diff options
Diffstat (limited to 'archival/unzip.c')
-rw-r--r-- | archival/unzip.c | 386 |
1 files changed, 386 insertions, 0 deletions
diff --git a/archival/unzip.c b/archival/unzip.c new file mode 100644 index 000000000..8b1c281c4 --- /dev/null +++ b/archival/unzip.c | |||
@@ -0,0 +1,386 @@ | |||
1 | /* vi: set sw=4 ts=4: */ | ||
2 | /* | ||
3 | * Mini unzip implementation for busybox | ||
4 | * | ||
5 | * Copyright (C) 2004 by Ed Clark | ||
6 | * | ||
7 | * Loosely based on original busybox unzip applet by Laurence Anderson. | ||
8 | * All options and features should work in this version. | ||
9 | * | ||
10 | * Licensed under the GPL v2 or later, see the file LICENSE in this tarball. | ||
11 | */ | ||
12 | |||
13 | /* For reference see | ||
14 | * http://www.pkware.com/company/standards/appnote/ | ||
15 | * http://www.info-zip.org/pub/infozip/doc/appnote-iz-latest.zip | ||
16 | */ | ||
17 | |||
18 | /* TODO | ||
19 | * Endian issues | ||
20 | * Zip64 + other methods | ||
21 | * Improve handling of zip format, ie. | ||
22 | * - deferred CRC, comp. & uncomp. lengths (zip header flags bit 3) | ||
23 | * - unix file permissions, etc. | ||
24 | * - central directory | ||
25 | */ | ||
26 | |||
27 | #include "busybox.h" | ||
28 | #include "unarchive.h" | ||
29 | |||
30 | #define ZIP_FILEHEADER_MAGIC SWAP_LE32(0x04034b50) | ||
31 | #define ZIP_CDS_MAGIC SWAP_LE32(0x02014b50) | ||
32 | #define ZIP_CDS_END_MAGIC SWAP_LE32(0x06054b50) | ||
33 | #define ZIP_DD_MAGIC SWAP_LE32(0x08074b50) | ||
34 | |||
35 | extern unsigned int gunzip_crc; | ||
36 | extern unsigned int gunzip_bytes_out; | ||
37 | |||
38 | typedef union { | ||
39 | unsigned char raw[26]; | ||
40 | struct { | ||
41 | unsigned short version; /* 0-1 */ | ||
42 | unsigned short flags; /* 2-3 */ | ||
43 | unsigned short method; /* 4-5 */ | ||
44 | unsigned short modtime; /* 6-7 */ | ||
45 | unsigned short moddate; /* 8-9 */ | ||
46 | unsigned int crc32 ATTRIBUTE_PACKED; /* 10-13 */ | ||
47 | unsigned int cmpsize ATTRIBUTE_PACKED; /* 14-17 */ | ||
48 | unsigned int ucmpsize ATTRIBUTE_PACKED; /* 18-21 */ | ||
49 | unsigned short filename_len; /* 22-23 */ | ||
50 | unsigned short extra_len; /* 24-25 */ | ||
51 | } formatted ATTRIBUTE_PACKED; | ||
52 | } zip_header_t; | ||
53 | |||
54 | static void unzip_skip(int fd, off_t skip) | ||
55 | { | ||
56 | if (lseek(fd, skip, SEEK_CUR) == (off_t)-1) { | ||
57 | if ((errno != ESPIPE) || (bb_copyfd_size(fd, -1, skip) != skip)) { | ||
58 | bb_error_msg_and_die("seek failure"); | ||
59 | } | ||
60 | } | ||
61 | } | ||
62 | |||
63 | static void unzip_create_leading_dirs(char *fn) | ||
64 | { | ||
65 | /* Create all leading directories */ | ||
66 | char *name = xstrdup(fn); | ||
67 | if (bb_make_directory(dirname(name), 0777, FILEUTILS_RECUR)) { | ||
68 | bb_error_msg_and_die("exiting"); /* bb_make_directory is noisy */ | ||
69 | } | ||
70 | free(name); | ||
71 | } | ||
72 | |||
73 | static int unzip_extract(zip_header_t *zip_header, int src_fd, int dst_fd) | ||
74 | { | ||
75 | if (zip_header->formatted.method == 0) { | ||
76 | /* Method 0 - stored (not compressed) */ | ||
77 | off_t size = zip_header->formatted.ucmpsize; | ||
78 | if (size && (bb_copyfd_size(src_fd, dst_fd, size) != size)) { | ||
79 | bb_error_msg_and_die("cannot complete extraction"); | ||
80 | } | ||
81 | |||
82 | } else { | ||
83 | /* Method 8 - inflate */ | ||
84 | inflate_init(zip_header->formatted.cmpsize); | ||
85 | inflate_unzip(src_fd, dst_fd); | ||
86 | inflate_cleanup(); | ||
87 | /* Validate decompression - crc */ | ||
88 | if (zip_header->formatted.crc32 != (gunzip_crc ^ 0xffffffffL)) { | ||
89 | bb_error_msg("invalid compressed data--%s error", "crc"); | ||
90 | return 1; | ||
91 | } | ||
92 | /* Validate decompression - size */ | ||
93 | if (zip_header->formatted.ucmpsize != gunzip_bytes_out) { | ||
94 | bb_error_msg("invalid compressed data--%s error", "length"); | ||
95 | return 1; | ||
96 | } | ||
97 | } | ||
98 | return 0; | ||
99 | } | ||
100 | |||
101 | int unzip_main(int argc, char **argv) | ||
102 | { | ||
103 | zip_header_t zip_header; | ||
104 | enum {v_silent, v_normal, v_list} verbosity = v_normal; | ||
105 | enum {o_prompt, o_never, o_always} overwrite = o_prompt; | ||
106 | unsigned int total_size = 0; | ||
107 | unsigned int total_entries = 0; | ||
108 | int src_fd = -1, dst_fd = -1; | ||
109 | char *src_fn = NULL, *dst_fn = NULL; | ||
110 | llist_t *zaccept = NULL; | ||
111 | llist_t *zreject = NULL; | ||
112 | char *base_dir = NULL; | ||
113 | int failed, i, opt, opt_range = 0, list_header_done = 0; | ||
114 | char key_buf[512]; | ||
115 | struct stat stat_buf; | ||
116 | |||
117 | while((opt = getopt(argc, argv, "-d:lnopqx")) != -1) { | ||
118 | switch (opt_range) { | ||
119 | case 0: /* Options */ | ||
120 | switch (opt) { | ||
121 | case 'l': /* List */ | ||
122 | verbosity = v_list; | ||
123 | break; | ||
124 | |||
125 | case 'n': /* Never overwrite existing files */ | ||
126 | overwrite = o_never; | ||
127 | break; | ||
128 | |||
129 | case 'o': /* Always overwrite existing files */ | ||
130 | overwrite = o_always; | ||
131 | break; | ||
132 | |||
133 | case 'p': /* Extract files to stdout and fall through to set verbosity */ | ||
134 | dst_fd = STDOUT_FILENO; | ||
135 | |||
136 | case 'q': /* Be quiet */ | ||
137 | verbosity = (verbosity == v_normal) ? v_silent : verbosity; | ||
138 | break; | ||
139 | |||
140 | case 1 : /* The zip file */ | ||
141 | src_fn = xstrndup(optarg, strlen(optarg)+4); | ||
142 | opt_range++; | ||
143 | break; | ||
144 | |||
145 | default: | ||
146 | bb_show_usage(); | ||
147 | |||
148 | } | ||
149 | break; | ||
150 | |||
151 | case 1: /* Include files */ | ||
152 | if (opt == 1) { | ||
153 | llist_add_to(&zaccept, optarg); | ||
154 | |||
155 | } else if (opt == 'd') { | ||
156 | base_dir = optarg; | ||
157 | opt_range += 2; | ||
158 | |||
159 | } else if (opt == 'x') { | ||
160 | opt_range++; | ||
161 | |||
162 | } else { | ||
163 | bb_show_usage(); | ||
164 | } | ||
165 | break; | ||
166 | |||
167 | case 2 : /* Exclude files */ | ||
168 | if (opt == 1) { | ||
169 | llist_add_to(&zreject, optarg); | ||
170 | |||
171 | } else if (opt == 'd') { /* Extract to base directory */ | ||
172 | base_dir = optarg; | ||
173 | opt_range++; | ||
174 | |||
175 | } else { | ||
176 | bb_show_usage(); | ||
177 | } | ||
178 | break; | ||
179 | |||
180 | default: | ||
181 | bb_show_usage(); | ||
182 | } | ||
183 | } | ||
184 | |||
185 | if (src_fn == NULL) { | ||
186 | bb_show_usage(); | ||
187 | } | ||
188 | |||
189 | /* Open input file */ | ||
190 | if (strcmp("-", src_fn) == 0) { | ||
191 | src_fd = STDIN_FILENO; | ||
192 | /* Cannot use prompt mode since zip data is arriving on STDIN */ | ||
193 | overwrite = (overwrite == o_prompt) ? o_never : overwrite; | ||
194 | |||
195 | } else { | ||
196 | static const char *const extn[] = {"", ".zip", ".ZIP"}; | ||
197 | int orig_src_fn_len = strlen(src_fn); | ||
198 | for(i = 0; (i < 3) && (src_fd == -1); i++) { | ||
199 | strcpy(src_fn + orig_src_fn_len, extn[i]); | ||
200 | src_fd = open(src_fn, O_RDONLY); | ||
201 | } | ||
202 | if (src_fd == -1) { | ||
203 | src_fn[orig_src_fn_len] = 0; | ||
204 | bb_error_msg_and_die("cannot open %s, %s.zip, %s.ZIP", src_fn, src_fn, src_fn); | ||
205 | } | ||
206 | } | ||
207 | |||
208 | /* Change dir if necessary */ | ||
209 | if (base_dir) | ||
210 | xchdir(base_dir); | ||
211 | |||
212 | if (verbosity != v_silent) | ||
213 | printf("Archive: %s\n", src_fn); | ||
214 | |||
215 | failed = 0; | ||
216 | |||
217 | while (1) { | ||
218 | unsigned int magic; | ||
219 | |||
220 | /* Check magic number */ | ||
221 | xread(src_fd, &magic, 4); | ||
222 | if (magic == ZIP_CDS_MAGIC) { | ||
223 | break; | ||
224 | } else if (magic != ZIP_FILEHEADER_MAGIC) { | ||
225 | bb_error_msg_and_die("invalid zip magic %08X", magic); | ||
226 | } | ||
227 | |||
228 | /* Read the file header */ | ||
229 | xread(src_fd, zip_header.raw, 26); | ||
230 | zip_header.formatted.version = SWAP_LE32(zip_header.formatted.version); | ||
231 | zip_header.formatted.flags = SWAP_LE32(zip_header.formatted.flags); | ||
232 | zip_header.formatted.method = SWAP_LE32(zip_header.formatted.method); | ||
233 | zip_header.formatted.modtime = SWAP_LE32(zip_header.formatted.modtime); | ||
234 | zip_header.formatted.moddate = SWAP_LE32(zip_header.formatted.moddate); | ||
235 | zip_header.formatted.crc32 = SWAP_LE32(zip_header.formatted.crc32); | ||
236 | zip_header.formatted.cmpsize = SWAP_LE32(zip_header.formatted.cmpsize); | ||
237 | zip_header.formatted.ucmpsize = SWAP_LE32(zip_header.formatted.ucmpsize); | ||
238 | zip_header.formatted.filename_len = SWAP_LE32(zip_header.formatted.filename_len); | ||
239 | zip_header.formatted.extra_len = SWAP_LE32(zip_header.formatted.extra_len); | ||
240 | if ((zip_header.formatted.method != 0) && (zip_header.formatted.method != 8)) { | ||
241 | bb_error_msg_and_die("unsupported compression method %d", zip_header.formatted.method); | ||
242 | } | ||
243 | |||
244 | /* Read filename */ | ||
245 | free(dst_fn); | ||
246 | dst_fn = xzalloc(zip_header.formatted.filename_len + 1); | ||
247 | xread(src_fd, dst_fn, zip_header.formatted.filename_len); | ||
248 | |||
249 | /* Skip extra header bytes */ | ||
250 | unzip_skip(src_fd, zip_header.formatted.extra_len); | ||
251 | |||
252 | if ((verbosity == v_list) && !list_header_done){ | ||
253 | printf(" Length Date Time Name\n" | ||
254 | " -------- ---- ---- ----\n"); | ||
255 | list_header_done = 1; | ||
256 | } | ||
257 | |||
258 | /* Filter zip entries */ | ||
259 | if (find_list_entry(zreject, dst_fn) || | ||
260 | (zaccept && !find_list_entry(zaccept, dst_fn))) { /* Skip entry */ | ||
261 | i = 'n'; | ||
262 | |||
263 | } else { /* Extract entry */ | ||
264 | total_size += zip_header.formatted.ucmpsize; | ||
265 | |||
266 | if (verbosity == v_list) { /* List entry */ | ||
267 | unsigned int dostime = zip_header.formatted.modtime | (zip_header.formatted.moddate << 16); | ||
268 | printf("%9u %02u-%02u-%02u %02u:%02u %s\n", | ||
269 | zip_header.formatted.ucmpsize, | ||
270 | (dostime & 0x01e00000) >> 21, | ||
271 | (dostime & 0x001f0000) >> 16, | ||
272 | (((dostime & 0xfe000000) >> 25) + 1980) % 100, | ||
273 | (dostime & 0x0000f800) >> 11, | ||
274 | (dostime & 0x000007e0) >> 5, | ||
275 | dst_fn); | ||
276 | total_entries++; | ||
277 | i = 'n'; | ||
278 | |||
279 | } else if (dst_fd == STDOUT_FILENO) { /* Extracting to STDOUT */ | ||
280 | i = -1; | ||
281 | |||
282 | } else if (last_char_is(dst_fn, '/')) { /* Extract directory */ | ||
283 | if (stat(dst_fn, &stat_buf) == -1) { | ||
284 | if (errno != ENOENT) { | ||
285 | bb_perror_msg_and_die("cannot stat '%s'",dst_fn); | ||
286 | } | ||
287 | if (verbosity == v_normal) { | ||
288 | printf(" creating: %s\n", dst_fn); | ||
289 | } | ||
290 | unzip_create_leading_dirs(dst_fn); | ||
291 | if (bb_make_directory(dst_fn, 0777, 0)) { | ||
292 | bb_error_msg_and_die("exiting"); | ||
293 | } | ||
294 | } else { | ||
295 | if (!S_ISDIR(stat_buf.st_mode)) { | ||
296 | bb_error_msg_and_die("'%s' exists but is not directory", dst_fn); | ||
297 | } | ||
298 | } | ||
299 | i = 'n'; | ||
300 | |||
301 | } else { /* Extract file */ | ||
302 | _check_file: | ||
303 | if (stat(dst_fn, &stat_buf) == -1) { /* File does not exist */ | ||
304 | if (errno != ENOENT) { | ||
305 | bb_perror_msg_and_die("cannot stat '%s'",dst_fn); | ||
306 | } | ||
307 | i = 'y'; | ||
308 | |||
309 | } else { /* File already exists */ | ||
310 | if (overwrite == o_never) { | ||
311 | i = 'n'; | ||
312 | |||
313 | } else if (S_ISREG(stat_buf.st_mode)) { /* File is regular file */ | ||
314 | if (overwrite == o_always) { | ||
315 | i = 'y'; | ||
316 | } else { | ||
317 | printf("replace %s? [y]es, [n]o, [A]ll, [N]one, [r]ename: ", dst_fn); | ||
318 | if (!fgets(key_buf, 512, stdin)) { | ||
319 | bb_perror_msg_and_die("cannot read input"); | ||
320 | } | ||
321 | i = key_buf[0]; | ||
322 | } | ||
323 | |||
324 | } else { /* File is not regular file */ | ||
325 | bb_error_msg_and_die("'%s' exists but is not regular file",dst_fn); | ||
326 | } | ||
327 | } | ||
328 | } | ||
329 | } | ||
330 | |||
331 | switch (i) { | ||
332 | case 'A': | ||
333 | overwrite = o_always; | ||
334 | case 'y': /* Open file and fall into unzip */ | ||
335 | unzip_create_leading_dirs(dst_fn); | ||
336 | dst_fd = xopen(dst_fn, O_WRONLY | O_CREAT | O_TRUNC); | ||
337 | case -1: /* Unzip */ | ||
338 | if (verbosity == v_normal) { | ||
339 | printf(" inflating: %s\n", dst_fn); | ||
340 | } | ||
341 | if (unzip_extract(&zip_header, src_fd, dst_fd)) { | ||
342 | failed = 1; | ||
343 | } | ||
344 | if (dst_fd != STDOUT_FILENO) { | ||
345 | /* closing STDOUT is potentially bad for future business */ | ||
346 | close(dst_fd); | ||
347 | } | ||
348 | break; | ||
349 | |||
350 | case 'N': | ||
351 | overwrite = o_never; | ||
352 | case 'n': | ||
353 | /* Skip entry data */ | ||
354 | unzip_skip(src_fd, zip_header.formatted.cmpsize); | ||
355 | break; | ||
356 | |||
357 | case 'r': | ||
358 | /* Prompt for new name */ | ||
359 | printf("new name: "); | ||
360 | if (!fgets(key_buf, 512, stdin)) { | ||
361 | bb_perror_msg_and_die("cannot read input"); | ||
362 | } | ||
363 | free(dst_fn); | ||
364 | dst_fn = xstrdup(key_buf); | ||
365 | chomp(dst_fn); | ||
366 | goto _check_file; | ||
367 | |||
368 | default: | ||
369 | printf("error: invalid response [%c]\n",(char)i); | ||
370 | goto _check_file; | ||
371 | } | ||
372 | |||
373 | /* Data descriptor section */ | ||
374 | if (zip_header.formatted.flags & 4) { | ||
375 | /* skip over duplicate crc, compressed size and uncompressed size */ | ||
376 | unzip_skip(src_fd, 12); | ||
377 | } | ||
378 | } | ||
379 | |||
380 | if (verbosity == v_list) { | ||
381 | printf(" -------- -------\n" | ||
382 | "%9d %d files\n", total_size, total_entries); | ||
383 | } | ||
384 | |||
385 | return failed; | ||
386 | } | ||