xref: /netbsd-src/usr.bin/ctags/ctags.c (revision 89b375bec59bc116375cfc77bf119f37b996ded6)
1 /*	$NetBSD: ctags.c,v 1.18 2024/10/31 01:50:20 kre Exp $	*/
2 
3 /*
4  * Copyright (c) 1987, 1993, 1994, 1995
5  *	The Regents of the University of California.  All rights reserved.
6  *
7  * Redistribution and use in source and binary forms, with or without
8  * modification, are permitted provided that the following conditions
9  * are met:
10  * 1. Redistributions of source code must retain the above copyright
11  *    notice, this list of conditions and the following disclaimer.
12  * 2. Redistributions in binary form must reproduce the above copyright
13  *    notice, this list of conditions and the following disclaimer in the
14  *    documentation and/or other materials provided with the distribution.
15  * 3. Neither the name of the University nor the names of its contributors
16  *    may be used to endorse or promote products derived from this software
17  *    without specific prior written permission.
18  *
19  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
20  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
21  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
22  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
23  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
24  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
25  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
26  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
27  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
28  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
29  * SUCH DAMAGE.
30  */
31 
32 #if HAVE_NBTOOL_CONFIG_H
33 #include "nbtool_config.h"
34 #endif
35 
36 #include <sys/cdefs.h>
37 #if defined(__COPYRIGHT) && !defined(lint)
38 __COPYRIGHT("@(#) Copyright (c) 1987, 1993, 1994, 1995\
39  The Regents of the University of California.  All rights reserved.");
40 #endif /* not lint */
41 
42 #if defined(__RCSID) && !defined(lint)
43 #if 0
44 static char sccsid[] = "@(#)ctags.c	8.4 (Berkeley) 2/7/95";
45 #endif
46 __RCSID("$NetBSD: ctags.c,v 1.18 2024/10/31 01:50:20 kre Exp $");
47 #endif /* not lint */
48 
49 #include <err.h>
50 #include <limits.h>
51 #include <stdio.h>
52 #include <string.h>
53 #include <stdlib.h>
54 #include <unistd.h>
55 
56 #include "ctags.h"
57 
58 /*
59  * ctags: create a tags file
60  */
61 
62 NODE	*head;			/* head of the sorted binary tree */
63 
64 				/* boolean "func" (see init()) */
65 bool	_wht[256], _etk[256], _itk[256], _btk[256], _gd[256];
66 
67 FILE	*inf;			/* ioptr for current input file */
68 FILE	*outf;			/* ioptr for tags file */
69 
70 long	lineftell;		/* ftell after getc( inf ) == '\n' */
71 
72 int	lineno;			/* line number of current line */
73 int	dflag;			/* -d: non-macro defines */
74 int	tflag;			/* -t: create tags for typedefs */
75 int	vflag;			/* -v: vgrind style index output */
76 int	wflag;			/* -w: suppress warnings */
77 int	xflag;			/* -x: cxref style output */
78 
79 char	*curfile;		/* current input file name */
80 char	searchar = '/';		/* use /.../ searches by default */
81 char	lbuf[LINE_MAX];
82 
83 void	init(void);
84 void	find_entries(char *);
85 
86 int
87 main(int argc, char **argv)
88 {
89 	static const char	*outfile = "tags";	/* output file */
90 	int	aflag;				/* -a: append to tags */
91 	int	uflag;				/* -u: update tags */
92 	int	exit_val;			/* exit value */
93 	int	step;				/* step through args */
94 	int	ch;				/* getopts char */
95 	char	*cmd;				/* not nice */
96 	char	tname[128];			/* disgusting */
97 	size_t	sz;
98 
99 	aflag = uflag = NO;
100 	while ((ch = getopt(argc, argv, "BFadf:tuwvx")) != -1)
101 		switch(ch) {
102 		case 'B':
103 			searchar = '?';
104 			break;
105 		case 'F':
106 			searchar = '/';
107 			break;
108 		case 'a':
109 			aflag++;
110 			break;
111 		case 'd':
112 			dflag++;
113 			break;
114 		case 'f':
115 			outfile = optarg;
116 			break;
117 		case 't':
118 			tflag++;
119 			break;
120 		case 'u':
121 			uflag++;
122 			break;
123 		case 'w':
124 			wflag++;
125 			break;
126 		case 'v':
127 			vflag++;
128 			/* FALLTHROUGH */
129 		case 'x':
130 			xflag++;
131 			break;
132 		case '?':
133 		default:
134 			goto usage;
135 		}
136 	argv += optind;
137 	argc -= optind;
138 	if (!argc) {
139  usage:;	(void)fprintf(stderr,
140 			"usage: ctags [-BFadtuwvx] [-f tagsfile] file ...\n");
141 		exit(EXIT_FAILURE);
142 	}
143 
144 	if ((sz = shquote(outfile, tname, sizeof tname)) >= sizeof tname) {
145 		/* nb: (size_t)-1 > sizeof tname */
146 		if (sz == (size_t)-1)
147 			err(EXIT_FAILURE, "Output file name '%s' too long",
148 			    outfile);
149 		else
150 			errx(EXIT_FAILURE, "Output file name '%s' too long",
151 			    outfile);
152 	}
153 
154 	init();
155 
156 	exit_val = EXIT_SUCCESS;
157 	for (step = 0; step < argc; ++step)
158 		if (!(inf = fopen(argv[step], "r"))) {
159 			warn("%s", argv[step]);
160 			exit_val = EXIT_FAILURE;
161 		}
162 		else {
163 			curfile = argv[step];
164 			find_entries(argv[step]);
165 			(void)fclose(inf);
166 		}
167 
168 	if (head) {
169 		if (xflag)
170 			put_entries(head);
171 		else {
172 			if (uflag) {
173 				for (step = 0; step < argc; step++) {
174 					char pattern[140];
175 
176 					if (asprintf(&cmd, "\t%s\t",
177 					    argv[step]) == -1)
178 						err(EXIT_FAILURE,
179 						   "Cannot generate '\\t%s\\t'",
180 						   argv[step]);
181 
182 					if ((sz = shquote(cmd, pattern,
183 					    sizeof pattern)) >= sizeof pattern)
184 					{
185 						if (sz == (size_t)-1)
186 						    err(EXIT_FAILURE, "'%s': "
187 							"quoting pattern", cmd);
188 						else
189 						    errx(EXIT_FAILURE, "'%s': "
190 							"failed to quote", cmd);
191 					}
192 					(void)free(cmd);
193 
194 					if (asprintf(&cmd,
195 				    	 "OTAGS=$(mktemp -t tags.$$) || exit\n"
196 					 "\ttest -n \"${OTAGS}\" || exit\n"
197 					 "\ttrap 'rm -f \"${OTAGS}\"' EXIT\n"
198 					 "\tmv %s \"${OTAGS}\" || exit\n"
199 					 "\tfgrep -v %s \"${OTAGS}\" >%s\n"
200 					 "\tX=$? ; "
201 					 "test \"$X\" -le 1 || exit $X",
202 					    tname, pattern, tname) == -1)
203 						err(EXIT_FAILURE,
204 						  "Command to update %s for -u"
205 						     " %s",
206 						  argv[step], outfile);
207 
208 					if (system(cmd) != 0)
209 						errx(EXIT_FAILURE,
210 						  "Update (-u) of %s failed.   "
211 						  "Cmd:\n    %s", outfile, cmd);
212 
213 					(void)free(cmd);
214 				}
215 				++aflag;
216 			}
217 			if (!(outf = fopen(outfile, aflag ? "a" : "w")))
218 				err(EXIT_FAILURE, "%s", outfile);
219 			put_entries(head);
220 			(void)fflush(outf);
221 			if (ferror(outf))
222 				err(EXIT_FAILURE, "output error (%s)", outfile);
223 			(void)fclose(outf);
224 			if (uflag) {
225 				if (asprintf(&cmd, "sort -o %s %s",
226 				    tname, tname) == -1)
227 					err(EXIT_FAILURE,
228 					    "sort command (-u) for %s",
229 					    outfile);
230 				if (system(cmd) != 0)
231 					errx(EXIT_FAILURE, "-u: sort %s failed"
232 					    "\t[ %s ]", outfile, cmd);
233 				(void)free(cmd);
234 			}
235 		}
236 	}
237 	if ((vflag || xflag) && (fflush(stdout) != 0 || ferror(stdout) != 0))
238 		errx(EXIT_FAILURE, "write error (stdout)");
239 	exit(exit_val);
240 }
241 
242 /*
243  * init --
244  *	this routine sets up the boolean pseudo-functions which work by
245  *	setting boolean flags dependent upon the corresponding character.
246  *	Every char which is NOT in that string is false with respect to
247  *	the pseudo-function.  Therefore, all of the array "_wht" is NO
248  *	by default and then the elements subscripted by the chars in
249  *	CWHITE are set to YES.  Thus, "_wht" of a char is YES if it is in
250  *	the string CWHITE, else NO.
251  */
252 void
253 init(void)
254 {
255 	int		i;
256 	const char	*sp;
257 
258 	for (i = 0; i < 256; i++) {
259 		_wht[i] = _etk[i] = _itk[i] = _btk[i] = NO;
260 		_gd[i] = YES;
261 	}
262 #define	CWHITE	" \f\t\n"
263 	for (sp = CWHITE; *sp; sp++)	/* white space chars */
264 		_wht[(unsigned)*sp] = YES;
265 #define	CTOKEN	" \t\n\"'#()[]{}=-+%*/&|^~!<>;,.:?"
266 	for (sp = CTOKEN; *sp; sp++)	/* token ending chars */
267 		_etk[(unsigned)*sp] = YES;
268 #define	CINTOK	"ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyz0123456789"
269 	for (sp = CINTOK; *sp; sp++)	/* valid in-token chars */
270 		_itk[(unsigned)*sp] = YES;
271 #define	CBEGIN	"ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyz"
272 	for (sp = CBEGIN; *sp; sp++)	/* token starting chars */
273 		_btk[(unsigned)*sp] = YES;
274 #define	CNOTGD	",;"
275 	for (sp = CNOTGD; *sp; sp++)	/* invalid after-function chars */
276 		_gd[(unsigned)*sp] = NO;
277 }
278 
279 /*
280  * find_entries --
281  *	this routine opens the specified file and calls the function
282  *	which searches the file.
283  */
284 void
285 find_entries(char *file)
286 {
287 	char	*cp;
288 
289 	lineno = 0;				/* should be 1 ?? KB */
290 	if ((cp = strrchr(file, '.')) != NULL) {
291 		if (cp[1] == 'l' && !cp[2]) {
292 			int	c;
293 
294 			for (;;) {
295 				if (GETC(==, EOF))
296 					return;
297 				if (!iswhite(c)) {
298 					rewind(inf);
299 					break;
300 				}
301 			}
302 #define	LISPCHR	";(["
303 /* lisp */		if (strchr(LISPCHR, c)) {
304 				l_entries();
305 				return;
306 			}
307 /* lex */		else {
308 				/*
309 				 * we search all 3 parts of a lex file
310 				 * for C references.  This may be wrong.
311 				 */
312 				toss_yysec();
313 				(void)strlcpy(lbuf, "%%$", sizeof(lbuf));
314 				pfnote("yylex", lineno);
315 				rewind(inf);
316 			}
317 		}
318 /* yacc */	else if (cp[1] == 'y' && !cp[2]) {
319 			/*
320 			 * we search only the 3rd part of a yacc file
321 			 * for C references.  This may be wrong.
322 			 */
323 			toss_yysec();
324 			(void)strlcpy(lbuf, "%%$", sizeof(lbuf));
325 			pfnote("yyparse", lineno);
326 			y_entries();
327 		}
328 /* fortran */	else if ((cp[1] != 'c' && cp[1] != 'h') && !cp[2]) {
329 			if (PF_funcs())
330 				return;
331 			rewind(inf);
332 		}
333 	}
334 /* C */	c_entries();
335 }
336