer.
200 			 */
201 			if (word_start == NULL)
202 				goto done;
203 
204 			*word_end++ = '\0';
205 			if (words_len == words_cap) {
206 				size_t new_size;
207 				words_cap *= 2;		/* ramp up fast */
208 				new_size = (words_cap + 1) * sizeof(char *);
209 				words = bmake_realloc(words, new_size);
210 			}
211 			words[words_len++] = word_start;
212 			word_start = NULL;
213 			if (ch == '\n' || ch == '\0') {
214 				if (expand && inquote) {
215 					free(words);
216 					free(words_buf);
217 					return (Words){ NULL, 0, NULL };
218 				}
219 				goto done;
220 			}
221 			continue;
222 		case '\\':
223 			if (!expand) {
224 				if (word_start == NULL)
225 					word_start = word_end;
226 				*word_end++ = '\\';
227 				/* catch '\' at end of line */
228 				if (str_p[1] == '\0')
229 					continue;
230 				ch = *++str_p;
231 				break;
232 			}
233 
234 			switch (ch = *++str_p) {
235 			case '\0':
236 			case '\n':
237 				/* hmmm; fix it up as best we can */
238 				ch = '\\';
239 				str_p--;
240 				break;
241 			case 'b':
242 				ch = '\b';
243 				break;
244 			case 'f':
245 				ch = '\f';
246 				break;
247 			case 'n':
248 				ch = '\n';
249 				break;
250 			case 'r':
251 				ch = '\r';
252 				break;
253 			case 't':
254 				ch = '\t';
255 				break;
256 			}
257 			break;
258 		}
259 		if (word_start == NULL)
260 			word_start = word_end;
261 		*word_end++ = ch;
262 	}
263 done:
264 	words[words_len] = NULL;	/* useful for argv */
265 	return (Words){ words, words_len, words_buf };
266 }
267 
268 /*
269  * Str_Match -- Test if a string matches a pattern like "*.[ch]".
270  * The following special characters are known *?\[] (as in fnmatch(3)).
271  *
272  * XXX: this function does not detect or report malformed patterns.
273  */
274 Boolean
275 Str_Match(const char *str, const char *pat)
276 {
277 	for (;;) {
278 		/*
279 		 * See if we're at the end of both the pattern and the
280 		 * string. If so, we succeeded.  If we're at the end of the
281 		 * pattern but not at the end of the string, we failed.
282 		 */
283 		if (*pat == '\0')
284 			return *str == '\0';
285 		if (*str == '\0' && *pat != '*')
286 			return FALSE;
287 
288 		/*
289 		 * A '*' in the pattern matches any substring.  We handle this
290 		 * by calling ourselves for each suffix of the string.
291 		 */
292 		if (*pat == '*') {
293 			pat++;
294 			while (*pat == '*')
295 				pat++;
296 			if (*pat == '\0')
297 				return TRUE;
298 			while (*str != '\0') {
299 				if (Str_Match(str, pat))
300 					return TRUE;
301 				str++;
302 			}
303 			return FALSE;
304 		}
305 
306 		/* A '?' in the pattern matches any single character. */
307 		if (*pat == '?')
308 			goto thisCharOK;
309 
310 		/*
311 		 * A '[' in the pattern matches a character from a list.
312 		 * The '[' is followed by the list of acceptable characters,
313 		 * or by ranges (two characters separated by '-'). In these
314 		 * character lists, the backslash is an ordinary character.
315 		 */
316 		if (*pat == '[') {
317 			Boolean neg = pat[1] == '^';
318 			pat += neg ? 2 : 1;
319 
320 			for (;;) {
321 				if (*pat == ']' || *pat == '\0') {
322 					if (neg)
323 						break;
324 					return FALSE;
325 				}
326 				/* XXX: This naive comparison makes the parser
327 				 * for the pattern dependent on the actual of
328 				 * the string.  This is unpredictable. */
329 				if (*pat == *str)
330 					break;
331 				if (pat[1] == '-') {
332 					if (pat[2] == '\0')
333 						return neg;
334 					if (*pat <= *str && pat[2] >= *str)
335 						break;
336 					if (*pat >= *str && pat[2] <= *str)
337 						break;
338 					pat += 2;
339 				}
340 				pat++;
341 			}
342 			if (neg && *pat != ']' && *pat != '\0')
343 				return FALSE;
344 			while (*pat != ']' && *pat != '\0')
345 				pat++;
346 			if (*pat == '\0')
347 				pat--;
348 			goto thisCharOK;
349 		}
350 
351 		/*
352 		 * A backslash in the pattern matches the character following
353 		 * it exactly.
354 		 */
355 		if (*pat == '\\') {
356 			pat++;
357 			if (*pat == '\0')
358 				return FALSE;
359 		}
360 
361 		if (*pat != *str)
362 			return FALSE;
363 
364 	thisCharOK:
365 		pat++;
366 		str++;
367 	}
368 }
369