aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorGlenn L McGrath <bug1@ihug.co.nz>2001-10-25 14:53:50 +0000
committerGlenn L McGrath <bug1@ihug.co.nz>2001-10-25 14:53:50 +0000
commit633f153808c3ec316f023cdb72b98c4e5a929bca (patch)
treebe37d276ee997ec744c43796667cf7c1da3f1c5f
parentef0eab514d87e162f618889209b8520d2a30c73b (diff)
downloadbusybox-w32-633f153808c3ec316f023cdb72b98c4e5a929bca.tar.gz
busybox-w32-633f153808c3ec316f023cdb72b98c4e5a929bca.tar.bz2
busybox-w32-633f153808c3ec316f023cdb72b98c4e5a929bca.zip
Move libbb/unarchive functions to ./archival/libunarchive
Diffstat
-rw-r--r--libbb/unarchive.c627
1 files changed, 0 insertions, 627 deletions
diff --git a/libbb/unarchive.c b/libbb/unarchive.c
deleted file mode 100644
index 9c599a415..000000000
--- a/libbb/unarchive.c
+++ /dev/null
@@ -1,627 +0,0 @@
1/*
2 * Copyright (C) 2000 by Glenn McGrath
3 * Copyright (C) 2001 by Laurence Anderson
4 *
5 * Based on previous work by busybox developers and others.
6 *
7 * This program is free software; you can redistribute it and/or modify
8 * it under the terms of the GNU General Public License as published by
9 * the Free Software Foundation; either version 2 of the License, or
10 * (at your option) any later version.
11 *
12 * This program is distributed in the hope that it will be useful,
13 * but WITHOUT ANY WARRANTY; without even the implied warranty of
14 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
15 * GNU Library General Public License for more details.
16 *
17 * You should have received a copy of the GNU General Public License
18 * along with this program; if not, write to the Free Software
19 * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
20 */
21
22#include <stdio.h>
23#include <errno.h>
24#include <fnmatch.h>
25#include <stdlib.h>
26#include <string.h>
27#include <unistd.h>
28#include <utime.h>
29#include "libbb.h"
30
31extern void seek_sub_file(FILE *src_stream, const int count);
32extern char *extract_archive(FILE *src_stream, FILE *out_stream, const file_header_t *file_entry,
33 const int function, const char *prefix);
34
35
36#ifdef L_archive_offset
37off_t archive_offset;
38#else
39extern off_t archive_offset;
40#endif
41
42#ifdef L_seek_sub_file
43void seek_sub_file(FILE *src_stream, const int count)
44{
45 int i;
46 /* Try to fseek as faster */
47 archive_offset += count;
48 if (fseek(src_stream, count, SEEK_CUR) != 0 && errno == ESPIPE) {
49 for (i = 0; i < count; i++) {
50 fgetc(src_stream);
51 }
52 }
53 return;
54}
55#endif
56
57
58
59#ifdef L_extract_archive
60/* Extract the data postioned at src_stream to either filesystem, stdout or
61 * buffer depending on the value of 'function' which is defined in libbb.h
62 *
63 * prefix doesnt have to be just a directory, it may prefix the filename as well.
64 *
65 * e.g. '/var/lib/dpkg/info/dpkg.' will extract all files to the base bath
66 * '/var/lib/dpkg/info/' and all files/dirs created in that dir will have
67 * 'dpkg.' as their prefix
68 *
69 * For this reason if prefix does point to a dir then it must end with a
70 * trailing '/' or else the last dir will be assumed to be the file prefix
71 */
72char *extract_archive(FILE *src_stream, FILE *out_stream, const file_header_t *file_entry,
73 const int function, const char *prefix)
74{
75 FILE *dst_stream = NULL;
76 char *full_name = NULL;
77 char *buffer = NULL;
78 struct utimbuf t;
79
80 /* prefix doesnt have to be a proper path it may prepend
81 * the filename as well */
82 if (prefix != NULL) {
83 /* strip leading '/' in filename to extract as prefix may not be dir */
84 /* Cant use concat_path_file here as prefix might not be a directory */
85 char *path = file_entry->name;
86 if (strncmp("./", path, 2) == 0) {
87 path += 2;
88 if (strlen(path) == 0) {
89 return(NULL);
90 }
91 }
92 full_name = xmalloc(strlen(prefix) + strlen(path) + 1);
93 strcpy(full_name, prefix);
94 strcat(full_name, path);
95 } else {
96 full_name = file_entry->name;
97 }
98 if (function & extract_to_stdout) {
99 if (S_ISREG(file_entry->mode)) {
100 copy_file_chunk(src_stream, out_stream, file_entry->size);
101 archive_offset += file_entry->size;
102 }
103 }
104 else if (function & extract_one_to_buffer) {
105 if (S_ISREG(file_entry->mode)) {
106 buffer = (char *) xmalloc(file_entry->size + 1);
107 fread(buffer, 1, file_entry->size, src_stream);
108 buffer[file_entry->size] = '\0';
109 archive_offset += file_entry->size;
110 return(buffer);
111 }
112 }
113 else if (function & extract_all_to_fs) {
114 struct stat oldfile;
115 int stat_res;
116 stat_res = lstat (full_name, &oldfile);
117 if (stat_res == 0) { /* The file already exists */
118 if ((function & extract_unconditional) || (oldfile.st_mtime < file_entry->mtime)) {
119 if (!S_ISDIR(oldfile.st_mode)) {
120 unlink(full_name); /* Directories might not be empty etc */
121 }
122 } else {
123 if ((function & extract_quiet) != extract_quiet) {
124 error_msg("%s not created: newer or same age file exists", file_entry->name);
125 }
126 seek_sub_file(src_stream, file_entry->size);
127 return (NULL);
128 }
129 }
130 if (function & extract_create_leading_dirs) { /* Create leading directories with default umask */
131 char *buf, *parent;
132 buf = xstrdup(full_name);
133 parent = dirname(buf);
134 if (make_directory (parent, -1, FILEUTILS_RECUR) != 0) {
135 if ((function & extract_quiet) != extract_quiet) {
136 error_msg("couldn't create leading directories");
137 }
138 }
139 free (buf);
140 }
141 switch(file_entry->mode & S_IFMT) {
142 case S_IFREG:
143 if (file_entry->link_name) { /* Found a cpio hard link */
144 if (link(file_entry->link_name, full_name) != 0) {
145 if ((function & extract_quiet) != extract_quiet) {
146 perror_msg("Cannot link from %s to '%s'",
147 file_entry->name, file_entry->link_name);
148 }
149 }
150 } else {
151 if ((dst_stream = wfopen(full_name, "w")) == NULL) {
152 seek_sub_file(src_stream, file_entry->size);
153 return NULL;
154 }
155 archive_offset += file_entry->size;
156 copy_file_chunk(src_stream, dst_stream, file_entry->size);
157 fclose(dst_stream);
158 }
159 break;
160 case S_IFDIR:
161 if (stat_res != 0) {
162 if (mkdir(full_name, file_entry->mode) < 0) {
163 if ((function & extract_quiet) != extract_quiet) {
164 perror_msg("extract_archive: %s", full_name);
165 }
166 }
167 }
168 break;
169 case S_IFLNK:
170 if (symlink(file_entry->link_name, full_name) < 0) {
171 if ((function & extract_quiet) != extract_quiet) {
172 perror_msg("Cannot create symlink from %s to '%s'", file_entry->name, file_entry->link_name);
173 }
174 return NULL;
175 }
176 break;
177 case S_IFSOCK:
178 case S_IFBLK:
179 case S_IFCHR:
180 case S_IFIFO:
181 if (mknod(full_name, file_entry->mode, file_entry->device) == -1) {
182 if ((function & extract_quiet) != extract_quiet) {
183 perror_msg("Cannot create node %s", file_entry->name);
184 }
185 return NULL;
186 }
187 break;
188 }
189
190 /* Changing a symlink's properties normally changes the properties of the
191 * file pointed to, so dont try and change the date or mode, lchown does
192 * does the right thing, but isnt available in older versions of libc */
193 if (S_ISLNK(file_entry->mode)) {
194#if (__GLIBC__ > 2) && (__GLIBC_MINOR__ > 1)
195 lchown(full_name, file_entry->uid, file_entry->gid);
196#endif
197 } else {
198 if (function & extract_preserve_date) {
199 t.actime = file_entry->mtime;
200 t.modtime = file_entry->mtime;
201 utime(full_name, &t);
202 }
203 chmod(full_name, file_entry->mode);
204 chown(full_name, file_entry->uid, file_entry->gid);
205 }
206 } else {
207 /* If we arent extracting data we have to skip it,
208 * if data size is 0 then then just do it anyway
209 * (saves testing for it) */
210 seek_sub_file(src_stream, file_entry->size);
211 }
212
213 /* extract_list and extract_verbose_list can be used in conjunction
214 * with one of the above four extraction functions, so do this seperately */
215 if (function & extract_verbose_list) {
216 fprintf(out_stream, "%s %d/%d %8d %s ", mode_string(file_entry->mode),
217 file_entry->uid, file_entry->gid,
218 (int) file_entry->size, time_string(file_entry->mtime));
219 }
220 if ((function & extract_list) || (function & extract_verbose_list)){
221 /* fputs doesnt add a trailing \n, so use fprintf */
222 fprintf(out_stream, "%s\n", full_name);
223 }
224
225 free(full_name);
226
227 return(NULL); /* Maybe we should say if failed */
228}
229#endif
230
231#ifdef L_unarchive
232char *unarchive(FILE *src_stream, FILE *out_stream, file_header_t *(*get_headers)(FILE *),
233 const int extract_function, const char *prefix, char **include_name, char **exclude_name)
234{
235 file_header_t *file_entry;
236 int extract_flag = TRUE;
237 int i;
238 char *buffer = NULL;
239
240 archive_offset = 0;
241 while ((file_entry = get_headers(src_stream)) != NULL) {
242
243 if (include_name != NULL) {
244 extract_flag = FALSE;
245 for(i = 0; include_name[i] != 0; i++) {
246 if (fnmatch(include_name[i], file_entry->name, FNM_LEADING_DIR) == 0) {
247 extract_flag = TRUE;
248 break;
249 }
250 }
251 } else {
252 extract_flag = TRUE;
253 }
254
255 /* If the file entry is in the exclude list dont extract it */
256 if (exclude_name != NULL) {
257 for(i = 0; exclude_name[i] != 0; i++) {
258 if (fnmatch(exclude_name[i], file_entry->name, FNM_LEADING_DIR) == 0) {
259 extract_flag = FALSE;
260 break;
261 }
262 }
263 }
264
265 if (extract_flag == TRUE) {
266 buffer = extract_archive(src_stream, out_stream, file_entry, extract_function, prefix);
267 } else {
268 /* seek past the data entry */
269 seek_sub_file(src_stream, file_entry->size);
270 }
271 free(file_entry->name); /* may be null, but doesn't matter */
272 if (file_entry->link_name) {
273 free(file_entry->link_name);
274 }
275 free(file_entry);
276 }
277 return(buffer);
278}
279#endif
280
281#ifdef L_get_header_ar
282file_header_t *get_header_ar(FILE *src_stream)
283{
284 file_header_t *typed;
285 union {
286 char raw[60];
287 struct {
288 char name[16];
289 char date[12];
290 char uid[6];
291 char gid[6];
292 char mode[8];
293 char size[10];
294 char magic[2];
295 } formated;
296 } ar;
297 static char *ar_long_names;
298
299 if (fread(ar.raw, 1, 60, src_stream) != 60) {
300 return(NULL);
301 }
302 archive_offset += 60;
303 /* align the headers based on the header magic */
304 if ((ar.formated.magic[0] != '`') || (ar.formated.magic[1] != '\n')) {
305 /* some version of ar, have an extra '\n' after each data entry,
306 * this puts the next header out by 1 */
307 if (ar.formated.magic[1] != '`') {
308 error_msg("Invalid magic");
309 return(NULL);
310 }
311 /* read the next char out of what would be the data section,
312 * if its a '\n' then it is a valid header offset by 1*/
313 archive_offset++;
314 if (fgetc(src_stream) != '\n') {
315 error_msg("Invalid magic");
316 return(NULL);
317 }
318 /* fix up the header, we started reading 1 byte too early */
319 /* raw_header[60] wont be '\n' as it should, but it doesnt matter */
320 memmove(ar.raw, &ar.raw[1], 59);
321 }
322
323 typed = (file_header_t *) xcalloc(1, sizeof(file_header_t));
324
325 typed->size = (size_t) atoi(ar.formated.size);
326 /* long filenames have '/' as the first character */
327 if (ar.formated.name[0] == '/') {
328 if (ar.formated.name[1] == '/') {
329 /* If the second char is a '/' then this entries data section
330 * stores long filename for multiple entries, they are stored
331 * in static variable long_names for use in future entries */
332 ar_long_names = (char *) xrealloc(ar_long_names, typed->size);
333 fread(ar_long_names, 1, typed->size, src_stream);
334 archive_offset += typed->size;
335 /* This ar entries data section only contained filenames for other records
336 * they are stored in the static ar_long_names for future reference */
337 return (get_header_ar(src_stream)); /* Return next header */
338 } else if (ar.formated.name[1] == ' ') {
339 /* This is the index of symbols in the file for compilers */
340 seek_sub_file(src_stream, typed->size);
341 return (get_header_ar(src_stream)); /* Return next header */
342 } else {
343 /* The number after the '/' indicates the offset in the ar data section
344 (saved in variable long_name) that conatains the real filename */
345 if (!ar_long_names) {
346 error_msg("Cannot resolve long file name");
347 return (NULL);
348 }
349 typed->name = xstrdup(ar_long_names + atoi(&ar.formated.name[1]));
350 }
351 } else {
352 /* short filenames */
353 typed->name = xcalloc(1, 16);
354 strncpy(typed->name, ar.formated.name, 16);
355 }
356 typed->name[strcspn(typed->name, " /")]='\0';
357
358 /* convert the rest of the now valid char header to its typed struct */
359 parse_mode(ar.formated.mode, &typed->mode);
360 typed->mtime = atoi(ar.formated.date);
361 typed->uid = atoi(ar.formated.uid);
362 typed->gid = atoi(ar.formated.gid);
363
364 return(typed);
365}
366#endif
367
368#ifdef L_get_header_cpio
369struct hardlinks {
370 file_header_t *entry;
371 int inode;
372 struct hardlinks *next;
373};
374
375file_header_t *get_header_cpio(FILE *src_stream)
376{
377 file_header_t *cpio_entry = NULL;
378 char cpio_header[110];
379 int namesize;
380 char dummy[16];
381 int major, minor, nlink, inode;
382 static struct hardlinks *saved_hardlinks = NULL;
383 static int pending_hardlinks = 0;
384
385 if (pending_hardlinks) { /* Deal with any pending hardlinks */
386 struct hardlinks *tmp = saved_hardlinks, *oldtmp = NULL;
387 while (tmp) {
388 if (tmp->entry->link_name) { /* Found a hardlink ready to be extracted */
389 cpio_entry = tmp->entry;
390 if (oldtmp) oldtmp->next = tmp->next; /* Remove item from linked list */
391 else saved_hardlinks = tmp->next;
392 free(tmp);
393 return (cpio_entry);
394 }
395 oldtmp = tmp;
396 tmp = tmp->next;
397 }
398 pending_hardlinks = 0; /* No more pending hardlinks, read next file entry */
399 }
400
401 /* There can be padding before archive header */
402 seek_sub_file(src_stream, (4 - (archive_offset % 4)) % 4);
403 if (fread(cpio_header, 1, 110, src_stream) == 110) {
404 archive_offset += 110;
405 if (strncmp(cpio_header, "07070", 5) != 0) {
406 error_msg("Unsupported format or invalid magic");
407 return(NULL);
408 }
409 switch (cpio_header[5]) {
410 case '2': /* "crc" header format */
411 /* Doesnt do the crc check yet */
412 case '1': /* "newc" header format */
413 cpio_entry = (file_header_t *) xcalloc(1, sizeof(file_header_t));
414 sscanf(cpio_header, "%6c%8x%8x%8x%8x%8x%8lx%8lx%16c%8x%8x%8x%8c",
415 dummy, &inode, (unsigned int*)&cpio_entry->mode,
416 (unsigned int*)&cpio_entry->uid, (unsigned int*)&cpio_entry->gid,
417 &nlink, &cpio_entry->mtime, &cpio_entry->size,
418 dummy, &major, &minor, &namesize, dummy);
419
420 cpio_entry->name = (char *) xcalloc(1, namesize);
421 fread(cpio_entry->name, 1, namesize, src_stream); /* Read in filename */
422 archive_offset += namesize;
423 /* Skip padding before file contents */
424 seek_sub_file(src_stream, (4 - (archive_offset % 4)) % 4);
425 if (strcmp(cpio_entry->name, "TRAILER!!!") == 0) {
426 printf("%d blocks\n", (int) (archive_offset % 512 ? (archive_offset / 512) + 1 : archive_offset / 512)); /* Always round up */
427 if (saved_hardlinks) { /* Bummer - we still have unresolved hardlinks */
428 struct hardlinks *tmp = saved_hardlinks, *oldtmp = NULL;
429 while (tmp) {
430 error_msg("%s not created: cannot resolve hardlink", tmp->entry->name);
431 oldtmp = tmp;
432 tmp = tmp->next;
433 free (oldtmp->entry->name);
434 free (oldtmp->entry);
435 free (oldtmp);
436 }
437 saved_hardlinks = NULL;
438 pending_hardlinks = 0;
439 }
440 return(NULL);
441 }
442
443 if (S_ISLNK(cpio_entry->mode)) {
444 cpio_entry->link_name = (char *) xcalloc(1, cpio_entry->size + 1);
445 fread(cpio_entry->link_name, 1, cpio_entry->size, src_stream);
446 archive_offset += cpio_entry->size;
447 cpio_entry->size = 0; /* Stop possiable seeks in future */
448 }
449 if (nlink > 1 && !S_ISDIR(cpio_entry->mode)) {
450 if (cpio_entry->size == 0) { /* Put file on a linked list for later */
451 struct hardlinks *new = xmalloc(sizeof(struct hardlinks));
452 new->next = saved_hardlinks;
453 new->inode = inode;
454 new->entry = cpio_entry;
455 saved_hardlinks = new;
456 return(get_header_cpio(src_stream)); /* Recurse to next file */
457 } else { /* Found the file with data in */
458 struct hardlinks *tmp = saved_hardlinks;
459 pending_hardlinks = 1;
460 while (tmp) {
461 if (tmp->inode == inode) {
462 tmp->entry->link_name = xstrdup(cpio_entry->name);
463 nlink--;
464 }
465 tmp = tmp->next;
466 }
467 if (nlink > 1) error_msg("error resolving hardlink: did you create the archive with GNU cpio 2.0-2.2?");
468 }
469 }
470 cpio_entry->device = (major << 8) | minor;
471 break;
472 default:
473 error_msg("Unsupported format");
474 return(NULL);
475 }
476 if (ferror(src_stream) || feof(src_stream)) {
477 perror_msg("Stream error");
478 return(NULL);
479 }
480 }
481 return(cpio_entry);
482}
483#endif
484
485#ifdef L_get_header_tar
486file_header_t *get_header_tar(FILE *tar_stream)
487{
488 union {
489 unsigned char raw[512];
490 struct {
491 char name[100]; /* 0-99 */
492 char mode[8]; /* 100-107 */
493 char uid[8]; /* 108-115 */
494 char gid[8]; /* 116-123 */
495 char size[12]; /* 124-135 */
496 char mtime[12]; /* 136-147 */
497 char chksum[8]; /* 148-155 */
498 char typeflag; /* 156-156 */
499 char linkname[100]; /* 157-256 */
500 char magic[6]; /* 257-262 */
501 char version[2]; /* 263-264 */
502 char uname[32]; /* 265-296 */
503 char gname[32]; /* 297-328 */
504 char devmajor[8]; /* 329-336 */
505 char devminor[8]; /* 337-344 */
506 char prefix[155]; /* 345-499 */
507 char padding[12]; /* 500-512 */
508 } formated;
509 } tar;
510 file_header_t *tar_entry = NULL;
511 long sum = 0;
512 long i;
513
514 if (archive_offset % 512 != 0) {
515 seek_sub_file(tar_stream, 512 - (archive_offset % 512));
516 }
517
518 if (fread(tar.raw, 1, 512, tar_stream) != 512) {
519 /* Unfortunatly its common for tar files to have all sorts of
520 * trailing garbage, fail silently */
521// error_msg("Couldnt read header");
522 return(NULL);
523 }
524 archive_offset += 512;
525
526 /* Check header has valid magic, unfortunately some tar files
527 * have empty (0'ed) tar entries at the end, which will
528 * cause this to fail, so fail silently for now
529 */
530 if (strncmp(tar.formated.magic, "ustar", 5) != 0) {
531 return(NULL);
532 }
533
534 /* Do checksum on headers */
535 for (i = 0; i < 148 ; i++) {
536 sum += tar.raw[i];
537 }
538 sum += ' ' * 8;
539 for (i = 156; i < 512 ; i++) {
540 sum += tar.raw[i];
541 }
542 if (sum != strtol(tar.formated.chksum, NULL, 8)) {
543 error_msg("Invalid tar header checksum");
544 return(NULL);
545 }
546
547 /* convert to type'ed variables */
548 tar_entry = xcalloc(1, sizeof(file_header_t));
549 tar_entry->name = xstrdup(tar.formated.name);
550
551 parse_mode(tar.formated.mode, &tar_entry->mode);
552 tar_entry->uid = strtol(tar.formated.uid, NULL, 8);
553 tar_entry->gid = strtol(tar.formated.gid, NULL, 8);
554 tar_entry->size = strtol(tar.formated.size, NULL, 8);
555 tar_entry->mtime = strtol(tar.formated.mtime, NULL, 8);
556 tar_entry->link_name = strlen(tar.formated.linkname) ?
557 xstrdup(tar.formated.linkname) : NULL;
558 tar_entry->device = (strtol(tar.formated.devmajor, NULL, 8) << 8) +
559 strtol(tar.formated.devminor, NULL, 8);
560
561 return(tar_entry);
562}
563#endif
564
565#ifdef L_deb_extract
566char *deb_extract(const char *package_filename, FILE *out_stream,
567 const int extract_function, const char *prefix, const char *filename)
568{
569 FILE *deb_stream;
570 FILE *uncompressed_stream = NULL;
571 file_header_t *ar_header = NULL;
572 char **file_list = NULL;
573 char *output_buffer = NULL;
574 char *ared_file = NULL;
575 char ar_magic[8];
576 int gunzip_pid;
577
578 if (filename != NULL) {
579 file_list = xmalloc(sizeof(char *) * 2);
580 file_list[0] = xstrdup(filename);
581 file_list[1] = NULL;
582 }
583
584 if (extract_function & extract_control_tar_gz) {
585 ared_file = xstrdup("control.tar.gz");
586 }
587 else if (extract_function & extract_data_tar_gz) {
588 ared_file = xstrdup("data.tar.gz");
589 }
590
591 /* open the debian package to be worked on */
592 deb_stream = wfopen(package_filename, "r");
593 if (deb_stream == NULL) {
594 return(NULL);
595 }
596 /* set the buffer size */
597 setvbuf(deb_stream, NULL, _IOFBF, 0x8000);
598
599 /* check ar magic */
600 fread(ar_magic, 1, 8, deb_stream);
601 if (strncmp(ar_magic,"!<arch>",7) != 0) {
602 error_msg_and_die("invalid magic");
603 }
604 archive_offset = 8;
605
606 while ((ar_header = get_header_ar(deb_stream)) != NULL) {
607 if (strcmp(ared_file, ar_header->name) == 0) {
608 /* open a stream of decompressed data */
609 uncompressed_stream = gz_open(deb_stream, &gunzip_pid);
610 archive_offset = 0;
611 output_buffer = unarchive(uncompressed_stream, out_stream, get_header_tar, extract_function, prefix, file_list, NULL);
612 }
613 seek_sub_file(deb_stream, ar_header->size);
614 free(ar_header->name);
615 free(ar_header);
616 }
617 gz_close(gunzip_pid);
618 fclose(deb_stream);
619 fclose(uncompressed_stream);
620 free(ared_file);
621 if (filename != NULL) {
622 free(file_list[0]);
623 free(file_list);
624 }
625 return(output_buffer);
626}
627#endif
generated by cgit v1.2.3-55-g6feb (git 2.39.1) at 2025-03-22 15:46:38 +0000