xref: /freebsd/usr.bin/sed/main.c (revision f0adf7f5cdd241db2f2c817683191a6ef64a4e95)
1 /*-
2  * Copyright (c) 1992 Diomidis Spinellis.
3  * Copyright (c) 1992, 1993
4  *	The Regents of the University of California.  All rights reserved.
5  *
6  * This code is derived from software contributed to Berkeley by
7  * Diomidis Spinellis of Imperial College, University of London.
8  *
9  * Redistribution and use in source and binary forms, with or without
10  * modification, are permitted provided that the following conditions
11  * are met:
12  * 1. Redistributions of source code must retain the above copyright
13  *    notice, this list of conditions and the following disclaimer.
14  * 2. Redistributions in binary form must reproduce the above copyright
15  *    notice, this list of conditions and the following disclaimer in the
16  *    documentation and/or other materials provided with the distribution.
17  * 3. All advertising materials mentioning features or use of this software
18  *    must display the following acknowledgement:
19  *	This product includes software developed by the University of
20  *	California, Berkeley and its contributors.
21  * 4. Neither the name of the University nor the names of its contributors
22  *    may be used to endorse or promote products derived from this software
23  *    without specific prior written permission.
24  *
25  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
26  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
27  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
28  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
29  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
30  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
31  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
32  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
33  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
34  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
35  * SUCH DAMAGE.
36  */
37 
38 #include <sys/cdefs.h>
39 __FBSDID("$FreeBSD$");
40 
41 #ifndef lint
42 static const char copyright[] =
43 "@(#) Copyright (c) 1992, 1993\n\
44 	The Regents of the University of California.  All rights reserved.\n";
45 #endif
46 
47 #ifndef lint
48 static const char sccsid[] = "@(#)main.c	8.2 (Berkeley) 1/3/94";
49 #endif
50 
51 #include <sys/types.h>
52 #include <sys/mman.h>
53 #include <sys/param.h>
54 #include <sys/stat.h>
55 
56 #include <err.h>
57 #include <errno.h>
58 #include <fcntl.h>
59 #include <limits.h>
60 #include <locale.h>
61 #include <regex.h>
62 #include <stddef.h>
63 #include <stdio.h>
64 #include <stdlib.h>
65 #include <string.h>
66 #include <unistd.h>
67 
68 #include "defs.h"
69 #include "extern.h"
70 
71 /*
72  * Linked list of units (strings and files) to be compiled
73  */
74 struct s_compunit {
75 	struct s_compunit *next;
76 	enum e_cut {CU_FILE, CU_STRING} type;
77 	char *s;			/* Pointer to string or fname */
78 };
79 
80 /*
81  * Linked list pointer to compilation units and pointer to current
82  * next pointer.
83  */
84 static struct s_compunit *script, **cu_nextp = &script;
85 
86 /*
87  * Linked list of files to be processed
88  */
89 struct s_flist {
90 	char *fname;
91 	struct s_flist *next;
92 };
93 
94 /*
95  * Linked list pointer to files and pointer to current
96  * next pointer.
97  */
98 static struct s_flist *files, **fl_nextp = &files;
99 
100 FILE *infile;			/* Current input file */
101 FILE *outfile;			/* Current output file */
102 
103 int aflag, eflag, nflag;
104 int rflags = 0;
105 static int rval;		/* Exit status */
106 
107 /*
108  * Current file and line number; line numbers restart across compilation
109  * units, but span across input files.
110  */
111 const char *fname;		/* File name. */
112 const char *outfname;		/* Output file name */
113 static char oldfname[PATH_MAX];	/* Old file name (for in-place editing) */
114 static char tmpfname[PATH_MAX];	/* Temporary file name (for in-place editing) */
115 const char *inplace;		/* Inplace edit file extension. */
116 u_long linenum;
117 
118 static void add_compunit(enum e_cut, char *);
119 static void add_file(char *);
120 static int inplace_edit(char **);
121 static void usage(void);
122 
123 int
124 main(int argc, char *argv[])
125 {
126 	int c, fflag;
127 	char *temp_arg;
128 
129 	(void) setlocale(LC_ALL, "");
130 
131 	fflag = 0;
132 	inplace = NULL;
133 
134 	while ((c = getopt(argc, argv, "Eae:f:i:n")) != -1)
135 		switch (c) {
136 		case 'E':
137 			rflags = REG_EXTENDED;
138 			break;
139 		case 'a':
140 			aflag = 1;
141 			break;
142 		case 'e':
143 			eflag = 1;
144 			if ((temp_arg = malloc(strlen(optarg) + 2)) == NULL)
145 				err(1, "malloc");
146 			strcpy(temp_arg, optarg);
147 			strcat(temp_arg, "\n");
148 			add_compunit(CU_STRING, temp_arg);
149 			break;
150 		case 'f':
151 			fflag = 1;
152 			add_compunit(CU_FILE, optarg);
153 			break;
154 		case 'i':
155 			inplace = optarg;
156 			break;
157 		case 'n':
158 			nflag = 1;
159 			break;
160 		default:
161 		case '?':
162 			usage();
163 		}
164 	argc -= optind;
165 	argv += optind;
166 
167 	/* First usage case; script is the first arg */
168 	if (!eflag && !fflag && *argv) {
169 		add_compunit(CU_STRING, *argv);
170 		argv++;
171 	}
172 
173 	compile();
174 
175 	/* Continue with first and start second usage */
176 	if (*argv)
177 		for (; *argv; argv++)
178 			add_file(*argv);
179 	else
180 		add_file(NULL);
181 	process();
182 	cfclose(prog, NULL);
183 	if (fclose(stdout))
184 		err(1, "stdout");
185 	exit(rval);
186 }
187 
188 static void
189 usage(void)
190 {
191 	(void)fprintf(stderr, "%s\n%s\n",
192 		"usage: sed script [-Ean] [-i extension] [file ...]",
193 		"       sed [-an] [-i extension] [-e script] ... [-f script_file] ... [file ...]");
194 	exit(1);
195 }
196 
197 /*
198  * Like fgets, but go through the chain of compilation units chaining them
199  * together.  Empty strings and files are ignored.
200  */
201 char *
202 cu_fgets(char *buf, int n, int *more)
203 {
204 	static enum {ST_EOF, ST_FILE, ST_STRING} state = ST_EOF;
205 	static FILE *f;		/* Current open file */
206 	static char *s;		/* Current pointer inside string */
207 	static char string_ident[30];
208 	char *p;
209 
210 again:
211 	switch (state) {
212 	case ST_EOF:
213 		if (script == NULL) {
214 			if (more != NULL)
215 				*more = 0;
216 			return (NULL);
217 		}
218 		linenum = 0;
219 		switch (script->type) {
220 		case CU_FILE:
221 			if ((f = fopen(script->s, "r")) == NULL)
222 				err(1, "%s", script->s);
223 			fname = script->s;
224 			state = ST_FILE;
225 			goto again;
226 		case CU_STRING:
227 			if ((snprintf(string_ident,
228 			    sizeof(string_ident), "\"%s\"", script->s)) >=
229 			    sizeof(string_ident) - 1)
230 				(void)strcpy(string_ident +
231 				    sizeof(string_ident) - 6, " ...\"");
232 			fname = string_ident;
233 			s = script->s;
234 			state = ST_STRING;
235 			goto again;
236 		}
237 	case ST_FILE:
238 		if ((p = fgets(buf, n, f)) != NULL) {
239 			linenum++;
240 			if (linenum == 1 && buf[0] == '#' && buf[1] == 'n')
241 				nflag = 1;
242 			if (more != NULL)
243 				*more = !feof(f);
244 			return (p);
245 		}
246 		script = script->next;
247 		(void)fclose(f);
248 		state = ST_EOF;
249 		goto again;
250 	case ST_STRING:
251 		if (linenum == 0 && s[0] == '#' && s[1] == 'n')
252 			nflag = 1;
253 		p = buf;
254 		for (;;) {
255 			if (n-- <= 1) {
256 				*p = '\0';
257 				linenum++;
258 				if (more != NULL)
259 					*more = 1;
260 				return (buf);
261 			}
262 			switch (*s) {
263 			case '\0':
264 				state = ST_EOF;
265 				if (s == script->s) {
266 					script = script->next;
267 					goto again;
268 				} else {
269 					script = script->next;
270 					*p = '\0';
271 					linenum++;
272 					if (more != NULL)
273 						*more = 0;
274 					return (buf);
275 				}
276 			case '\n':
277 				*p++ = '\n';
278 				*p = '\0';
279 				s++;
280 				linenum++;
281 				if (more != NULL)
282 					*more = 0;
283 				return (buf);
284 			default:
285 				*p++ = *s++;
286 			}
287 		}
288 	}
289 	/* NOTREACHED */
290 	return (NULL);
291 }
292 
293 /*
294  * Like fgets, but go through the list of files chaining them together.
295  * Set len to the length of the line.
296  */
297 int
298 mf_fgets(SPACE *sp, enum e_spflag spflag)
299 {
300 	struct stat sb;
301 	size_t len;
302 	char *p;
303 	int c;
304 	static int firstfile;
305 
306 	if (infile == NULL) {
307 		/* stdin? */
308 		if (files->fname == NULL) {
309 			if (inplace != NULL)
310 				errx(1, "-i may not be used with stdin");
311 			infile = stdin;
312 			fname = "stdin";
313 			outfile = stdout;
314 			outfname = "stdout";
315 		}
316 		firstfile = 1;
317 	}
318 
319 	for (;;) {
320 		if (infile != NULL && (c = getc(infile)) != EOF) {
321 			(void)ungetc(c, infile);
322 			break;
323 		}
324 		/* If we are here then either eof or no files are open yet */
325 		if (infile == stdin) {
326 			sp->len = 0;
327 			return (0);
328 		}
329 		if (infile != NULL) {
330 			fclose(infile);
331 			if (*oldfname != '\0') {
332 				if (rename(fname, oldfname) != 0) {
333 					warn("rename()");
334 					unlink(tmpfname);
335 					exit(1);
336 				}
337 				*oldfname = '\0';
338 			}
339 			if (*tmpfname != '\0') {
340 				if (outfile != NULL && outfile != stdout)
341 					fclose(outfile);
342 				outfile = NULL;
343 				rename(tmpfname, fname);
344 				*tmpfname = '\0';
345 			}
346 			outfname = NULL;
347 		}
348 		if (firstfile == 0)
349 			files = files->next;
350 		else
351 			firstfile = 0;
352 		if (files == NULL) {
353 			sp->len = 0;
354 			return (0);
355 		}
356 		fname = files->fname;
357 		if (inplace != NULL) {
358 			if (lstat(fname, &sb) != 0)
359 				err(1, "%s", fname);
360 			if (!(sb.st_mode & S_IFREG))
361 				errx(1, "%s: %s %s", fname,
362 				    "in-place editing only",
363 				    "works for regular files");
364 			if (*inplace != '\0') {
365 				strlcpy(oldfname, fname,
366 				    sizeof(oldfname));
367 				len = strlcat(oldfname, inplace,
368 				    sizeof(oldfname));
369 				if (len > sizeof(oldfname))
370 					errx(1, "%s: name too long", fname);
371 			}
372 			len = snprintf(tmpfname, sizeof(tmpfname),
373 			    "%s/.!%ld!%s", dirname(fname), (long)getpid(),
374 			    basename(fname));
375 			if (len >= sizeof(tmpfname))
376 				errx(1, "%s: name too long", fname);
377 			unlink(tmpfname);
378 			if ((outfile = fopen(tmpfname, "w")) == NULL)
379 				err(1, "%s", fname);
380 			fchown(fileno(outfile), sb.st_uid, sb.st_gid);
381 			fchmod(fileno(outfile), sb.st_mode & ALLPERMS);
382 			outfname = tmpfname;
383 		} else {
384 			outfile = stdout;
385 			outfname = "stdout";
386 		}
387 		if ((infile = fopen(fname, "r")) == NULL) {
388 			warn("%s", fname);
389 			rval = 1;
390 			continue;
391 		}
392 	}
393 	/*
394 	 * We are here only when infile is open and we still have something
395 	 * to read from it.
396 	 *
397 	 * Use fgetln so that we can handle essentially infinite input data.
398 	 * Can't use the pointer into the stdio buffer as the process space
399 	 * because the ungetc() can cause it to move.
400 	 */
401 	p = fgetln(infile, &len);
402 	if (ferror(infile))
403 		errx(1, "%s: %s", fname, strerror(errno ? errno : EIO));
404 	if (len != 0 && p[len - 1] == '\n')
405 		len--;
406 	cspace(sp, p, len, spflag);
407 
408 	linenum++;
409 
410 	return (1);
411 }
412 
413 /*
414  * Add a compilation unit to the linked list
415  */
416 static void
417 add_compunit(enum e_cut type, char *s)
418 {
419 	struct s_compunit *cu;
420 
421 	if ((cu = malloc(sizeof(struct s_compunit))) == NULL)
422 		err(1, "malloc");
423 	cu->type = type;
424 	cu->s = s;
425 	cu->next = NULL;
426 	*cu_nextp = cu;
427 	cu_nextp = &cu->next;
428 }
429 
430 /*
431  * Add a file to the linked list
432  */
433 static void
434 add_file(char *s)
435 {
436 	struct s_flist *fp;
437 
438 	if ((fp = malloc(sizeof(struct s_flist))) == NULL)
439 		err(1, "malloc");
440 	fp->next = NULL;
441 	*fl_nextp = fp;
442 	fp->fname = s;
443 	fl_nextp = &fp->next;
444 }
445 
446 int
447 lastline(void)
448 {
449 	int ch;
450 
451 	if (files->next != NULL)
452 		return (0);
453 	if ((ch = getc(infile)) == EOF)
454 		return (1);
455 	ungetc(ch, infile);
456 	return (0);
457 }
458