1/*	$NetBSD: main.c,v 1.38 2021/03/11 15:45:55 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 1.38 2021/03/11 15:45:55 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, "EGI::ae:f:gi::lnru")) != -1)
145		switch (c) {
146		case 'r':		/* Gnu sed compat */
147		case 'E':
148			rflags |= REG_EXTENDED;
149			break;
150		case 'G':
151			rflags &= ~REG_GNU;
152			break;
153		case 'I':
154			inplace = optarg ? optarg : __UNCONST("");
155			ispan = 1;	/* span across input files */
156			break;
157		case 'a':
158			aflag = 1;
159			break;
160		case 'e':
161			eflag = 1;
162			temp_arg = xmalloc(strlen(optarg) + 2);
163			strcpy(temp_arg, optarg);
164			strcat(temp_arg, "\n");
165			add_compunit(CU_STRING, temp_arg);
166			break;
167		case 'f':
168			fflag = 1;
169			add_compunit(CU_FILE, optarg);
170			break;
171		case 'g':
172			rflags |= REG_GNU;
173			break;
174		case 'i':
175			inplace = optarg ? optarg : __UNCONST("");
176			ispan = 0;	/* don't span across input files */
177			break;
178		case 'l':
179#ifdef _IOLBF
180			c = setvbuf(stdout, NULL, _IOLBF, 0);
181#else
182			c = setlinebuf(stdout);
183#endif
184			if (c)
185				warn("setting line buffered output failed");
186			break;
187		case 'n':
188			nflag = 1;
189			break;
190		case 'u':
191#ifdef _IONBF
192			c = setvbuf(stdout, NULL, _IONBF, 0);
193#else
194			c = -1;
195			errno = EOPNOTSUPP;
196#endif
197			if (c)
198				warn("setting unbuffered output failed");
199			break;
200		default:
201		case '?':
202			usage();
203		}
204	argc -= optind;
205	argv += optind;
206
207	/* First usage case; script is the first arg */
208	if (!eflag && !fflag && *argv) {
209		add_compunit(CU_STRING, *argv);
210		argv++;
211	}
212
213	compile();
214
215	/* Continue with first and start second usage */
216	if (*argv)
217		for (; *argv; argv++)
218			add_file(*argv);
219	else
220		add_file(NULL);
221	process();
222	cfclose(prog, NULL);
223	if (fclose(stdout))
224		err(1, "stdout");
225	exit(rval);
226}
227
228static void
229usage(void)
230{
231	(void)fprintf(stderr,
232	    "Usage:  %s [-aEGglnru] command [file ...]\n"
233	    "\t%s [-aEGglnru] [-e command] [-f command_file] [-I[extension]]\n"
234	    "\t    [-i[extension]] [file ...]\n", getprogname(), getprogname());
235	exit(1);
236}
237
238/*
239 * Like fgets, but go through the chain of compilation units chaining them
240 * together.  Empty strings and files are ignored.
241 */
242char *
243cu_fgets(char *buf, int n, int *more)
244{
245	static enum {ST_EOF, ST_FILE, ST_STRING} state = ST_EOF;
246	static FILE *f;		/* Current open file */
247	static char *s;		/* Current pointer inside string */
248	static char string_ident[30];
249	char *p;
250
251again:
252	switch (state) {
253	case ST_EOF:
254		if (script == NULL) {
255			if (more != NULL)
256				*more = 0;
257			return (NULL);
258		}
259		linenum = 0;
260		switch (script->type) {
261		case CU_FILE:
262			if ((f = fopen(script->s, "r")) == NULL)
263				err(1, "%s", script->s);
264			fname = script->s;
265			state = ST_FILE;
266			goto again;
267		case CU_STRING:
268			if (((size_t)snprintf(string_ident,
269			    sizeof(string_ident), "\"%s\"", script->s)) >=
270			    sizeof(string_ident) - 1)
271				(void)strcpy(string_ident +
272				    sizeof(string_ident) - 6, " ...\"");
273			fname = string_ident;
274			s = script->s;
275			state = ST_STRING;
276			goto again;
277		default:
278			abort();
279		}
280	case ST_FILE:
281		if ((p = fgets(buf, n, f)) != NULL) {
282			linenum++;
283			if (linenum == 1 && buf[0] == '#' && buf[1] == 'n')
284				nflag = 1;
285			if (more != NULL)
286				*more = !feof(f);
287			return (p);
288		}
289		script = script->next;
290		(void)fclose(f);
291		state = ST_EOF;
292		goto again;
293	case ST_STRING:
294		if (linenum == 0 && s[0] == '#' && s[1] == 'n')
295			nflag = 1;
296		p = buf;
297		for (;;) {
298			if (n-- <= 1) {
299				*p = '\0';
300				linenum++;
301				if (more != NULL)
302					*more = 1;
303				return (buf);
304			}
305			switch (*s) {
306			case '\0':
307				state = ST_EOF;
308				if (s == script->s) {
309					script = script->next;
310					goto again;
311				} else {
312					script = script->next;
313					*p = '\0';
314					linenum++;
315					if (more != NULL)
316						*more = 0;
317					return (buf);
318				}
319			case '\n':
320				*p++ = '\n';
321				*p = '\0';
322				s++;
323				linenum++;
324				if (more != NULL)
325					*more = 0;
326				return (buf);
327			default:
328				*p++ = *s++;
329			}
330		}
331	}
332	/* NOTREACHED */
333	return (NULL);
334}
335
336/*
337 * Like fgets, but go through the list of files chaining them together.
338 * Set len to the length of the line.
339 */
340int
341mf_fgets(SPACE *sp, enum e_spflag spflag)
342{
343	struct stat sb;
344	size_t len;
345	static char *p = NULL;
346	static size_t plen = 0;
347	int c;
348	static int firstfile;
349
350	if (infile == NULL) {
351		/* stdin? */
352		if (files->fname == NULL) {
353			if (inplace != NULL)
354				errx(1, "-I or -i may not be used with stdin");
355			infile = stdin;
356			fname = "stdin";
357			outfile = stdout;
358			outfname = "stdout";
359		}
360		firstfile = 1;
361	}
362
363	for (;;) {
364		if (infile != NULL && (c = getc(infile)) != EOF) {
365			(void)ungetc(c, infile);
366			break;
367		}
368		/* If we are here then either eof or no files are open yet */
369		if (infile == stdin) {
370			sp->len = 0;
371			return (0);
372		}
373		if (infile != NULL) {
374			fclose(infile);
375			if (*oldfname != '\0') {
376				/* if there was a backup file, remove it */
377				unlink(oldfname);
378				/*
379				 * Backup the original.  Note that hard links
380				 * are not supported on all filesystems.
381				 */
382				if ((link(fname, oldfname) != 0) &&
383				   (rename(fname, oldfname) != 0)) {
384					warn("rename()");
385					if (*tmpfname)
386						unlink(tmpfname);
387					exit(1);
388				}
389				*oldfname = '\0';
390			}
391			if (*tmpfname != '\0') {
392				if (outfile != NULL && outfile != stdout)
393					if (fclose(outfile) != 0) {
394						warn("fclose()");
395						unlink(tmpfname);
396						exit(1);
397					}
398				outfile = NULL;
399				if (rename(tmpfname, fname) != 0) {
400					/* this should not happen really! */
401					warn("rename()");
402					unlink(tmpfname);
403					exit(1);
404				}
405				*tmpfname = '\0';
406			}
407			outfname = NULL;
408		}
409		if (firstfile == 0)
410			files = files->next;
411		else
412			firstfile = 0;
413		if (files == NULL) {
414			sp->len = 0;
415			return (0);
416		}
417		fname = files->fname;
418		if (inplace != NULL) {
419			if (lstat(fname, &sb) != 0)
420				err(1, "%s", fname);
421			if (!(sb.st_mode & S_IFREG))
422				errx(1, "%s: %s %s", fname,
423				    "in-place editing only",
424				    "works for regular files");
425			if (*inplace != '\0') {
426				strlcpy(oldfname, fname,
427				    sizeof(oldfname));
428				len = strlcat(oldfname, inplace,
429				    sizeof(oldfname));
430				if (len > sizeof(oldfname))
431					errx(1, "%s: name too long", fname);
432			}
433			char d_name[PATH_MAX], f_name[PATH_MAX];
434			(void)strlcpy(d_name, fname, sizeof(d_name));
435			(void)strlcpy(f_name, fname, sizeof(f_name));
436			len = (size_t)snprintf(tmpfname, sizeof(tmpfname),
437			    "%s/.!%ld!%s", dirname(d_name), (long)getpid(),
438			    basename(f_name));
439			if (len >= sizeof(tmpfname))
440				errx(1, "%s: name too long", fname);
441			unlink(tmpfname);
442			if (outfile != NULL && outfile != stdout)
443				fclose(outfile);
444			if ((outfile = fopen(tmpfname, "w")) == NULL)
445				err(1, "%s", fname);
446			fchown(fileno(outfile), sb.st_uid, sb.st_gid);
447			fchmod(fileno(outfile), sb.st_mode & ALLPERMS);
448			outfname = tmpfname;
449			if (!ispan) {
450				linenum = 0;
451				resetstate();
452			}
453		} else {
454			outfile = stdout;
455			outfname = "stdout";
456		}
457		if ((infile = fopen(fname, "r")) == NULL) {
458			warn("%s", fname);
459			rval = 1;
460			continue;
461		}
462	}
463	/*
464	 * We are here only when infile is open and we still have something
465	 * to read from it.
466	 *
467	 * Use getline() so that we can handle essentially infinite input
468	 * data.  The p and plen are static so each invocation gives
469	 * getline() the same buffer which is expanded as needed.
470	 */
471	ssize_t slen = getline(&p, &plen, infile);
472	if (slen == -1)
473		err(1, "%s", fname);
474	if (slen != 0 && p[slen - 1] == '\n') {
475		sp->append_newline = 1;
476		slen--;
477	} else if (!lastline()) {
478		sp->append_newline = 1;
479	} else {
480		sp->append_newline = 0;
481	}
482	cspace(sp, p, (size_t)slen, spflag);
483
484	linenum++;
485
486	return (1);
487}
488
489/*
490 * Add a compilation unit to the linked list
491 */
492static void
493add_compunit(enum e_cut type, char *s)
494{
495	struct s_compunit *cu;
496
497	cu = xmalloc(sizeof(struct s_compunit));
498	cu->type = type;
499	cu->s = s;
500	cu->next = NULL;
501	*cu_nextp = cu;
502	cu_nextp = &cu->next;
503}
504
505/*
506 * Add a file to the linked list
507 */
508static void
509add_file(char *s)
510{
511	struct s_flist *fp;
512
513	fp = xmalloc(sizeof(struct s_flist));
514	fp->next = NULL;
515	*fl_nextp = fp;
516	fp->fname = s;
517	fl_nextp = &fp->next;
518}
519
520static int
521next_files_have_lines(void)
522{
523	struct s_flist *file;
524	FILE *file_fd;
525	int ch;
526
527	file = files;
528	while ((file = file->next) != NULL) {
529		if ((file_fd = fopen(file->fname, "r")) == NULL)
530			continue;
531
532		if ((ch = getc(file_fd)) != EOF) {
533			/*
534			 * This next file has content, therefore current
535			 * file doesn't contains the last line.
536			 */
537			ungetc(ch, file_fd);
538			fclose(file_fd);
539			return (1);
540		}
541
542		fclose(file_fd);
543	}
544
545	return (0);
546}
547
548int
549lastline(void)
550{
551	int ch;
552
553	if (feof(infile)) {
554		return !(
555		    (inplace == NULL || ispan) &&
556		    next_files_have_lines());
557	}
558	if ((ch = getc(infile)) == EOF) {
559		return !(
560		    (inplace == NULL || ispan) &&
561		    next_files_have_lines());
562	}
563	ungetc(ch, infile);
564	return (0);
565}
566