xref: /netbsd-src/usr.bin/make/str.c (revision 1f2744e6e4915c9da2a3f980279398c4cf7d5e6d)
1 /*-
2  * Copyright (c) 1988, 1989, 1990 The Regents of the University of California.
3  * Copyright (c) 1988, 1989 by Adam de Boor
4  * Copyright (c) 1989 by Berkeley Softworks
5  * 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. All advertising materials mentioning features or use of this software
19  *    must display the following acknowledgement:
20  *	This product includes software developed by the University of
21  *	California, Berkeley and its contributors.
22  * 4. Neither the name of the University nor the names of its contributors
23  *    may be used to endorse or promote products derived from this software
24  *    without specific prior written permission.
25  *
26  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
27  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
28  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
29  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
30  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
31  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
32  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
33  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
34  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
35  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
36  * SUCH DAMAGE.
37  */
38 
39 #ifndef lint
40 /* from: static char     sccsid[] = "@(#)str.c	5.8 (Berkeley) 6/1/90"; */
41 static char *rcsid = "$Id: str.c,v 1.9 1995/02/04 23:44:41 christos Exp $";
42 #endif				/* not lint */
43 
44 #include "make.h"
45 
46 static char **argv, *buffer;
47 static int argmax, curlen;
48 
49 /*
50  * str_init --
51  *	Initialize the strings package
52  *
53  */
54 void
55 str_init()
56 {
57     char *p1;
58     argv = (char **)emalloc((argmax = 50) * sizeof(char *));
59     argv[0] = Var_Value(".MAKE", VAR_GLOBAL, &p1);
60 }
61 
62 
63 /*
64  * str_end --
65  *	Cleanup the strings package
66  *
67  */
68 void
69 str_end()
70 {
71     if (argv[0]) {
72 	free(argv[0]);
73 	free((Address) argv);
74     }
75     if (buffer)
76 	free(buffer);
77 }
78 
79 /*-
80  * str_concat --
81  *	concatenate the two strings, inserting a space or slash between them,
82  *	freeing them if requested.
83  *
84  * returns --
85  *	the resulting string in allocated space.
86  */
87 char *
88 str_concat(s1, s2, flags)
89 	char *s1, *s2;
90 	int flags;
91 {
92 	register int len1, len2;
93 	register char *result;
94 
95 	/* get the length of both strings */
96 	len1 = strlen(s1);
97 	len2 = strlen(s2);
98 
99 	/* allocate length plus separator plus EOS */
100 	result = emalloc((u_int)(len1 + len2 + 2));
101 
102 	/* copy first string into place */
103 	memcpy(result, s1, len1);
104 
105 	/* add separator character */
106 	if (flags & STR_ADDSPACE) {
107 		result[len1] = ' ';
108 		++len1;
109 	} else if (flags & STR_ADDSLASH) {
110 		result[len1] = '/';
111 		++len1;
112 	}
113 
114 	/* copy second string plus EOS into place */
115 	memcpy(result + len1, s2, len2 + 1);
116 
117 	/* free original strings */
118 	if (flags & STR_DOFREE) {
119 		(void)free(s1);
120 		(void)free(s2);
121 	}
122 	return(result);
123 }
124 
125 /*-
126  * brk_string --
127  *	Fracture a string into an array of words (as delineated by tabs or
128  *	spaces) taking quotation marks into account.  Leading tabs/spaces
129  *	are ignored.
130  *
131  * returns --
132  *	Pointer to the array of pointers to the words.  To make life easier,
133  *	the first word is always the value of the .MAKE variable.
134  */
135 char **
136 brk_string(str, store_argc, expand)
137 	register char *str;
138 	int *store_argc;
139 	Boolean expand;
140 {
141 	register int argc, ch;
142 	register char inquote, *p, *start, *t;
143 	int len;
144 
145 	/* skip leading space chars. */
146 	for (; *str == ' ' || *str == '\t'; ++str)
147 		continue;
148 
149 	/* allocate room for a copy of the string */
150 	if ((len = strlen(str) + 1) > curlen) {
151 		if (buffer)
152 		    free(buffer);
153 		buffer = emalloc(curlen = len);
154 	}
155 
156 	/*
157 	 * copy the string; at the same time, parse backslashes,
158 	 * quotes and build the argument list.
159 	 */
160 	argc = 1;
161 	inquote = '\0';
162 	for (p = str, start = t = buffer;; ++p) {
163 		switch(ch = *p) {
164 		case '"':
165 		case '\'':
166 			if (inquote)
167 				if (inquote == ch)
168 					inquote = '\0';
169 				else
170 					break;
171 			else {
172 				inquote = (char) ch;
173 				/* Don't miss "" or '' */
174 				if (start == NULL && p[1] == inquote) {
175 					start = t + 1;
176 					break;
177 				}
178 			}
179 			if (!expand) {
180 				if (!start)
181 					start = t;
182 				*t++ = ch;
183 			}
184 			continue;
185 		case ' ':
186 		case '\t':
187 		case '\n':
188 			if (inquote)
189 				break;
190 			if (!start)
191 				continue;
192 			/* FALLTHROUGH */
193 		case '\0':
194 			/*
195 			 * end of a token -- make sure there's enough argv
196 			 * space and save off a pointer.
197 			 */
198 			if (!start)
199 			    goto done;
200 
201 			*t++ = '\0';
202 			if (argc == argmax) {
203 				argmax *= 2;		/* ramp up fast */
204 				if (!(argv = (char **)realloc(argv,
205 				    argmax * sizeof(char *))))
206 				enomem();
207 			}
208 			argv[argc++] = start;
209 			start = (char *)NULL;
210 			if (ch == '\n' || ch == '\0')
211 				goto done;
212 			continue;
213 		case '\\':
214 			if (!expand) {
215 				if (!start)
216 					start = t;
217 				*t++ = '\\';
218 				ch = *++p;
219 				break;
220 			}
221 
222 			switch (ch = *++p) {
223 			case '\0':
224 			case '\n':
225 				/* hmmm; fix it up as best we can */
226 				ch = '\\';
227 				--p;
228 				break;
229 			case 'b':
230 				ch = '\b';
231 				break;
232 			case 'f':
233 				ch = '\f';
234 				break;
235 			case 'n':
236 				ch = '\n';
237 				break;
238 			case 'r':
239 				ch = '\r';
240 				break;
241 			case 't':
242 				ch = '\t';
243 				break;
244 			}
245 			break;
246 		}
247 		if (!start)
248 			start = t;
249 		*t++ = (char) ch;
250 	}
251 done:	argv[argc] = (char *)NULL;
252 	*store_argc = argc;
253 	return(argv);
254 }
255 
256 /*
257  * Str_FindSubstring -- See if a string contains a particular substring.
258  *
259  * Results: If string contains substring, the return value is the location of
260  * the first matching instance of substring in string.  If string doesn't
261  * contain substring, the return value is NULL.  Matching is done on an exact
262  * character-for-character basis with no wildcards or special characters.
263  *
264  * Side effects: None.
265  */
266 char *
267 Str_FindSubstring(string, substring)
268 	register char *string;		/* String to search. */
269 	char *substring;		/* Substring to find in string */
270 {
271 	register char *a, *b;
272 
273 	/*
274 	 * First scan quickly through the two strings looking for a single-
275 	 * character match.  When it's found, then compare the rest of the
276 	 * substring.
277 	 */
278 
279 	for (b = substring; *string != 0; string += 1) {
280 		if (*string != *b)
281 			continue;
282 		a = string;
283 		for (;;) {
284 			if (*b == 0)
285 				return(string);
286 			if (*a++ != *b++)
287 				break;
288 		}
289 		b = substring;
290 	}
291 	return((char *) NULL);
292 }
293 
294 /*
295  * Str_Match --
296  *
297  * See if a particular string matches a particular pattern.
298  *
299  * Results: Non-zero is returned if string matches pattern, 0 otherwise. The
300  * matching operation permits the following special characters in the
301  * pattern: *?\[] (see the man page for details on what these mean).
302  *
303  * Side effects: None.
304  */
305 int
306 Str_Match(string, pattern)
307 	register char *string;		/* String */
308 	register char *pattern;		/* Pattern */
309 {
310 	char c2;
311 
312 	for (;;) {
313 		/*
314 		 * See if we're at the end of both the pattern and the
315 		 * string. If, we succeeded.  If we're at the end of the
316 		 * pattern but not at the end of the string, we failed.
317 		 */
318 		if (*pattern == 0)
319 			return(!*string);
320 		if (*string == 0 && *pattern != '*')
321 			return(0);
322 		/*
323 		 * Check for a "*" as the next pattern character.  It matches
324 		 * any substring.  We handle this by calling ourselves
325 		 * recursively for each postfix of string, until either we
326 		 * match or we reach the end of the string.
327 		 */
328 		if (*pattern == '*') {
329 			pattern += 1;
330 			if (*pattern == 0)
331 				return(1);
332 			while (*string != 0) {
333 				if (Str_Match(string, pattern))
334 					return(1);
335 				++string;
336 			}
337 			return(0);
338 		}
339 		/*
340 		 * Check for a "?" as the next pattern character.  It matches
341 		 * any single character.
342 		 */
343 		if (*pattern == '?')
344 			goto thisCharOK;
345 		/*
346 		 * Check for a "[" as the next pattern character.  It is
347 		 * followed by a list of characters that are acceptable, or
348 		 * by a range (two characters separated by "-").
349 		 */
350 		if (*pattern == '[') {
351 			++pattern;
352 			for (;;) {
353 				if ((*pattern == ']') || (*pattern == 0))
354 					return(0);
355 				if (*pattern == *string)
356 					break;
357 				if (pattern[1] == '-') {
358 					c2 = pattern[2];
359 					if (c2 == 0)
360 						return(0);
361 					if ((*pattern <= *string) &&
362 					    (c2 >= *string))
363 						break;
364 					if ((*pattern >= *string) &&
365 					    (c2 <= *string))
366 						break;
367 					pattern += 2;
368 				}
369 				++pattern;
370 			}
371 			while ((*pattern != ']') && (*pattern != 0))
372 				++pattern;
373 			goto thisCharOK;
374 		}
375 		/*
376 		 * If the next pattern character is '/', just strip off the
377 		 * '/' so we do exact matching on the character that follows.
378 		 */
379 		if (*pattern == '\\') {
380 			++pattern;
381 			if (*pattern == 0)
382 				return(0);
383 		}
384 		/*
385 		 * There's no special character.  Just make sure that the
386 		 * next characters of each string match.
387 		 */
388 		if (*pattern != *string)
389 			return(0);
390 thisCharOK:	++pattern;
391 		++string;
392 	}
393 }
394 
395 
396 /*-
397  *-----------------------------------------------------------------------
398  * Str_SYSVMatch --
399  *	Check word against pattern for a match (% is wild),
400  *
401  * Results:
402  *	Returns the beginning position of a match or null. The number
403  *	of characters matched is returned in len.
404  *
405  * Side Effects:
406  *	None
407  *
408  *-----------------------------------------------------------------------
409  */
410 char *
411 Str_SYSVMatch(word, pattern, len)
412     char	*word;		/* Word to examine */
413     char	*pattern;	/* Pattern to examine against */
414     int		*len;		/* Number of characters to substitute */
415 {
416     char *p = pattern;
417     char *w = word;
418     char *m;
419 
420     if (*p == '\0') {
421 	/* Null pattern is the whole string */
422 	*len = strlen(w);
423 	return w;
424     }
425 
426     if ((m = strchr(p, '%')) != NULL) {
427 	/* check that the prefix matches */
428 	for (; p != m && *w && *w == *p; w++, p++)
429 	     continue;
430 
431 	if (p != m)
432 	    return NULL;	/* No match */
433 
434 	if (*++p == '\0') {
435 	    /* No more pattern, return the rest of the string */
436 	    *len = strlen(w);
437 	    return w;
438 	}
439     }
440 
441     m = w;
442 
443     /* Find a matching tail */
444     do
445 	if (strcmp(p, w) == 0) {
446 	    *len = w - m;
447 	    return m;
448 	}
449     while (*w++ != '\0');
450 
451     return NULL;
452 }
453 
454 
455 /*-
456  *-----------------------------------------------------------------------
457  * Str_SYSVSubst --
458  *	Substitute '%' on the pattern with len characters from src.
459  *	If the pattern does not contain a '%' prepend len characters
460  *	from src.
461  *
462  * Results:
463  *	None
464  *
465  * Side Effects:
466  *	Places result on buf
467  *
468  *-----------------------------------------------------------------------
469  */
470 void
471 Str_SYSVSubst(buf, pat, src, len)
472     Buffer buf;
473     char *pat;
474     char *src;
475     int   len;
476 {
477     char *m;
478 
479     if ((m = strchr(pat, '%')) != NULL) {
480 	/* Copy the prefix */
481 	Buf_AddBytes(buf, m - pat, (Byte *) pat);
482 	/* skip the % */
483 	pat = m + 1;
484     }
485 
486     /* Copy the pattern */
487     Buf_AddBytes(buf, len, (Byte *) src);
488 
489     /* append the rest */
490     Buf_AddBytes(buf, strlen(pat), (Byte *) pat);
491 }
492