main.c revision 1.36 
1/*	$NetBSD: main.c,v 1.36 2020/05/15 22:39:54 christos Exp $	*/
2
3/*-
4 * Copyright (c) 2013 Johann 'Myrkraverk' Oskarsson.
5 * Copyright (c) 1992 Diomidis Spinellis.
6 * Copyright (c) 1992, 1993
7 *	The Regents of the University of California.  All rights reserved.
8 *
9 * This code is derived from software contributed to Berkeley by
10 * Diomidis Spinellis of Imperial College, University of London.
11 *
12 * Redistribution and use in source and binary forms, with or without
13 * modification, are permitted provided that the following conditions
14 * are met:
15 * 1. Redistributions of source code must retain the above copyright
16 *    notice, this list of conditions and the following disclaimer.
17 * 2. Redistributions in binary form must reproduce the above copyright
18 *    notice, this list of conditions and the following disclaimer in the
19 *    documentation and/or other materials provided with the distribution.
20 * 3. Neither the name of the University nor the names of its contributors
21 *    may be used to endorse or promote products derived from this software
22 *    without specific prior written permission.
23 *
24 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
25 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
26 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
27 * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
28 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
29 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
30 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
31 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
32 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
33 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
34 * SUCH DAMAGE.
35 */
36
37#if HAVE_NBTOOL_CONFIG_H
38#include "nbtool_config.h"
39#endif
40
41#include <sys/cdefs.h>
42__RCSID("$NetBSD: main.c,v 1.36 2020/05/15 22:39:54 christos Exp $");
43#ifdef __FBSDID
44__FBSDID("$FreeBSD: head/usr.bin/sed/main.c 252231 2013-06-26 04:14:19Z pfg $");
45#endif
46
47#ifndef lint
48__COPYRIGHT("@(#) Copyright (c) 1992, 1993\
49	The Regents of the University of California.  All rights reserved.");
50#endif
51
52#if 0
53static const char sccsid[] = "@(#)main.c	8.2 (Berkeley) 1/3/94";
54#endif
55
56#include <sys/types.h>
57#include <sys/mman.h>
58#include <sys/param.h>
59#include <sys/stat.h>
60
61#include <err.h>
62#include <errno.h>
63#include <fcntl.h>
64#include <libgen.h>
65#include <limits.h>
66#include <locale.h>
67#include <regex.h>
68#include <stddef.h>
69#define _WITH_GETLINE
70#include <stdio.h>
71#include <stdlib.h>
72#include <string.h>
73#include <unistd.h>
74
75#include "defs.h"
76#include "extern.h"
77
78/*
79 * Linked list of units (strings and files) to be compiled
80 */
81struct s_compunit {
82	struct s_compunit *next;
83	enum e_cut {CU_FILE, CU_STRING} type;
84	char *s;			/* Pointer to string or fname */
85};
86
87/*
88 * Linked list pointer to compilation units and pointer to current
89 * next pointer.
90 */
91static struct s_compunit *script, **cu_nextp = &script;
92
93/*
94 * Linked list of files to be processed
95 */
96struct s_flist {
97	char *fname;
98	struct s_flist *next;
99};
100
101/*
102 * Linked list pointer to files and pointer to current
103 * next pointer.
104 */
105static struct s_flist *files, **fl_nextp = &files;
106
107FILE *infile;			/* Current input file */
108FILE *outfile;			/* Current output file */
109
110int aflag, eflag, nflag;
111int rflags = 0;
112static int rval;		/* Exit status */
113
114static int ispan;		/* Whether inplace editing spans across files */
115
116/*
117 * Current file and line number; line numbers restart across compilation
118 * units, but span across input files.  The latter is optional if editing
119 * in place.
120 */
121const char *fname;		/* File name. */
122const char *outfname;		/* Output file name */
123static char oldfname[PATH_MAX];	/* Old file name (for in-place editing) */
124static char tmpfname[PATH_MAX];	/* Temporary file name (for in-place editing) */
125static const char *inplace;	/* Inplace edit file extension. */
126u_long linenum;
127
128static void add_compunit(enum e_cut, char *);
129static void add_file(char *);
130static void usage(void) __dead;
131
132int
133main(int argc, char *argv[])
134{
135	int c, fflag;
136	char *temp_arg;
137
138	setprogname(argv[0]);
139	(void) setlocale(LC_ALL, "");
140
141	fflag = 0;
142	inplace = NULL;
143
144	while ((c = getopt(argc, argv, "EI::ae:f:i::lnru")) != -1)
145		switch (c) {
146		case 'r':		/* Gnu sed compat */
147		case 'E':
148			rflags = REG_EXTENDED;
149			break;
150		case 'I':
151			inplace = optarg ? optarg : __UNCONST("");
152			ispan = 1;	/* span across input files */
153			break;
154		case 'a':
155			aflag = 1;
156			break;
157		case 'e':
158			eflag = 1;
159			temp_arg = xmalloc(strlen(optarg) + 2);
160			strcpy(temp_arg, optarg);
161			strcat(temp_arg, "\n");
162			add_compunit(CU_STRING, temp_arg);
163			break;
164		case 'f':
165			fflag = 1;
166			add_compunit(CU_FILE, optarg);
167			break;
168		case 'i':
169			inplace = optarg ? optarg : __UNCONST("");
170			ispan = 0;	/* don't span across input files */
171			break;
172		case 'l':
173#ifdef _IOLBF
174			c = setvbuf(stdout, NULL, _IOLBF, 0);
175#else
176			c = setlinebuf(stdout);
177#endif
178			if (c)
179				warn("setting line buffered output failed");
180			break;
181		case 'n':
182			nflag = 1;
183			break;
184		case 'u':
185#ifdef _IONBF
186			c = setvbuf(stdout, NULL, _IONBF, 0);
187#else
188			c = -1;
189			errno = EOPNOTSUPP;
190#endif
191			if (c)
192				warn("setting unbuffered output failed");
193			break;
194		default:
195		case '?':
196			usage();
197		}
198	argc -= optind;
199	argv += optind;
200
201	/* First usage case; script is the first arg */
202	if (!eflag && !fflag && *argv) {
203		add_compunit(CU_STRING, *argv);
204		argv++;
205	}
206
207	compile();
208
209	/* Continue with first and start second usage */
210	if (*argv)
211		for (; *argv; argv++)
212			add_file(*argv);
213	else
214		add_file(NULL);
215	process();
216	cfclose(prog, NULL);
217	if (fclose(stdout))
218		err(1, "stdout");
219	exit(rval);
220}
221
222static void
223usage(void)
224{
225	(void)fprintf(stderr,
226	    "Usage:  %s [-aElnru] command [file ...]\n"
227	    "\t%s [-aElnru] [-e command] [-f command_file] [-I[extension]]\n"
228	    "\t    [-i[extension]] [file ...]\n", getprogname(), getprogname());
229	exit(1);
230}
231
232/*
233 * Like fgets, but go through the chain of compilation units chaining them
234 * together.  Empty strings and files are ignored.
235 */
236char *
237cu_fgets(char *buf, int n, int *more)
238{
239	static enum {ST_EOF, ST_FILE, ST_STRING} state = ST_EOF;
240	static FILE *f;		/* Current open file */
241	static char *s;		/* Current pointer inside string */
242	static char string_ident[30];
243	char *p;
244
245again:
246	switch (state) {
247	case ST_EOF:
248		if (script == NULL) {
249			if (more != NULL)
250				*more = 0;
251			return (NULL);
252		}
253		linenum = 0;
254		switch (script->type) {
255		case CU_FILE:
256			if ((f = fopen(script->s, "r")) == NULL)
257				err(1, "%s", script->s);
258			fname = script->s;
259			state = ST_FILE;
260			goto again;
261		case CU_STRING:
262			if (((size_t)snprintf(string_ident,
263			    sizeof(string_ident), "\"%s\"", script->s)) >=
264			    sizeof(string_ident) - 1)
265				(void)strcpy(string_ident +
266				    sizeof(string_ident) - 6, " ...\"");
267			fname = string_ident;
268			s = script->s;
269			state = ST_STRING;
270			goto again;
271		default:
272			abort();
273		}
274	case ST_FILE:
275		if ((p = fgets(buf, n, f)) != NULL) {
276			linenum++;
277			if (linenum == 1 && buf[0] == '#' && buf[1] == 'n')
278				nflag = 1;
279			if (more != NULL)
280				*more = !feof(f);
281			return (p);
282		}
283		script = script->next;
284		(void)fclose(f);
285		state = ST_EOF;
286		goto again;
287	case ST_STRING:
288		if (linenum == 0 && s[0] == '#' && s[1] == 'n')
289			nflag = 1;
290		p = buf;
291		for (;;) {
292			if (n-- <= 1) {
293				*p = '\0';
294				linenum++;
295				if (more != NULL)
296					*more = 1;
297				return (buf);
298			}
299			switch (*s) {
300			case '\0':
301				state = ST_EOF;
302				if (s == script->s) {
303					script = script->next;
304					goto again;
305				} else {
306					script = script->next;
307					*p = '\0';
308					linenum++;
309					if (more != NULL)
310						*more = 0;
311					return (buf);
312				}
313			case '\n':
314				*p++ = '\n';
315				*p = '\0';
316				s++;
317				linenum++;
318				if (more != NULL)
319					*more = 0;
320				return (buf);
321			default:
322				*p++ = *s++;
323			}
324		}
325	}
326	/* NOTREACHED */
327	return (NULL);
328}
329
330/*
331 * Like fgets, but go through the list of files chaining them together.
332 * Set len to the length of the line.
333 */
334int
335mf_fgets(SPACE *sp, enum e_spflag spflag)
336{
337	struct stat sb;
338	size_t len;
339	static char *p = NULL;
340	static size_t plen = 0;
341	int c;
342	static int firstfile;
343
344	if (infile == NULL) {
345		/* stdin? */
346		if (files->fname == NULL) {
347			if (inplace != NULL)
348				errx(1, "-I or -i may not be used with stdin");
349			infile = stdin;
350			fname = "stdin";
351			outfile = stdout;
352			outfname = "stdout";
353		}
354		firstfile = 1;
355	}
356
357	for (;;) {
358		if (infile != NULL && (c = getc(infile)) != EOF) {
359			(void)ungetc(c, infile);
360			break;
361		}
362		/* If we are here then either eof or no files are open yet */
363		if (infile == stdin) {
364			sp->len = 0;
365			return (0);
366		}
367		if (infile != NULL) {
368			fclose(infile);
369			if (*oldfname != '\0') {
370				/* if there was a backup file, remove it */
371				unlink(oldfname);
372				/*
373				 * Backup the original.  Note that hard links
374				 * are not supported on all filesystems.
375				 */
376				if ((link(fname, oldfname) != 0) &&
377				   (rename(fname, oldfname) != 0)) {
378					warn("rename()");
379					if (*tmpfname)
380						unlink(tmpfname);
381					exit(1);
382				}
383				*oldfname = '\0';
384			}
385			if (*tmpfname != '\0') {
386				if (outfile != NULL && outfile != stdout)
387					if (fclose(outfile) != 0) {
388						warn("fclose()");
389						unlink(tmpfname);
390						exit(1);
391					}
392				outfile = NULL;
393				if (rename(tmpfname, fname) != 0) {
394					/* this should not happen really! */
395					warn("rename()");
396					unlink(tmpfname);
397					exit(1);
398				}
399				*tmpfname = '\0';
400			}
401			outfname = NULL;
402		}
403		if (firstfile == 0)
404			files = files->next;
405		else
406			firstfile = 0;
407		if (files == NULL) {
408			sp->len = 0;
409			return (0);
410		}
411		fname = files->fname;
412		if (inplace != NULL) {
413			if (lstat(fname, &sb) != 0)
414				err(1, "%s", fname);
415			if (!(sb.st_mode & S_IFREG))
416				errx(1, "%s: %s %s", fname,
417				    "in-place editing only",
418				    "works for regular files");
419			if (*inplace != '\0') {
420				strlcpy(oldfname, fname,
421				    sizeof(oldfname));
422				len = strlcat(oldfname, inplace,
423				    sizeof(oldfname));
424				if (len > sizeof(oldfname))
425					errx(1, "%s: name too long", fname);
426			}
427			char d_name[PATH_MAX], f_name[PATH_MAX];
428			(void)strlcpy(d_name, fname, sizeof(d_name));
429			(void)strlcpy(f_name, fname, sizeof(f_name));
430			len = (size_t)snprintf(tmpfname, sizeof(tmpfname),
431			    "%s/.!%ld!%s", dirname(d_name), (long)getpid(),
432			    basename(f_name));
433			if (len >= sizeof(tmpfname))
434				errx(1, "%s: name too long", fname);
435			unlink(tmpfname);
436			if (outfile != NULL && outfile != stdout)
437				fclose(outfile);
438			if ((outfile = fopen(tmpfname, "w")) == NULL)
439				err(1, "%s", fname);
440			fchown(fileno(outfile), sb.st_uid, sb.st_gid);
441			fchmod(fileno(outfile), sb.st_mode & ALLPERMS);
442			outfname = tmpfname;
443			if (!ispan) {
444				linenum = 0;
445				resetstate();
446			}
447		} else {
448			outfile = stdout;
449			outfname = "stdout";
450		}
451		if ((infile = fopen(fname, "r")) == NULL) {
452			warn("%s", fname);
453			rval = 1;
454			continue;
455		}
456	}
457	/*
458	 * We are here only when infile is open and we still have something
459	 * to read from it.
460	 *
461	 * Use getline() so that we can handle essentially infinite input
462	 * data.  The p and plen are static so each invocation gives
463	 * getline() the same buffer which is expanded as needed.
464	 */
465	ssize_t slen = getline(&p, &plen, infile);
466	if (slen == -1)
467		err(1, "%s", fname);
468	if (slen != 0 && p[slen - 1] == '\n') {
469		sp->append_newline = 1;
470		slen--;
471	} else if (!lastline()) {
472		sp->append_newline = 1;
473	} else {
474		sp->append_newline = 0;
475	}
476	cspace(sp, p, (size_t)slen, spflag);
477
478	linenum++;
479
480	return (1);
481}
482
483/*
484 * Add a compilation unit to the linked list
485 */
486static void
487add_compunit(enum e_cut type, char *s)
488{
489	struct s_compunit *cu;
490
491	cu = xmalloc(sizeof(struct s_compunit));
492	cu->type = type;
493	cu->s = s;
494	cu->next = NULL;
495	*cu_nextp = cu;
496	cu_nextp = &cu->next;
497}
498
499/*
500 * Add a file to the linked list
501 */
502static void
503add_file(char *s)
504{
505	struct s_flist *fp;
506
507	fp = xmalloc(sizeof(struct s_flist));
508	fp->next = NULL;
509	*fl_nextp = fp;
510	fp->fname = s;
511	fl_nextp = &fp->next;
512}
513
514static int
515next_files_have_lines(void)
516{
517	struct s_flist *file;
518	FILE *file_fd;
519	int ch;
520
521	file = files;
522	while ((file = file->next) != NULL) {
523		if ((file_fd = fopen(file->fname, "r")) == NULL)
524			continue;
525
526		if ((ch = getc(file_fd)) != EOF) {
527			/*
528			 * This next file has content, therefore current
529			 * file doesn't contains the last line.
530			 */
531			ungetc(ch, file_fd);
532			fclose(file_fd);
533			return (1);
534		}
535
536		fclose(file_fd);
537	}
538
539	return (0);
540}
541
542int
543lastline(void)
544{
545	int ch;
546
547	if (feof(infile)) {
548		return !(
549		    (inplace == NULL || ispan) &&
550		    next_files_have_lines());
551	}
552	if ((ch = getc(infile)) == EOF) {
553		return !(
554		    (inplace == NULL || ispan) &&
555		    next_files_have_lines());
556	}
557	ungetc(ch, infile);
558	return (0);
559}
560