xref: /openbsd/usr.bin/sed/main.c (revision 73471bf0)
1 /*	$OpenBSD: main.c,v 1.42 2021/01/31 14:23:05 naddy Exp $	*/
2 
3 /*-
4  * Copyright (c) 1992 Diomidis Spinellis.
5  * Copyright (c) 1992, 1993
6  *	The Regents of the University of California.  All rights reserved.
7  *
8  * This code is derived from software contributed to Berkeley by
9  * Diomidis Spinellis of Imperial College, University of London.
10  *
11  * Redistribution and use in source and binary forms, with or without
12  * modification, are permitted provided that the following conditions
13  * are met:
14  * 1. Redistributions of source code must retain the above copyright
15  *    notice, this list of conditions and the following disclaimer.
16  * 2. Redistributions in binary form must reproduce the above copyright
17  *    notice, this list of conditions and the following disclaimer in the
18  *    documentation and/or other materials provided with the distribution.
19  * 3. Neither the name of the University nor the names of its contributors
20  *    may be used to endorse or promote products derived from this software
21  *    without specific prior written permission.
22  *
23  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
24  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
25  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
26  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
27  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
28  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
29  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
30  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
31  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
32  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
33  * SUCH DAMAGE.
34  */
35 
36 #include <sys/types.h>
37 #include <sys/ioctl.h>
38 #include <sys/stat.h>
39 
40 #include <ctype.h>
41 #include <errno.h>
42 #include <fcntl.h>
43 #include <limits.h>
44 #include <regex.h>
45 #include <stddef.h>
46 #include <stdio.h>
47 #include <stdlib.h>
48 #include <string.h>
49 #include <unistd.h>
50 #include <libgen.h>
51 
52 #include "defs.h"
53 #include "extern.h"
54 
55 /*
56  * Linked list of units (strings and files) to be compiled
57  */
58 struct s_compunit {
59 	struct s_compunit *next;
60 	enum e_cut {CU_FILE, CU_STRING} type;
61 	char *s;			/* Pointer to string or fname */
62 };
63 
64 /*
65  * Linked list pointer to compilation units and pointer to current
66  * next pointer.
67  */
68 static struct s_compunit *script, **cu_nextp = &script;
69 
70 /*
71  * Linked list of files to be processed
72  */
73 struct s_flist {
74 	char *fname;
75 	struct s_flist *next;
76 };
77 
78 /*
79  * Linked list pointer to files and pointer to current
80  * next pointer.
81  */
82 static struct s_flist *files, **fl_nextp = &files;
83 
84 FILE *infile;			/* Current input file */
85 FILE *outfile;			/* Current output file */
86 
87 int Eflag, aflag, eflag, nflag;
88 static int rval;	/* Exit status */
89 
90 /*
91  * Current file and line number; line numbers restart across compilation
92  * units, but span across input files.  The latter is optional if editing
93  * in place.
94  */
95 const char *fname;		/* File name. */
96 const char *outfname;		/* Output file name */
97 static char oldfname[PATH_MAX];	/* Old file name (for in-place editing) */
98 static char tmpfname[PATH_MAX];	/* Temporary file name (for in-place editing) */
99 char *inplace;			/* Inplace edit file extension */
100 u_long linenum;
101 
102 static void add_compunit(enum e_cut, char *);
103 static void add_file(char *);
104 static int next_files_have_lines(void);
105 
106 int termwidth;
107 
108 int pledge_wpath, pledge_rpath;
109 
110 int
111 main(int argc, char *argv[])
112 {
113 	struct winsize win;
114 	int c, fflag;
115 	char *p;
116 
117 	fflag = 0;
118 	inplace = NULL;
119 	while ((c = getopt(argc, argv, "Eae:f:i::nru")) != -1)
120 		switch (c) {
121 		case 'E':
122 		case 'r':
123 			Eflag = 1;
124 			break;
125 		case 'a':
126 			aflag = 1;
127 			break;
128 		case 'e':
129 			eflag = 1;
130 			add_compunit(CU_STRING, optarg);
131 			break;
132 		case 'f':
133 			fflag = 1;
134 			add_compunit(CU_FILE, optarg);
135 			break;
136 		case 'i':
137 			inplace = optarg ? optarg : "";
138 			break;
139 		case 'n':
140 			nflag = 1;
141 			break;
142 		case 'u':
143 			setvbuf(stdout, NULL, _IOLBF, 0);
144 			break;
145 		default:
146 		case '?':
147 			(void)fprintf(stderr,
148 			    "usage: sed [-aEnru] [-i[extension]] command [file ...]\n"
149 			    "       sed [-aEnru] [-e command] [-f command_file] [-i[extension]] [file ...]\n");
150 			exit(1);
151 		}
152 	argc -= optind;
153 	argv += optind;
154 
155 	termwidth = 0;
156 	if ((p = getenv("COLUMNS")) != NULL)
157 		termwidth = strtonum(p, 0, INT_MAX, NULL);
158 	if (termwidth == 0 && ioctl(STDOUT_FILENO, TIOCGWINSZ, &win) == 0 &&
159 	    win.ws_col > 0)
160 		termwidth = win.ws_col;
161 	if (termwidth == 0)
162 		termwidth = 80;
163 	if (termwidth <= 8)
164 		termwidth = 1;
165 	else
166 		termwidth -= 8;
167 
168 	if (inplace != NULL) {
169 		if (pledge("stdio rpath wpath cpath fattr chown", NULL) == -1)
170 			error(FATAL, "pledge: %s", strerror(errno));
171 	} else {
172 		if (pledge("stdio rpath wpath cpath", NULL) == -1)
173 			error(FATAL, "pledge: %s", strerror(errno));
174 	}
175 
176 	/* First usage case; script is the first arg */
177 	if (!eflag && !fflag && *argv) {
178 		add_compunit(CU_STRING, *argv);
179 		argv++;
180 	}
181 
182 	compile();
183 
184 	/* Continue with first and start second usage */
185 	if (*argv) {
186 		if (!pledge_wpath && inplace == NULL) {
187 			if (pledge("stdio rpath", NULL) == -1)
188 				error(FATAL, "pledge: %s", strerror(errno));
189 		}
190 		for (; *argv; argv++)
191 			add_file(*argv);
192 	} else {
193 		if (!pledge_wpath && !pledge_rpath) {
194 			if (pledge("stdio", NULL) == -1)
195 				error(FATAL, "pledge: %s", strerror(errno));
196 		} else if (pledge_rpath) {
197 			if (pledge("stdio rpath", NULL) == -1)
198 				error(FATAL, "pledge: %s", strerror(errno));
199 		} else if (pledge_wpath) {
200 			if (pledge("stdio wpath cpath", NULL) == -1)
201 				error(FATAL, "pledge: %s", strerror(errno));
202 		}
203 		add_file(NULL);
204 	}
205 	process();
206 	cfclose(prog, NULL);
207 	if (fclose(stdout))
208 		error(FATAL, "stdout: %s", strerror(errno));
209 	exit (rval);
210 }
211 
212 /*
213  * Like fgets, but go through the chain of compilation units chaining them
214  * together.  Empty strings and files are ignored.
215  */
216 char *
217 cu_fgets(char **outbuf, size_t *outsize)
218 {
219 	static enum {ST_EOF, ST_FILE, ST_STRING} state = ST_EOF;
220 	static FILE *f;		/* Current open file */
221 	static char *s;		/* Current pointer inside string */
222 	static char string_ident[30];
223 	size_t len;
224 	char *p;
225 
226 	if (*outbuf == NULL)
227 		*outsize = 0;
228 
229 again:
230 	switch (state) {
231 	case ST_EOF:
232 		if (script == NULL)
233 			return (NULL);
234 		linenum = 0;
235 		switch (script->type) {
236 		case CU_FILE:
237 			if ((f = fopen(script->s, "r")) == NULL)
238 				error(FATAL,
239 				    "%s: %s", script->s, strerror(errno));
240 			fname = script->s;
241 			state = ST_FILE;
242 			goto again;
243 		case CU_STRING:
244 			len = snprintf(string_ident, sizeof(string_ident),
245 			    "\"%s\"", script->s);
246 			if (len >= sizeof(string_ident))
247 				strlcpy(string_ident +
248 				    sizeof(string_ident) - 6, " ...\"", 5);
249 			fname = string_ident;
250 			s = script->s;
251 			state = ST_STRING;
252 			goto again;
253 		}
254 	case ST_FILE:
255 		if (getline(outbuf, outsize, f) != -1) {
256 			p = *outbuf;
257 			linenum++;
258 			if (linenum == 1 && p[0] == '#' && p[1] == 'n')
259 				nflag = 1;
260 			return (*outbuf);
261 		}
262 		script = script->next;
263 		(void)fclose(f);
264 		state = ST_EOF;
265 		goto again;
266 	case ST_STRING:
267 		if (linenum == 0 && s[0] == '#' && s[1] == 'n')
268 			nflag = 1;
269 		p = *outbuf;
270 		len = *outsize;
271 		for (;;) {
272 			if (len <= 1) {
273 				*outbuf = xrealloc(*outbuf,
274 				    *outsize + _POSIX2_LINE_MAX);
275 				p = *outbuf + *outsize - len;
276 				len += _POSIX2_LINE_MAX;
277 				*outsize += _POSIX2_LINE_MAX;
278 			}
279 			switch (*s) {
280 			case '\0':
281 				state = ST_EOF;
282 				if (s == script->s) {
283 					script = script->next;
284 					goto again;
285 				} else {
286 					script = script->next;
287 					*p = '\0';
288 					linenum++;
289 					return (*outbuf);
290 				}
291 			case '\n':
292 				*p++ = '\n';
293 				*p = '\0';
294 				s++;
295 				linenum++;
296 				return (*outbuf);
297 			default:
298 				*p++ = *s++;
299 				len--;
300 			}
301 		}
302 	}
303 
304 	return (NULL);
305 }
306 
307 void
308 finish_file(void)
309 {
310 	if (infile != NULL) {
311 		fclose(infile);
312 		if (*oldfname != '\0') {
313 			if (rename(fname, oldfname) != 0) {
314 				warning("rename()");
315 				unlink(tmpfname);
316 				exit(1);
317 			}
318 			*oldfname = '\0';
319 		}
320 		if (*tmpfname != '\0') {
321 			if (outfile != NULL && outfile != stdout)
322 				fclose(outfile);
323 			outfile = NULL;
324 			rename(tmpfname, fname);
325 			*tmpfname = '\0';
326 		}
327 		outfname = NULL;
328 	}
329 }
330 
331 /*
332  * Like fgets, but go through the list of files chaining them together.
333  * Set len to the length of the line.
334  */
335 int
336 mf_fgets(SPACE *sp, enum e_spflag spflag)
337 {
338 	struct stat sb;
339 	size_t len;
340 	char dirbuf[PATH_MAX];
341 	static char *p;
342 	static size_t psize;
343 	int c, fd;
344 	static int firstfile;
345 
346 	if (infile == NULL) {
347 		/* stdin? */
348 		if (files->fname == NULL) {
349 			if (inplace != NULL)
350 				error(FATAL, "-i may not be used with stdin");
351 			infile = stdin;
352 			fname = "stdin";
353 			outfile = stdout;
354 			outfname = "stdout";
355 		}
356 
357 		firstfile = 1;
358 	}
359 
360 	for (;;) {
361 		if (infile != NULL && (c = getc(infile)) != EOF) {
362 			(void)ungetc(c, infile);
363 			break;
364 		}
365 		/* If we are here then either eof or no files are open yet */
366 		if (infile == stdin) {
367 			sp->len = 0;
368 			return (0);
369 		}
370 		finish_file();
371 		if (firstfile == 0)
372 			files = files->next;
373 		else
374 			firstfile = 0;
375 		if (files == NULL) {
376 			sp->len = 0;
377 			return (0);
378 		}
379 		fname = files->fname;
380 		if (inplace != NULL) {
381 			if (lstat(fname, &sb) != 0)
382 				error(FATAL, "%s: %s", fname,
383 				    strerror(errno ? errno : EIO));
384 			if (!S_ISREG(sb.st_mode))
385 				error(FATAL, "%s: %s %s", fname,
386 				    "in-place editing only",
387 				    "works for regular files");
388 			if (*inplace != '\0') {
389 				strlcpy(oldfname, fname,
390 				    sizeof(oldfname));
391 				len = strlcat(oldfname, inplace,
392 				    sizeof(oldfname));
393 				if (len > sizeof(oldfname))
394 					error(FATAL, "%s: name too long", fname);
395 			}
396 			strlcpy(dirbuf, fname, sizeof(dirbuf));
397 			len = snprintf(tmpfname, sizeof(tmpfname),
398 			    "%s/sedXXXXXXXXXX", dirname(dirbuf));
399 			if (len >= sizeof(tmpfname))
400 				error(FATAL, "%s: name too long", fname);
401 			if ((fd = mkstemp(tmpfname)) == -1)
402 				error(FATAL, "%s: %s", fname, strerror(errno));
403 			if ((outfile = fdopen(fd, "w")) == NULL) {
404 				unlink(tmpfname);
405 				error(FATAL, "%s", fname);
406 			}
407 			fchown(fileno(outfile), sb.st_uid, sb.st_gid);
408 			fchmod(fileno(outfile), sb.st_mode & ALLPERMS);
409 			outfname = tmpfname;
410 			linenum = 0;
411 			resetstate();
412 		} else {
413 			outfile = stdout;
414 			outfname = "stdout";
415 		}
416 		if ((infile = fopen(fname, "r")) == NULL) {
417 			warning("%s", strerror(errno));
418 			rval = 1;
419 			continue;
420 		}
421 	}
422 
423 	/*
424 	 * We are here only when infile is open and we still have something
425 	 * to read from it.
426 	 *
427 	 * Use getline() so that we can handle essentially infinite input
428 	 * data.  The p and psize are static so each invocation gives
429 	 * getline() the same buffer which is expanded as needed.
430 	 */
431 	len = getline(&p, &psize, infile);
432 	if ((ssize_t)len == -1)
433 		error(FATAL, "%s: %s", fname, strerror(errno));
434 	if (len != 0 && p[len - 1] == '\n') {
435 		sp->append_newline = 1;
436 		len--;
437 	} else if (!lastline()) {
438 		sp->append_newline = 1;
439 	} else {
440 		sp->append_newline = 0;
441 	}
442 	cspace(sp, p, len, spflag);
443 
444 	linenum++;
445 
446 	return (1);
447 }
448 
449 /*
450  * Add a compilation unit to the linked list
451  */
452 static void
453 add_compunit(enum e_cut type, char *s)
454 {
455 	struct s_compunit *cu;
456 
457 	cu = xmalloc(sizeof(struct s_compunit));
458 	cu->type = type;
459 	cu->s = s;
460 	cu->next = NULL;
461 	*cu_nextp = cu;
462 	cu_nextp = &cu->next;
463 }
464 
465 /*
466  * Add a file to the linked list
467  */
468 static void
469 add_file(char *s)
470 {
471 	struct s_flist *fp;
472 
473 	fp = xmalloc(sizeof(struct s_flist));
474 	fp->next = NULL;
475 	*fl_nextp = fp;
476 	fp->fname = s;
477 	fl_nextp = &fp->next;
478 }
479 
480 
481 static int
482 next_files_have_lines()
483 {
484 	struct s_flist *file;
485 	FILE *file_fd;
486 	int ch;
487 
488 	file = files;
489 	while ((file = file->next) != NULL) {
490 		if ((file_fd = fopen(file->fname, "r")) == NULL)
491 			continue;
492 
493 		if ((ch = getc(file_fd)) != EOF) {
494 			/*
495 			 * This next file has content, therefore current
496 			 * file doesn't contains the last line.
497 			 */
498 			ungetc(ch, file_fd);
499 			fclose(file_fd);
500 			return (1);
501 		}
502 		fclose(file_fd);
503 	}
504 	return (0);
505 }
506 
507 int
508 lastline(void)
509 {
510 	int ch;
511 
512 	if (feof(infile)) {
513 		return !(
514 		    (inplace == NULL) &&
515 		    next_files_have_lines());
516 	}
517 	if ((ch = getc(infile)) == EOF) {
518 		return !(
519 		    (inplace == NULL) &&
520 		    next_files_have_lines());
521 	}
522 	ungetc(ch, infile);
523 	return (0);
524 }
525