xref: /openbsd-src/usr.bin/sed/main.c (revision 897fc685943471cf985a0fe38ba076ea6fe74fa5)
1 /*	$OpenBSD: main.c,v 1.36 2017/12/13 16:06:34 millert 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 
164 	if (inplace != NULL) {
165 		if (pledge("stdio rpath wpath cpath fattr chown", NULL) == -1)
166 			error(FATAL, "pledge: %s", strerror(errno));
167 	} else {
168 		if (pledge("stdio rpath wpath cpath", NULL) == -1)
169 			error(FATAL, "pledge: %s", strerror(errno));
170 	}
171 
172 	/* First usage case; script is the first arg */
173 	if (!eflag && !fflag && *argv) {
174 		add_compunit(CU_STRING, *argv);
175 		argv++;
176 	}
177 
178 	compile();
179 
180 	/* Continue with first and start second usage */
181 	if (*argv) {
182 		if (!pledge_wpath && inplace == NULL) {
183 			if (pledge("stdio rpath", NULL) == -1)
184 				error(FATAL, "pledge: %s", strerror(errno));
185 		}
186 		for (; *argv; argv++)
187 			add_file(*argv);
188 	} else {
189 		if (!pledge_wpath && !pledge_rpath) {
190 			if (pledge("stdio", NULL) == -1)
191 				error(FATAL, "pledge: %s", strerror(errno));
192 		} else if (pledge_rpath) {
193 			if (pledge("stdio rpath", NULL) == -1)
194 				error(FATAL, "pledge: %s", strerror(errno));
195 		} else if (pledge_wpath) {
196 			if (pledge("stdio wpath cpath", NULL) == -1)
197 				error(FATAL, "pledge: %s", strerror(errno));
198 		}
199 		add_file(NULL);
200 	}
201 	process();
202 	cfclose(prog, NULL);
203 	if (fclose(stdout))
204 		error(FATAL, "stdout: %s", strerror(errno));
205 	exit (rval);
206 }
207 
208 /*
209  * Like fgets, but go through the chain of compilation units chaining them
210  * together.  Empty strings and files are ignored.
211  */
212 char *
213 cu_fgets(char **outbuf, size_t *outsize)
214 {
215 	static enum {ST_EOF, ST_FILE, ST_STRING} state = ST_EOF;
216 	static FILE *f;		/* Current open file */
217 	static char *s;		/* Current pointer inside string */
218 	static char string_ident[30];
219 	size_t len;
220 	char *p;
221 
222 	if (*outbuf == NULL)
223 		*outsize = 0;
224 
225 again:
226 	switch (state) {
227 	case ST_EOF:
228 		if (script == NULL)
229 			return (NULL);
230 		linenum = 0;
231 		switch (script->type) {
232 		case CU_FILE:
233 			if ((f = fopen(script->s, "r")) == NULL)
234 				error(FATAL,
235 				    "%s: %s", script->s, strerror(errno));
236 			fname = script->s;
237 			state = ST_FILE;
238 			goto again;
239 		case CU_STRING:
240 			len = snprintf(string_ident, sizeof(string_ident),
241 			    "\"%s\"", script->s);
242 			if (len >= sizeof(string_ident))
243 				strlcpy(string_ident +
244 				    sizeof(string_ident) - 6, " ...\"", 5);
245 			fname = string_ident;
246 			s = script->s;
247 			state = ST_STRING;
248 			goto again;
249 		}
250 	case ST_FILE:
251 		if ((p = fgetln(f, &len)) != NULL) {
252 			linenum++;
253 			if (len >= *outsize) {
254 				free(*outbuf);
255 				*outsize = ROUNDLEN(len + 1);
256 				*outbuf = xmalloc(*outsize);
257 			}
258 			memcpy(*outbuf, p, len);
259 			(*outbuf)[len] = '\0';
260 			if (linenum == 1 && p[0] == '#' && p[1] == 'n')
261 				nflag = 1;
262 			return (*outbuf);
263 		}
264 		script = script->next;
265 		(void)fclose(f);
266 		state = ST_EOF;
267 		goto again;
268 	case ST_STRING:
269 		if (linenum == 0 && s[0] == '#' && s[1] == 'n')
270 			nflag = 1;
271 		p = *outbuf;
272 		len = *outsize;
273 		for (;;) {
274 			if (len <= 1) {
275 				*outbuf = xrealloc(*outbuf,
276 				    *outsize + _POSIX2_LINE_MAX);
277 				p = *outbuf + *outsize - len;
278 				len += _POSIX2_LINE_MAX;
279 				*outsize += _POSIX2_LINE_MAX;
280 			}
281 			switch (*s) {
282 			case '\0':
283 				state = ST_EOF;
284 				if (s == script->s) {
285 					script = script->next;
286 					goto again;
287 				} else {
288 					script = script->next;
289 					*p = '\0';
290 					linenum++;
291 					return (*outbuf);
292 				}
293 			case '\n':
294 				*p++ = '\n';
295 				*p = '\0';
296 				s++;
297 				linenum++;
298 				return (*outbuf);
299 			default:
300 				*p++ = *s++;
301 				len--;
302 			}
303 		}
304 	}
305 
306 	return (NULL);
307 }
308 
309 /*
310  * Like fgets, but go through the list of files chaining them together.
311  * Set len to the length of the line.
312  */
313 int
314 mf_fgets(SPACE *sp, enum e_spflag spflag)
315 {
316 	struct stat sb;
317 	size_t len;
318 	char *p;
319 	int c, fd;
320 	static int firstfile;
321 
322 	if (infile == NULL) {
323 		/* stdin? */
324 		if (files->fname == NULL) {
325 			if (inplace != NULL)
326 				error(FATAL, "-i may not be used with stdin");
327 			infile = stdin;
328 			fname = "stdin";
329 			outfile = stdout;
330 			outfname = "stdout";
331 		}
332 
333 		firstfile = 1;
334 	}
335 
336 	for (;;) {
337 		if (infile != NULL && (c = getc(infile)) != EOF) {
338 			(void)ungetc(c, infile);
339 			break;
340 		}
341 		/* If we are here then either eof or no files are open yet */
342 		if (infile == stdin) {
343 			sp->len = 0;
344 			return (0);
345 		}
346 		if (infile != NULL) {
347 			fclose(infile);
348 			if (*oldfname != '\0') {
349 				if (rename(fname, oldfname) != 0) {
350 					warning("rename()");
351 					unlink(tmpfname);
352 					exit(1);
353 				}
354 				*oldfname = '\0';
355 			}
356 			if (*tmpfname != '\0') {
357 				if (outfile != NULL && outfile != stdout)
358 					fclose(outfile);
359 				outfile = NULL;
360 				rename(tmpfname, fname);
361 				*tmpfname = '\0';
362 			}
363 			outfname = NULL;
364 		}
365 		if (firstfile == 0)
366 			files = files->next;
367 		else
368 			firstfile = 0;
369 		if (files == NULL) {
370 			sp->len = 0;
371 			return (0);
372 		}
373 		fname = files->fname;
374 		if (inplace != NULL) {
375 			if (lstat(fname, &sb) != 0)
376 				error(FATAL, "%s: %s", fname,
377 				    strerror(errno ? errno : EIO));
378 			if (!S_ISREG(sb.st_mode))
379 				error(FATAL, "%s: %s %s", fname,
380 				    "in-place editing only",
381 				    "works for regular files");
382 			if (*inplace != '\0') {
383 				strlcpy(oldfname, fname,
384 				    sizeof(oldfname));
385 				len = strlcat(oldfname, inplace,
386 				    sizeof(oldfname));
387 				if (len > sizeof(oldfname))
388 					error(FATAL, "%s: name too long", fname);
389 			}
390 			len = snprintf(tmpfname, sizeof(tmpfname), "%s/sedXXXXXXXXXX",
391 			    dirname(fname));
392 			if (len >= sizeof(tmpfname))
393 				error(FATAL, "%s: name too long", fname);
394 			if ((fd = mkstemp(tmpfname)) == -1)
395 				error(FATAL, "%s: %s", fname, strerror(errno));
396 			if ((outfile = fdopen(fd, "w")) == NULL) {
397 				unlink(tmpfname);
398 				error(FATAL, "%s", fname);
399 			}
400 			fchown(fileno(outfile), sb.st_uid, sb.st_gid);
401 			fchmod(fileno(outfile), sb.st_mode & ALLPERMS);
402 			outfname = tmpfname;
403 			linenum = 0;
404 			resetranges();
405 		} else {
406 			outfile = stdout;
407 			outfname = "stdout";
408 		}
409 		if ((infile = fopen(fname, "r")) == NULL) {
410 			warning("%s", strerror(errno));
411 			rval = 1;
412 			continue;
413 		}
414 	}
415 
416 	/*
417 	 * We are here only when infile is open and we still have something
418 	 * to read from it.
419 	 *
420 	 * Use fgetln so that we can handle essentially infinite input data.
421 	 * Can't use the pointer into the stdio buffer as the process space
422 	 * because the ungetc() can cause it to move.
423 	 */
424 	p = fgetln(infile, &len);
425 	if (ferror(infile))
426 		error(FATAL, "%s: %s", fname, strerror(errno ? errno : EIO));
427 	if (len != 0 && p[len - 1] == '\n') {
428 		sp->append_newline = 1;
429 		len--;
430 	} else if (!lastline()) {
431 		sp->append_newline = 1;
432 	} else {
433 		sp->append_newline = 0;
434 	}
435 	cspace(sp, p, len, spflag);
436 
437 	linenum++;
438 
439 	return (1);
440 }
441 
442 /*
443  * Add a compilation unit to the linked list
444  */
445 static void
446 add_compunit(enum e_cut type, char *s)
447 {
448 	struct s_compunit *cu;
449 
450 	cu = xmalloc(sizeof(struct s_compunit));
451 	cu->type = type;
452 	cu->s = s;
453 	cu->next = NULL;
454 	*cu_nextp = cu;
455 	cu_nextp = &cu->next;
456 }
457 
458 /*
459  * Add a file to the linked list
460  */
461 static void
462 add_file(char *s)
463 {
464 	struct s_flist *fp;
465 
466 	fp = xmalloc(sizeof(struct s_flist));
467 	fp->next = NULL;
468 	*fl_nextp = fp;
469 	fp->fname = s;
470 	fl_nextp = &fp->next;
471 }
472 
473 
474 static int
475 next_files_have_lines()
476 {
477 	struct s_flist *file;
478 	FILE *file_fd;
479 	int ch;
480 
481 	file = files;
482 	while ((file = file->next) != NULL) {
483 		if ((file_fd = fopen(file->fname, "r")) == NULL)
484 			continue;
485 
486 		if ((ch = getc(file_fd)) != EOF) {
487 			/*
488 			 * This next file has content, therefore current
489 			 * file doesn't contains the last line.
490 			 */
491 			ungetc(ch, file_fd);
492 			fclose(file_fd);
493 			return (1);
494 		}
495 		fclose(file_fd);
496 	}
497 	return (0);
498 }
499 
500 int
501 lastline(void)
502 {
503 	int ch;
504 
505 	if (feof(infile)) {
506 		return !(
507 		    (inplace == NULL) &&
508 		    next_files_have_lines());
509 	}
510 	if ((ch = getc(infile)) == EOF) {
511 		return !(
512 		    (inplace == NULL) &&
513 		    next_files_have_lines());
514 	}
515 	ungetc(ch, infile);
516 	return (0);
517 }
518