2
* Copyright (c) 2003-2007 Tim Kientzle
5
* Redistribution and use in source and binary forms, with or without
6
* modification, are permitted provided that the following conditions
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.
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.
26
#include "bsdtar_platform.h"
27
__FBSDID("$FreeBSD: src/usr.bin/tar/util.c,v 1.23 2008/12/15 06:00:25 kientzle Exp $");
29
#ifdef HAVE_SYS_STAT_H
32
#ifdef HAVE_SYS_TYPES_H
33
#include <sys/types.h> /* Linux doesn't define mode_t, etc. in sys/stat.h. */
58
/* If we don't have wctype, we need to hack up some version of iswprint(). */
59
#define iswprint isprint
65
static size_t bsdtar_expand_char(char *, size_t, char);
66
static const char *strip_components(const char *path, int elements);
68
#if defined(_WIN32) && !defined(__CYGWIN__)
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.
76
* #define mbtowc(wcp, p, n) ((*wcp = *p), 1)
81
* Print a string, taking care with any non-printable characters.
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).
93
safe_fprintf(FILE *f, const char *fmt, ...)
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. */
107
/* Use a stack-allocated buffer if we can, for speed and safety. */
109
fmtbuff_length = sizeof(fmtbuff_stack);
110
fmtbuff = fmtbuff_stack;
112
/* Try formatting into the stack buffer. */
114
length = vsnprintf(fmtbuff, fmtbuff_length, fmt, ap);
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);
122
/* Reformat the result into the heap buffer if we can. */
123
if (fmtbuff_heap != NULL) {
124
fmtbuff = fmtbuff_heap;
126
length = vsnprintf(fmtbuff, fmtbuff_length, fmt, ap);
129
/* Leave fmtbuff pointing to the truncated
130
* string in fmtbuff_stack. */
131
length = sizeof(fmtbuff_stack) - 1;
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. */
139
/* Write data, expanding unprintable characters. */
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) {
149
if (iswprint(wc) && wc != L'\\') {
150
/* Printable, copy the bytes through. */
154
/* Not printable, format the bytes. */
156
i += (unsigned)bsdtar_expand_char(
160
/* After any conversion failure, don't bother
161
* trying to convert the rest. */
162
i += (unsigned)bsdtar_expand_char(outbuff, i, *p++);
166
/* If our output buffer is full, dump it and keep going. */
167
if (i > (sizeof(outbuff) - 20)) {
169
fprintf(f, "%s", outbuff);
174
fprintf(f, "%s", outbuff);
176
/* If we allocated a heap-based formatting buffer, free it now. */
177
if (fmtbuff_heap != NULL)
182
* Render an arbitrary sequence of bytes into printable ASCII characters.
185
bsdtar_expand_char(char *buff, size_t offset, char c)
189
if (isprint((unsigned char)c) && 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;
199
/* On some platforms, \n and \r are the same. */
200
case '\r': buff[i++] = 'r'; break;
202
case '\t': buff[i++] = 't'; break;
203
case '\v': buff[i++] = 'v'; break;
204
case '\\': buff[i++] = '\\'; break;
206
sprintf(buff + i, "%03o", 0xFF & (int)c);
215
yes(const char *fmt, ...)
223
vfprintf(stderr, fmt, ap);
225
fprintf(stderr, " (y/N)? ");
228
l = read(2, buff, sizeof(buff) - 1);
233
for (p = buff; *p != '\0'; p++) {
234
if (isspace((unsigned char)*p))
250
* The logic here for -C <dir> attempts to avoid
251
* chdir() as long as possible. For example:
252
* "-C /foo -C /bar file" needs chdir("/bar") but not chdir("/foo")
253
* "-C /foo -C bar file" needs chdir("/foo/bar")
254
* "-C /foo -C bar /file1" does not need chdir()
255
* "-C /foo -C bar /file1 file2" needs chdir("/foo/bar") before file2
257
* The only correct way to handle this is to record a "pending" chdir
258
* request and combine multiple requests intelligently until we
259
* need to process a non-absolute file. set_chdir() adds the new dir
260
* to the pending list; do_chdir() actually executes any pending chdir.
262
* This way, programs that build tar command lines don't have to worry
263
* about -C with non-existent directories; such requests will only
264
* fail if the directory must be accessed.
266
* TODO: Make this handle Windows paths correctly.
269
set_chdir(struct bsdtar *bsdtar, const char *newdir)
271
if (newdir[0] == '/') {
272
/* The -C /foo -C /bar case; dump first one. */
273
free(bsdtar->pending_chdir);
274
bsdtar->pending_chdir = NULL;
276
if (bsdtar->pending_chdir == NULL)
277
/* Easy case: no previously-saved dir. */
278
bsdtar->pending_chdir = strdup(newdir);
280
/* The -C /foo -C bar case; concatenate */
281
char *old_pending = bsdtar->pending_chdir;
282
size_t old_len = strlen(old_pending);
283
bsdtar->pending_chdir = malloc(old_len + strlen(newdir) + 2);
284
if (old_pending[old_len - 1] == '/')
285
old_pending[old_len - 1] = '\0';
286
if (bsdtar->pending_chdir != NULL)
287
sprintf(bsdtar->pending_chdir, "%s/%s",
288
old_pending, newdir);
291
if (bsdtar->pending_chdir == NULL)
292
lafe_errc(1, errno, "No memory");
296
do_chdir(struct bsdtar *bsdtar)
298
if (bsdtar->pending_chdir == NULL)
301
if (chdir(bsdtar->pending_chdir) != 0) {
302
lafe_errc(1, 0, "could not chdir to '%s'\n",
303
bsdtar->pending_chdir);
305
free(bsdtar->pending_chdir);
306
bsdtar->pending_chdir = NULL;
310
strip_components(const char *p, int elements)
312
/* Skip as many elements as necessary. */
313
while (elements > 0) {
316
#if defined(_WIN32) && !defined(__CYGWIN__)
317
case '\\': /* Support \ path sep on Windows ONLY. */
322
/* Path is too short, skip it. */
327
/* Skip any / characters. This handles short paths that have
328
* additional / termination. This also handles the case where
329
* the logic above stops in the middle of a duplicate //
330
* sequence (which would otherwise get converted to an
335
#if defined(_WIN32) && !defined(__CYGWIN__)
336
case '\\': /* Support \ path sep on Windows ONLY. */
349
* Handle --strip-components and any future path-rewriting options.
350
* Returns non-zero if the pathname should not be extracted.
352
* TODO: Support pax-style regex path rewrites.
355
edit_pathname(struct bsdtar *bsdtar, struct archive_entry *entry)
357
const char *name = archive_entry_pathname(entry);
364
r = apply_substitution(bsdtar, name, &subst_name, 0);
366
lafe_warnc(0, "Invalid substitution, skipping entry");
370
archive_entry_copy_pathname(entry, subst_name);
371
if (*subst_name == '\0') {
376
name = archive_entry_pathname(entry);
379
if (archive_entry_hardlink(entry)) {
380
r = apply_substitution(bsdtar, archive_entry_hardlink(entry), &subst_name, 1);
382
lafe_warnc(0, "Invalid substitution, skipping entry");
386
archive_entry_copy_hardlink(entry, subst_name);
390
if (archive_entry_symlink(entry) != NULL) {
391
r = apply_substitution(bsdtar, archive_entry_symlink(entry), &subst_name, 1);
393
lafe_warnc(0, "Invalid substitution, skipping entry");
397
archive_entry_copy_symlink(entry, subst_name);
403
/* Strip leading dir names as per --strip-components option. */
404
if (bsdtar->strip_components > 0) {
405
const char *linkname = archive_entry_hardlink(entry);
407
name = strip_components(name, bsdtar->strip_components);
411
if (linkname != NULL) {
412
linkname = strip_components(linkname,
413
bsdtar->strip_components);
414
if (linkname == NULL)
416
archive_entry_copy_hardlink(entry, linkname);
420
/* By default, don't write or restore absolute pathnames. */
421
if (!bsdtar->option_absolute_paths) {
422
const char *rp, *p = name;
425
/* Remove leading "//./" or "//?/" or "//?/UNC/"
426
* (absolute path prefixes used by Windows API) */
427
if ((p[0] == '/' || p[0] == '\\') &&
428
(p[1] == '/' || p[1] == '\\') &&
429
(p[2] == '.' || p[2] == '?') &&
430
(p[3] == '/' || p[3] == '\\'))
433
(p[4] == 'U' || p[4] == 'u') &&
434
(p[5] == 'N' || p[5] == 'n') &&
435
(p[6] == 'C' || p[6] == 'c') &&
436
(p[7] == '/' || p[7] == '\\'))
444
/* Remove leading drive letter from archives created
446
if (((p[0] >= 'a' && p[0] <= 'z') ||
447
(p[0] >= 'A' && p[0] <= 'Z')) &&
452
/* Remove leading "/../", "//", etc. */
453
while (p[0] == '/' || p[0] == '\\') {
454
if (p[1] == '.' && p[2] == '.' &&
455
(p[3] == '/' || p[3] == '\\')) {
456
p += 3; /* Remove "/..", leave "/"
460
p += 1; /* Remove "/". */
464
if (p != name && !bsdtar->warned_lead_slash) {
465
/* Generate a warning the first time this happens. */
468
"Removing leading '%c' from member names",
472
"Removing leading drive letter from "
474
bsdtar->warned_lead_slash = 1;
477
/* Special case: Stripping everything yields ".". */
483
/* Strip redundant leading '/' characters. */
484
while (name[0] == '/' && name[1] == '/')
488
/* Safely replace name in archive_entry. */
489
if (name != archive_entry_pathname(entry)) {
490
char *q = strdup(name);
491
archive_entry_copy_pathname(entry, q);
498
* It would be nice to just use printf() for formatting large numbers,
499
* but the compatibility problems are quite a headache. Hence the
500
* following simple utility function.
503
tar_i64toa(int64_t n0)
505
static char buff[24];
506
int64_t n = n0 < 0 ? -n0 : n0;
507
char *p = buff + sizeof(buff);
511
*--p = '0' + (int)(n % 10);
520
* Like strcmp(), but try to be a little more aware of the fact that
521
* we're comparing two paths. Right now, it just handles leading
522
* "./" and trailing '/' specially, so that "a/b/" == "./a/b"
524
* TODO: Make this better, so that "./a//b/./c/" == "a/b/c"
525
* TODO: After this works, push it down into libarchive.
526
* TODO: Publish the path normalization routines in libarchive so
527
* that bsdtar can normalize paths and use fast strcmp() instead
530
* Note: This is currently only used within write.c, so should
531
* not handle \ path separators.
535
pathcmp(const char *a, const char *b)
537
/* Skip leading './' */
538
if (a[0] == '.' && a[1] == '/' && a[2] != '\0')
540
if (b[0] == '.' && b[1] == '/' && b[2] != '\0')
542
/* Find the first difference, or return (0) if none. */
550
* If one ends in '/' and the other one doesn't,
553
if (a[0] == '/' && a[1] == '\0' && b[0] == '\0')
555
if (a[0] == '\0' && b[0] == '/' && b[1] == '\0')
557
/* They're really different, return the correct sign. */
558
return (*(const unsigned char *)a - *(const unsigned char *)b);