xref: /freebsd/contrib/bmake/str.c (revision e9e8876a4d6afc1ad5315faaa191b25121a813d7)
1 /*	$NetBSD: str.c,v 1.86 2021/06/21 16:59:18 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 1.86 2021/06/21 16:59:18 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 /*
103  * Fracture a string into an array of words (as delineated by tabs or spaces)
104  * taking quotation marks into account.
105  *
106  * If expand is true, quotes are removed and escape sequences such as \r, \t,
107  * etc... are expanded. In this case, return NULL on parse errors.
108  *
109  * Returns the fractured words, which must be freed later using Words_Free,
110  * unless the returned Words.words was NULL.
111  */
112 SubstringWords
113 Substring_Words(const char *str, bool expand)
114 {
115 	size_t str_len;
116 	char *words_buf;
117 	size_t words_cap;
118 	Substring *words;
119 	size_t words_len;
120 	char inquote;
121 	char *word_start;
122 	char *word_end;
123 	const char *str_p;
124 
125 	/* XXX: why only hspace, not whitespace? */
126 	cpp_skip_hspace(&str);	/* skip leading space chars. */
127 
128 	/* words_buf holds the words, separated by '\0'. */
129 	str_len = strlen(str);
130 	words_buf = bmake_malloc(str_len + 1);
131 
132 	words_cap = str_len / 5 > 50 ? str_len / 5 : 50;
133 	words = bmake_malloc((words_cap + 1) * sizeof(words[0]));
134 
135 	/*
136 	 * copy the string; at the same time, parse backslashes,
137 	 * quotes and build the word list.
138 	 */
139 	words_len = 0;
140 	inquote = '\0';
141 	word_start = words_buf;
142 	word_end = words_buf;
143 	for (str_p = str;; str_p++) {
144 		char ch = *str_p;
145 		switch (ch) {
146 		case '"':
147 		case '\'':
148 			if (inquote != '\0') {
149 				if (inquote == ch)
150 					inquote = '\0';
151 				else
152 					break;
153 			} else {
154 				inquote = ch;
155 				/* Don't miss "" or '' */
156 				if (word_start == NULL && str_p[1] == inquote) {
157 					if (!expand) {
158 						word_start = word_end;
159 						*word_end++ = ch;
160 					} else
161 						word_start = word_end + 1;
162 					str_p++;
163 					inquote = '\0';
164 					break;
165 				}
166 			}
167 			if (!expand) {
168 				if (word_start == NULL)
169 					word_start = word_end;
170 				*word_end++ = ch;
171 			}
172 			continue;
173 		case ' ':
174 		case '\t':
175 		case '\n':
176 			if (inquote != '\0')
177 				break;
178 			if (word_start == NULL)
179 				continue;
180 			/* FALLTHROUGH */
181 		case '\0':
182 			/*
183 			 * end of a token -- make sure there's enough words
184 			 * space and save off a pointer.
185 			 */
186 			if (word_start == NULL)
187 				goto done;
188 
189 			*word_end++ = '\0';
190 			if (words_len == words_cap) {
191 				words_cap *= 2;
192 				words = bmake_realloc(words,
193 				    (words_cap + 1) * sizeof(words[0]));
194 			}
195 			words[words_len++] =
196 			    Substring_Init(word_start, word_end - 1);
197 			word_start = NULL;
198 			if (ch == '\n' || ch == '\0') {
199 				if (expand && inquote != '\0') {
200 					SubstringWords res;
201 
202 					free(words);
203 					free(words_buf);
204 
205 					res.words = NULL;
206 					res.len = 0;
207 					res.freeIt = NULL;
208 					return res;
209 				}
210 				goto done;
211 			}
212 			continue;
213 		case '\\':
214 			if (!expand) {
215 				if (word_start == NULL)
216 					word_start = word_end;
217 				*word_end++ = '\\';
218 				/* catch '\' at end of line */
219 				if (str_p[1] == '\0')
220 					continue;
221 				ch = *++str_p;
222 				break;
223 			}
224 
225 			switch (ch = *++str_p) {
226 			case '\0':
227 			case '\n':
228 				/* hmmm; fix it up as best we can */
229 				ch = '\\';
230 				str_p--;
231 				break;
232 			case 'b':
233 				ch = '\b';
234 				break;
235 			case 'f':
236 				ch = '\f';
237 				break;
238 			case 'n':
239 				ch = '\n';
240 				break;
241 			case 'r':
242 				ch = '\r';
243 				break;
244 			case 't':
245 				ch = '\t';
246 				break;
247 			}
248 			break;
249 		}
250 		if (word_start == NULL)
251 			word_start = word_end;
252 		*word_end++ = ch;
253 	}
254 done:
255 	words[words_len] = Substring_Init(NULL, NULL);	/* useful for argv */
256 
257 	{
258 		SubstringWords result;
259 
260 		result.words = words;
261 		result.len = words_len;
262 		result.freeIt = words_buf;
263 		return result;
264 	}
265 }
266 
267 Words
268 Str_Words(const char *str, bool expand)
269 {
270 	SubstringWords swords;
271 	Words words;
272 	size_t i;
273 
274 	swords = Substring_Words(str, expand);
275 	if (swords.words == NULL) {
276 		words.words = NULL;
277 		words.len = 0;
278 		words.freeIt = NULL;
279 		return words;
280 	}
281 
282 	words.words = bmake_malloc((swords.len + 1) * sizeof(words.words[0]));
283 	words.len = swords.len;
284 	words.freeIt = swords.freeIt;
285 	for (i = 0; i < swords.len + 1; i++)
286 		words.words[i] = UNCONST(swords.words[i].start);
287 	free(swords.words);
288 	return words;
289 }
290 
291 /*
292  * Str_Match -- Test if a string matches a pattern like "*.[ch]".
293  * The following special characters are known *?\[] (as in fnmatch(3)).
294  *
295  * XXX: this function does not detect or report malformed patterns.
296  */
297 bool
298 Str_Match(const char *str, const char *pat)
299 {
300 	for (;;) {
301 		/*
302 		 * See if we're at the end of both the pattern and the
303 		 * string. If so, we succeeded.  If we're at the end of the
304 		 * pattern but not at the end of the string, we failed.
305 		 */
306 		if (*pat == '\0')
307 			return *str == '\0';
308 		if (*str == '\0' && *pat != '*')
309 			return false;
310 
311 		/*
312 		 * A '*' in the pattern matches any substring.  We handle this
313 		 * by calling ourselves for each suffix of the string.
314 		 */
315 		if (*pat == '*') {
316 			pat++;
317 			while (*pat == '*')
318 				pat++;
319 			if (*pat == '\0')
320 				return true;
321 			while (*str != '\0') {
322 				if (Str_Match(str, pat))
323 					return true;
324 				str++;
325 			}
326 			return false;
327 		}
328 
329 		/* A '?' in the pattern matches any single character. */
330 		if (*pat == '?')
331 			goto thisCharOK;
332 
333 		/*
334 		 * A '[' in the pattern matches a character from a list.
335 		 * The '[' is followed by the list of acceptable characters,
336 		 * or by ranges (two characters separated by '-'). In these
337 		 * character lists, the backslash is an ordinary character.
338 		 */
339 		if (*pat == '[') {
340 			bool neg = pat[1] == '^';
341 			pat += neg ? 2 : 1;
342 
343 			for (;;) {
344 				if (*pat == ']' || *pat == '\0') {
345 					if (neg)
346 						break;
347 					return false;
348 				}
349 				/*
350 				 * XXX: This naive comparison makes the
351 				 * control flow of the pattern parser
352 				 * dependent on the actual value of the
353 				 * string.  This is unpredictable.  It may be
354 				 * though that the code only looks wrong but
355 				 * actually all code paths result in the same
356 				 * behavior.  This needs further tests.
357 				 */
358 				if (*pat == *str)
359 					break;
360 				if (pat[1] == '-') {
361 					if (pat[2] == '\0')
362 						return neg;
363 					if (*pat <= *str && pat[2] >= *str)
364 						break;
365 					if (*pat >= *str && pat[2] <= *str)
366 						break;
367 					pat += 2;
368 				}
369 				pat++;
370 			}
371 			if (neg && *pat != ']' && *pat != '\0')
372 				return false;
373 			while (*pat != ']' && *pat != '\0')
374 				pat++;
375 			if (*pat == '\0')
376 				pat--;
377 			goto thisCharOK;
378 		}
379 
380 		/*
381 		 * A backslash in the pattern matches the character following
382 		 * it exactly.
383 		 */
384 		if (*pat == '\\') {
385 			pat++;
386 			if (*pat == '\0')
387 				return false;
388 		}
389 
390 		if (*pat != *str)
391 			return false;
392 
393 	thisCharOK:
394 		pat++;
395 		str++;
396 	}
397 }
398