/usr.bin/tar/util.c

  1/*-
  2 * Copyright (c) 2003-2007 Tim Kientzle
  3 * All rights reserved.
  4 *
  5 * Redistribution and use in source and binary forms, with or without
  6 * modification, are permitted provided that the following conditions
  7 * are met:
  8 * 1. Redistributions of source code must retain the above copyright
  9 *    notice, this list of conditions and the following disclaimer.
 10 * 2. Redistributions in binary form must reproduce the above copyright
 11 *    notice, this list of conditions and the following disclaimer in the
 12 *    documentation and/or other materials provided with the distribution.
 13 *
 14 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR(S) ``AS IS'' AND ANY EXPRESS OR
 15 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
 16 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
 17 * IN NO EVENT SHALL THE AUTHOR(S) BE LIABLE FOR ANY DIRECT, INDIRECT,
 18 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
 19 * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
 20 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
 21 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 22 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
 23 * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 24 */
 25
 26#include "bsdtar_platform.h"
 27__FBSDID("$FreeBSD$");
 28
 29#ifdef HAVE_SYS_STAT_H
 30#include <sys/stat.h>
 31#endif
 32#ifdef HAVE_SYS_TYPES_H
 33#include <sys/types.h>  /* Linux doesn't define mode_t, etc. in sys/stat.h. */
 34#endif
 35#include <ctype.h>
 36#ifdef HAVE_ERRNO_H
 37#include <errno.h>
 38#endif
 39#ifdef HAVE_IO_H
 40#include <io.h>
 41#endif
 42#ifdef HAVE_STDARG_H
 43#include <stdarg.h>
 44#endif
 45#ifdef HAVE_STDINT_H
 46#include <stdint.h>
 47#endif
 48#include <stdio.h>
 49#ifdef HAVE_STDLIB_H
 50#include <stdlib.h>
 51#endif
 52#ifdef HAVE_STRING_H
 53#include <string.h>
 54#endif
 55#ifdef HAVE_WCTYPE_H
 56#include <wctype.h>
 57#else
 58/* If we don't have wctype, we need to hack up some version of iswprint(). */
 59#define iswprint isprint
 60#endif
 61
 62#include "bsdtar.h"
 63#include "err.h"
 64
 65static size_t	bsdtar_expand_char(char *, size_t, char);
 66static const char *strip_components(const char *path, int elements);
 67
 68#if defined(_WIN32) && !defined(__CYGWIN__)
 69#define read _read
 70#endif
 71
 72/* TODO:  Hack up a version of mbtowc for platforms with no wide
 73 * character support at all.  I think the following might suffice,
 74 * but it needs careful testing.
 75 * #if !HAVE_MBTOWC
 76 * #define mbtowc(wcp, p, n) ((*wcp = *p), 1)
 77 * #endif
 78 */
 79
 80/*
 81 * Print a string, taking care with any non-printable characters.
 82 *
 83 * Note that we use a stack-allocated buffer to receive the formatted
 84 * string if we can.  This is partly performance (avoiding a call to
 85 * malloc()), partly out of expedience (we have to call vsnprintf()
 86 * before malloc() anyway to find out how big a buffer we need; we may
 87 * as well point that first call at a small local buffer in case it
 88 * works), but mostly for safety (so we can use this to print messages
 89 * about out-of-memory conditions).
 90 */
 91
 92void
 93safe_fprintf(FILE *f, const char *fmt, ...)
 94{
 95	char fmtbuff_stack[256]; /* Place to format the printf() string. */
 96	char outbuff[256]; /* Buffer for outgoing characters. */
 97	char *fmtbuff_heap; /* If fmtbuff_stack is too small, we use malloc */
 98	char *fmtbuff;  /* Pointer to fmtbuff_stack or fmtbuff_heap. */
 99	int fmtbuff_length;
100	int length, n;
101	va_list ap;
102	const char *p;
103	unsigned i;
104	wchar_t wc;
105	char try_wc;
106
107	/* Use a stack-allocated buffer if we can, for speed and safety. */
108	fmtbuff_heap = NULL;
109	fmtbuff_length = sizeof(fmtbuff_stack);
110	fmtbuff = fmtbuff_stack;
111
112	/* Try formatting into the stack buffer. */
113	va_start(ap, fmt);
114	length = vsnprintf(fmtbuff, fmtbuff_length, fmt, ap);
115	va_end(ap);
116
117	/* If the result was too large, allocate a buffer on the heap. */
118	if (length >= fmtbuff_length) {
119		fmtbuff_length = length+1;
120		fmtbuff_heap = malloc(fmtbuff_length);
121
122		/* Reformat the result into the heap buffer if we can. */
123		if (fmtbuff_heap != NULL) {
124			fmtbuff = fmtbuff_heap;
125			va_start(ap, fmt);
126			length = vsnprintf(fmtbuff, fmtbuff_length, fmt, ap);
127			va_end(ap);
128		} else {
129			/* Leave fmtbuff pointing to the truncated
130			 * string in fmtbuff_stack. */
131			length = sizeof(fmtbuff_stack) - 1;
132		}
133	}
134
135	/* Note: mbrtowc() has a cleaner API, but mbtowc() seems a bit
136	 * more portable, so we use that here instead. */
137	n = mbtowc(NULL, NULL, 1); /* Reset the shift state. */
138
139	/* Write data, expanding unprintable characters. */
140	p = fmtbuff;
141	i = 0;
142	try_wc = 1;
143	while (*p != '\0') {
144
145		/* Convert to wide char, test if the wide
146		 * char is printable in the current locale. */
147		if (try_wc && (n = mbtowc(&wc, p, length)) != -1) {
148			length -= n;
149			if (iswprint(wc) && wc != L'\\') {
150				/* Printable, copy the bytes through. */
151				while (n-- > 0)
152					outbuff[i++] = *p++;
153			} else {
154				/* Not printable, format the bytes. */
155				while (n-- > 0)
156					i += (unsigned)bsdtar_expand_char(
157					    outbuff, i, *p++);
158			}
159		} else {
160			/* After any conversion failure, don't bother
161			 * trying to convert the rest. */
162			i += (unsigned)bsdtar_expand_char(outbuff, i, *p++);
163			try_wc = 0;
164		}
165
166		/* If our output buffer is full, dump it and keep going. */
167		if (i > (sizeof(outbuff) - 20)) {
168			outbuff[i] = '\0';
169			fprintf(f, "%s", outbuff);
170			i = 0;
171		}
172	}
173	outbuff[i] = '\0';
174	fprintf(f, "%s", outbuff);
175
176	/* If we allocated a heap-based formatting buffer, free it now. */
177	if (fmtbuff_heap != NULL)
178		free(fmtbuff_heap);
179}
180
181/*
182 * Render an arbitrary sequence of bytes into printable ASCII characters.
183 */
184static size_t
185bsdtar_expand_char(char *buff, size_t offset, char c)
186{
187	size_t i = offset;
188
189	if (isprint((unsigned char)c) && c != '\\')
190		buff[i++] = c;
191	else {
192		buff[i++] = '\\';
193		switch (c) {
194		case '\a': buff[i++] = 'a'; break;
195		case '\b': buff[i++] = 'b'; break;
196		case '\f': buff[i++] = 'f'; break;
197		case '\n': buff[i++] = 'n'; break;
198#if '\r' != '\n'
199		/* On some platforms, \n and \r are the same. */
200		case '\r': buff[i++] = 'r'; break;
201#endif
202		case '\t': buff[i++] = 't'; break;
203		case '\v': buff[i++] = 'v'; break;
204		case '\\': buff[i++] = '\\'; break;
205		default:
206			sprintf(buff + i, "%03o", 0xFF & (int)c);
207			i += 3;
208		}
209	}
210
211	return (i - offset);
212}
213
214int
215yes(const char *fmt, ...)
216{
217	char buff[32];
218	char *p;
219	ssize_t l;
220
221	va_list ap;
222	va_start(ap, fmt);
223	vfprintf(stderr, fmt, ap);
224	va_end(ap);
225	fprintf(stderr, " (y/N)? ");
226	fflush(stderr);
227
228	l = read(2, buff, sizeof(buff) - 1);
229	if (l < 0) {
230	  fprintf(stderr, "Keyboard read failed\n");
231	  exit(1);
232	}
233	if (l == 0)
234		return (0);
235	buff[l] = 0;
236
237	for (p = buff; *p != '\0'; p++) {
238		if (isspace((unsigned char)*p))
239			continue;
240		switch(*p) {
241		case 'y': case 'Y':
242			return (1);
243		case 'n': case 'N':
244			return (0);
245		default:
246			return (0);
247		}
248	}
249
250	return (0);
251}
252
253/*-
254 * The logic here for -C <dir> attempts to avoid
255 * chdir() as long as possible.  For example:
256 * "-C /foo -C /bar file"          needs chdir("/bar") but not chdir("/foo")
257 * "-C /foo -C bar file"           needs chdir("/foo/bar")
258 * "-C /foo -C bar /file1"         does not need chdir()
259 * "-C /foo -C bar /file1 file2"   needs chdir("/foo/bar") before file2
260 *
261 * The only correct way to handle this is to record a "pending" chdir
262 * request and combine multiple requests intelligently until we
263 * need to process a non-absolute file.  set_chdir() adds the new dir
264 * to the pending list; do_chdir() actually executes any pending chdir.
265 *
266 * This way, programs that build tar command lines don't have to worry
267 * about -C with non-existent directories; such requests will only
268 * fail if the directory must be accessed.
269 *
270 * TODO: Make this handle Windows paths correctly.
271 */
272void
273set_chdir(struct bsdtar *bsdtar, const char *newdir)
274{
275	if (newdir[0] == '/') {
276		/* The -C /foo -C /bar case; dump first one. */
277		free(bsdtar->pending_chdir);
278		bsdtar->pending_chdir = NULL;
279	}
280	if (bsdtar->pending_chdir == NULL)
281		/* Easy case: no previously-saved dir. */
282		bsdtar->pending_chdir = strdup(newdir);
283	else {
284		/* The -C /foo -C bar case; concatenate */
285		char *old_pending = bsdtar->pending_chdir;
286		size_t old_len = strlen(old_pending);
287		bsdtar->pending_chdir = malloc(old_len + strlen(newdir) + 2);
288		if (old_pending[old_len - 1] == '/')
289			old_pending[old_len - 1] = '\0';
290		if (bsdtar->pending_chdir != NULL)
291			sprintf(bsdtar->pending_chdir, "%s/%s",
292			    old_pending, newdir);
293		free(old_pending);
294	}
295	if (bsdtar->pending_chdir == NULL)
296		lafe_errc(1, errno, "No memory");
297}
298
299void
300do_chdir(struct bsdtar *bsdtar)
301{
302	if (bsdtar->pending_chdir == NULL)
303		return;
304
305	if (chdir(bsdtar->pending_chdir) != 0) {
306		lafe_errc(1, 0, "could not chdir to '%s'\n",
307		    bsdtar->pending_chdir);
308	}
309	free(bsdtar->pending_chdir);
310	bsdtar->pending_chdir = NULL;
311}
312
313static const char *
314strip_components(const char *p, int elements)
315{
316	/* Skip as many elements as necessary. */
317	while (elements > 0) {
318		switch (*p++) {
319		case '/':
320#if defined(_WIN32) && !defined(__CYGWIN__)
321		case '\\': /* Support \ path sep on Windows ONLY. */
322#endif
323			elements--;
324			break;
325		case '\0':
326			/* Path is too short, skip it. */
327			return (NULL);
328		}
329	}
330
331	/* Skip any / characters.  This handles short paths that have
332	 * additional / termination.  This also handles the case where
333	 * the logic above stops in the middle of a duplicate //
334	 * sequence (which would otherwise get converted to an
335	 * absolute path). */
336	for (;;) {
337		switch (*p) {
338		case '/':
339#if defined(_WIN32) && !defined(__CYGWIN__)
340		case '\\': /* Support \ path sep on Windows ONLY. */
341#endif
342			++p;
343			break;
344		case '\0':
345			return (NULL);
346		default:
347			return (p);
348		}
349	}
350}
351
352/*
353 * Handle --strip-components and any future path-rewriting options.
354 * Returns non-zero if the pathname should not be extracted.
355 *
356 * TODO: Support pax-style regex path rewrites.
357 */
358int
359edit_pathname(struct bsdtar *bsdtar, struct archive_entry *entry)
360{
361	const char *name = archive_entry_pathname(entry);
362#if HAVE_REGEX_H
363	char *subst_name;
364	int r;
365#endif
366
367#if HAVE_REGEX_H
368	r = apply_substitution(bsdtar, name, &subst_name, 0);
369	if (r == -1) {
370		lafe_warnc(0, "Invalid substitution, skipping entry");
371		return 1;
372	}
373	if (r == 1) {
374		archive_entry_copy_pathname(entry, subst_name);
375		if (*subst_name == '\0') {
376			free(subst_name);
377			return -1;
378		} else
379			free(subst_name);
380		name = archive_entry_pathname(entry);
381	}
382
383	if (archive_entry_hardlink(entry)) {
384		r = apply_substitution(bsdtar, archive_entry_hardlink(entry), &subst_name, 1);
385		if (r == -1) {
386			lafe_warnc(0, "Invalid substitution, skipping entry");
387			return 1;
388		}
389		if (r == 1) {
390			archive_entry_copy_hardlink(entry, subst_name);
391			free(subst_name);
392		}
393	}
394	if (archive_entry_symlink(entry) != NULL) {
395		r = apply_substitution(bsdtar, archive_entry_symlink(entry), &subst_name, 1);
396		if (r == -1) {
397			lafe_warnc(0, "Invalid substitution, skipping entry");
398			return 1;
399		}
400		if (r == 1) {
401			archive_entry_copy_symlink(entry, subst_name);
402			free(subst_name);
403		}
404	}
405#endif
406
407	/* Strip leading dir names as per --strip-components option. */
408	if (bsdtar->strip_components > 0) {
409		const char *linkname = archive_entry_hardlink(entry);
410
411		name = strip_components(name, bsdtar->strip_components);
412		if (name == NULL)
413			return (1);
414
415		if (linkname != NULL) {
416			linkname = strip_components(linkname,
417			    bsdtar->strip_components);
418			if (linkname == NULL)
419				return (1);
420			archive_entry_copy_hardlink(entry, linkname);
421		}
422	}
423
424	/* By default, don't write or restore absolute pathnames. */
425	if (!bsdtar->option_absolute_paths) {
426		const char *rp, *p = name;
427		int slashonly = 1;
428
429		/* Remove leading "//./" or "//?/" or "//?/UNC/"
430		 * (absolute path prefixes used by Windows API) */
431		if ((p[0] == '/' || p[0] == '\\') &&
432		    (p[1] == '/' || p[1] == '\\') &&
433		    (p[2] == '.' || p[2] == '?') &&
434		    (p[3] == '/' || p[3] == '\\'))
435		{
436			if (p[2] == '?' &&
437			    (p[4] == 'U' || p[4] == 'u') &&
438			    (p[5] == 'N' || p[5] == 'n') &&
439			    (p[6] == 'C' || p[6] == 'c') &&
440			    (p[7] == '/' || p[7] == '\\'))
441				p += 8;
442			else
443				p += 4;
444			slashonly = 0;
445		}
446		do {
447			rp = p;
448			/* Remove leading drive letter from archives created
449			 * on Windows. */
450			if (((p[0] >= 'a' && p[0] <= 'z') ||
451			     (p[0] >= 'A' && p[0] <= 'Z')) &&
452				 p[1] == ':') {
453				p += 2;
454				slashonly = 0;
455			}
456			/* Remove leading "/../", "//", etc. */
457			while (p[0] == '/' || p[0] == '\\') {
458				if (p[1] == '.' && p[2] == '.' &&
459					(p[3] == '/' || p[3] == '\\')) {
460					p += 3; /* Remove "/..", leave "/"
461							 * for next pass. */
462					slashonly = 0;
463				} else
464					p += 1; /* Remove "/". */
465			}
466		} while (rp != p);
467
468		if (p != name && !bsdtar->warned_lead_slash) {
469			/* Generate a warning the first time this happens. */
470			if (slashonly)
471				lafe_warnc(0,
472				    "Removing leading '%c' from member names",
473				    name[0]);
474			else
475				lafe_warnc(0,
476				    "Removing leading drive letter from "
477				    "member names");
478			bsdtar->warned_lead_slash = 1;
479		}
480
481		/* Special case: Stripping everything yields ".". */
482		if (*p == '\0')
483			name = ".";
484		else
485			name = p;
486	} else {
487		/* Strip redundant leading '/' characters. */
488		while (name[0] == '/' && name[1] == '/')
489			name++;
490	}
491
492	/* Safely replace name in archive_entry. */
493	if (name != archive_entry_pathname(entry)) {
494		char *q = strdup(name);
495		archive_entry_copy_pathname(entry, q);
496		free(q);
497	}
498	return (0);
499}
500
501/*
502 * It would be nice to just use printf() for formatting large numbers,
503 * but the compatibility problems are quite a headache.  Hence the
504 * following simple utility function.
505 */
506const char *
507tar_i64toa(int64_t n0)
508{
509	static char buff[24];
510	int64_t n = n0 < 0 ? -n0 : n0;
511	char *p = buff + sizeof(buff);
512
513	*--p = '\0';
514	do {
515		*--p = '0' + (int)(n % 10);
516		n /= 10;
517	} while (n > 0);
518	if (n0 < 0)
519		*--p = '-';
520	return p;
521}
522
523/*
524 * Like strcmp(), but try to be a little more aware of the fact that
525 * we're comparing two paths.  Right now, it just handles leading
526 * "./" and trailing '/' specially, so that "a/b/" == "./a/b"
527 *
528 * TODO: Make this better, so that "./a//b/./c/" == "a/b/c"
529 * TODO: After this works, push it down into libarchive.
530 * TODO: Publish the path normalization routines in libarchive so
531 * that bsdtar can normalize paths and use fast strcmp() instead
532 * of this.
533 *
534 * Note: This is currently only used within write.c, so should
535 * not handle \ path separators.
536 */
537
538int
539pathcmp(const char *a, const char *b)
540{
541	/* Skip leading './' */
542	if (a[0] == '.' && a[1] == '/' && a[2] != '\0')
543		a += 2;
544	if (b[0] == '.' && b[1] == '/' && b[2] != '\0')
545		b += 2;
546	/* Find the first difference, or return (0) if none. */
547	while (*a == *b) {
548		if (*a == '\0')
549			return (0);
550		a++;
551		b++;
552	}
553	/*
554	 * If one ends in '/' and the other one doesn't,
555	 * they're the same.
556	 */
557	if (a[0] == '/' && a[1] == '\0' && b[0] == '\0')
558		return (0);
559	if (a[0] == '\0' && b[0] == '/' && b[1] == '\0')
560		return (0);
561	/* They're really different, return the correct sign. */
562	return (*(const unsigned char *)a - *(const unsigned char *)b);
563}