/src/libzip/lib/zip_source_compress.c
Line | Count | Source (jump to first uncovered line) |
1 | | /* |
2 | | zip_source_compress.c -- (de)compression routines |
3 | | Copyright (C) 2017-2023 Dieter Baron and Thomas Klausner |
4 | | |
5 | | This file is part of libzip, a library to manipulate ZIP archives. |
6 | | The authors can be contacted at <info@libzip.org> |
7 | | |
8 | | Redistribution and use in source and binary forms, with or without |
9 | | modification, are permitted provided that the following conditions |
10 | | are met: |
11 | | 1. Redistributions of source code must retain the above copyright |
12 | | notice, this list of conditions and the following disclaimer. |
13 | | 2. Redistributions in binary form must reproduce the above copyright |
14 | | notice, this list of conditions and the following disclaimer in |
15 | | the documentation and/or other materials provided with the |
16 | | distribution. |
17 | | 3. The names of the authors may not be used to endorse or promote |
18 | | products derived from this software without specific prior |
19 | | written permission. |
20 | | |
21 | | THIS SOFTWARE IS PROVIDED BY THE AUTHORS ``AS IS'' AND ANY EXPRESS |
22 | | OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED |
23 | | WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE |
24 | | ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR ANY |
25 | | DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL |
26 | | DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE |
27 | | GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS |
28 | | INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER |
29 | | IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR |
30 | | OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN |
31 | | IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. |
32 | | */ |
33 | | |
34 | | #include <stdlib.h> |
35 | | #include <string.h> |
36 | | |
37 | | #include "zipint.h" |
38 | | |
39 | | struct context { |
40 | | zip_error_t error; |
41 | | |
42 | | bool end_of_input; |
43 | | bool end_of_stream; |
44 | | bool can_store; |
45 | | bool is_stored; /* only valid if end_of_stream is true */ |
46 | | bool compress; |
47 | | bool check_consistency; |
48 | | zip_int32_t method; |
49 | | |
50 | | zip_uint64_t size; |
51 | | zip_int64_t first_read; |
52 | | zip_uint8_t buffer[BUFSIZE]; |
53 | | |
54 | | zip_compression_algorithm_t *algorithm; |
55 | | void *ud; |
56 | | }; |
57 | | |
58 | | |
59 | | struct implementation { |
60 | | zip_uint16_t method; |
61 | | zip_compression_algorithm_t *compress; |
62 | | zip_compression_algorithm_t *decompress; |
63 | | }; |
64 | | |
65 | | static struct implementation implementations[] = { |
66 | | {ZIP_CM_DEFLATE, &zip_algorithm_deflate_compress, &zip_algorithm_deflate_decompress}, |
67 | | #if defined(HAVE_LIBBZ2) |
68 | | {ZIP_CM_BZIP2, &zip_algorithm_bzip2_compress, &zip_algorithm_bzip2_decompress}, |
69 | | #endif |
70 | | #if defined(HAVE_LIBLZMA) |
71 | | {ZIP_CM_LZMA, &zip_algorithm_xz_compress, &zip_algorithm_xz_decompress}, |
72 | | /* Disabled - because 7z isn't able to unpack ZIP+LZMA2 |
73 | | archives made this way - and vice versa. |
74 | | |
75 | | {ZIP_CM_LZMA2, &zip_algorithm_xz_compress, &zip_algorithm_xz_decompress}, |
76 | | */ |
77 | | {ZIP_CM_XZ, &zip_algorithm_xz_compress, &zip_algorithm_xz_decompress}, |
78 | | #endif |
79 | | #if defined(HAVE_LIBZSTD) |
80 | | {ZIP_CM_ZSTD, &zip_algorithm_zstd_compress, &zip_algorithm_zstd_decompress}, |
81 | | #endif |
82 | | |
83 | | }; |
84 | | |
85 | | static size_t implementations_size = sizeof(implementations) / sizeof(implementations[0]); |
86 | | |
87 | | static zip_source_t *compression_source_new(zip_t *za, zip_source_t *src, zip_int32_t method, bool compress, zip_uint32_t compression_flags); |
88 | | static zip_int64_t compress_callback(zip_source_t *, void *, void *, zip_uint64_t, zip_source_cmd_t); |
89 | | static void context_free(struct context *ctx); |
90 | | static struct context *context_new(zip_int32_t method, bool compress, zip_uint32_t compression_flags, zip_compression_algorithm_t *algorithm, bool check_consistency); |
91 | | static zip_int64_t compress_read(zip_source_t *, struct context *, void *, zip_uint64_t); |
92 | | |
93 | 0 | zip_compression_algorithm_t *_zip_get_compression_algorithm(zip_int32_t method, bool compress) { |
94 | 0 | size_t i; |
95 | 0 | zip_uint16_t real_method = ZIP_CM_ACTUAL(method); |
96 | |
|
97 | 0 | for (i = 0; i < implementations_size; i++) { |
98 | 0 | if (implementations[i].method == real_method) { |
99 | 0 | if (compress) { |
100 | 0 | return implementations[i].compress; |
101 | 0 | } |
102 | 0 | else { |
103 | 0 | return implementations[i].decompress; |
104 | 0 | } |
105 | 0 | } |
106 | 0 | } |
107 | | |
108 | 0 | return NULL; |
109 | 0 | } |
110 | | |
111 | 0 | ZIP_EXTERN int zip_compression_method_supported(zip_int32_t method, int compress) { |
112 | 0 | if (method == ZIP_CM_STORE) { |
113 | 0 | return 1; |
114 | 0 | } |
115 | 0 | return _zip_get_compression_algorithm(method, compress) != NULL; |
116 | 0 | } |
117 | | |
118 | 0 | zip_source_t *zip_source_compress(zip_t *za, zip_source_t *src, zip_int32_t method, zip_uint32_t compression_flags) { |
119 | 0 | return compression_source_new(za, src, method, true, compression_flags); |
120 | 0 | } |
121 | | |
122 | | zip_source_t * |
123 | 0 | zip_source_decompress(zip_t *za, zip_source_t *src, zip_int32_t method) { |
124 | 0 | return compression_source_new(za, src, method, false, 0); |
125 | 0 | } |
126 | | |
127 | | |
128 | 0 | static zip_source_t *compression_source_new(zip_t *za, zip_source_t *src, zip_int32_t method, bool compress, zip_uint32_t compression_flags) { |
129 | 0 | struct context *ctx; |
130 | 0 | zip_source_t *s2; |
131 | 0 | zip_compression_algorithm_t *algorithm = NULL; |
132 | |
|
133 | 0 | if (src == NULL) { |
134 | 0 | zip_error_set(&za->error, ZIP_ER_INVAL, 0); |
135 | 0 | return NULL; |
136 | 0 | } |
137 | | |
138 | 0 | if ((algorithm = _zip_get_compression_algorithm(method, compress)) == NULL) { |
139 | 0 | zip_error_set(&za->error, ZIP_ER_COMPNOTSUPP, 0); |
140 | 0 | return NULL; |
141 | 0 | } |
142 | | |
143 | 0 | if ((ctx = context_new(method, compress, compression_flags, algorithm, za->open_flags & ZIP_CHECKCONS)) == NULL) { |
144 | 0 | zip_error_set(&za->error, ZIP_ER_MEMORY, 0); |
145 | 0 | return NULL; |
146 | 0 | } |
147 | | |
148 | 0 | if ((s2 = zip_source_layered(za, src, compress_callback, ctx)) == NULL) { |
149 | 0 | context_free(ctx); |
150 | 0 | return NULL; |
151 | 0 | } |
152 | | |
153 | 0 | return s2; |
154 | 0 | } |
155 | | |
156 | | |
157 | 0 | static struct context *context_new(zip_int32_t method, bool compress, zip_uint32_t compression_flags, zip_compression_algorithm_t *algorithm, bool check_consistency) { |
158 | 0 | struct context *ctx; |
159 | |
|
160 | 0 | if ((ctx = (struct context *)malloc(sizeof(*ctx))) == NULL) { |
161 | 0 | return NULL; |
162 | 0 | } |
163 | 0 | zip_error_init(&ctx->error); |
164 | 0 | ctx->can_store = compress ? ZIP_CM_IS_DEFAULT(method) : false; |
165 | 0 | ctx->algorithm = algorithm; |
166 | 0 | ctx->method = method; |
167 | 0 | ctx->compress = compress; |
168 | 0 | ctx->end_of_input = false; |
169 | 0 | ctx->end_of_stream = false; |
170 | 0 | ctx->is_stored = false; |
171 | 0 | ctx->check_consistency = check_consistency; |
172 | |
|
173 | 0 | if ((ctx->ud = ctx->algorithm->allocate(ZIP_CM_ACTUAL(method), compression_flags, &ctx->error)) == NULL) { |
174 | 0 | zip_error_fini(&ctx->error); |
175 | 0 | free(ctx); |
176 | 0 | return NULL; |
177 | 0 | } |
178 | | |
179 | 0 | return ctx; |
180 | 0 | } |
181 | | |
182 | | |
183 | | static void |
184 | 0 | context_free(struct context *ctx) { |
185 | 0 | if (ctx == NULL) { |
186 | 0 | return; |
187 | 0 | } |
188 | | |
189 | 0 | ctx->algorithm->deallocate(ctx->ud); |
190 | 0 | zip_error_fini(&ctx->error); |
191 | |
|
192 | 0 | free(ctx); |
193 | 0 | } |
194 | | |
195 | | |
196 | | static zip_int64_t |
197 | 0 | compress_read(zip_source_t *src, struct context *ctx, void *data, zip_uint64_t len) { |
198 | 0 | zip_compression_status_t ret; |
199 | 0 | bool end; |
200 | 0 | zip_int64_t n; |
201 | 0 | zip_uint64_t out_offset; |
202 | 0 | zip_uint64_t out_len; |
203 | |
|
204 | 0 | if (zip_error_code_zip(&ctx->error) != ZIP_ER_OK) { |
205 | 0 | return -1; |
206 | 0 | } |
207 | | |
208 | 0 | if (len == 0 || ctx->end_of_stream) { |
209 | 0 | return 0; |
210 | 0 | } |
211 | | |
212 | 0 | out_offset = 0; |
213 | |
|
214 | 0 | end = false; |
215 | 0 | while (!end && out_offset < len) { |
216 | 0 | out_len = len - out_offset; |
217 | 0 | ret = ctx->algorithm->process(ctx->ud, (zip_uint8_t *)data + out_offset, &out_len); |
218 | |
|
219 | 0 | if (ret != ZIP_COMPRESSION_ERROR) { |
220 | 0 | out_offset += out_len; |
221 | 0 | } |
222 | |
|
223 | 0 | switch (ret) { |
224 | 0 | case ZIP_COMPRESSION_END: |
225 | 0 | ctx->end_of_stream = true; |
226 | |
|
227 | 0 | if (!ctx->end_of_input) { |
228 | 0 | n = zip_source_read(src, ctx->buffer, 1); |
229 | 0 | if (n < 0) { |
230 | 0 | zip_error_set_from_source(&ctx->error, src); |
231 | 0 | end = true; |
232 | 0 | break; |
233 | 0 | } |
234 | 0 | else if (n == 0) { |
235 | 0 | ctx->end_of_input = true; |
236 | 0 | n = ctx->algorithm->end_of_input(ctx->ud) ? 1 : 0; |
237 | 0 | } |
238 | | |
239 | 0 | if (n > 0 && ctx->check_consistency) { |
240 | | /* garbage after stream, or compression ended before all data read */ |
241 | 0 | zip_error_set(&ctx->error, ZIP_ER_INCONS, ZIP_ER_DETAIL_COMPRESSED_DATA_TRAILING_GARBAGE); |
242 | 0 | end = true; |
243 | 0 | break; |
244 | 0 | } |
245 | 0 | } |
246 | | |
247 | 0 | if (ctx->first_read < 0) { |
248 | | /* we got end of processed stream before reading any input data */ |
249 | 0 | zip_error_set(&ctx->error, ZIP_ER_INTERNAL, 0); |
250 | 0 | end = true; |
251 | 0 | break; |
252 | 0 | } |
253 | 0 | if (ctx->can_store && (zip_uint64_t)ctx->first_read <= out_offset) { |
254 | 0 | ctx->is_stored = true; |
255 | 0 | ctx->size = (zip_uint64_t)ctx->first_read; |
256 | 0 | (void)memcpy_s(data, ctx->size, ctx->buffer, ctx->size); |
257 | 0 | return (zip_int64_t)ctx->size; |
258 | 0 | } |
259 | 0 | end = true; |
260 | 0 | break; |
261 | | |
262 | 0 | case ZIP_COMPRESSION_OK: |
263 | 0 | break; |
264 | | |
265 | 0 | case ZIP_COMPRESSION_NEED_DATA: |
266 | 0 | if (ctx->end_of_input) { |
267 | | /* TODO: error: stream not ended, but no more input */ |
268 | 0 | end = true; |
269 | 0 | break; |
270 | 0 | } |
271 | | |
272 | 0 | if ((n = zip_source_read(src, ctx->buffer, sizeof(ctx->buffer))) < 0) { |
273 | 0 | zip_error_set_from_source(&ctx->error, src); |
274 | 0 | end = true; |
275 | 0 | break; |
276 | 0 | } |
277 | 0 | else if (n == 0) { |
278 | 0 | ctx->end_of_input = true; |
279 | 0 | ctx->algorithm->end_of_input(ctx->ud); |
280 | 0 | if (ctx->first_read < 0) { |
281 | 0 | ctx->first_read = 0; |
282 | 0 | } |
283 | 0 | } |
284 | 0 | else { |
285 | 0 | if (ctx->first_read >= 0) { |
286 | | /* we overwrote a previously filled ctx->buffer */ |
287 | 0 | ctx->can_store = false; |
288 | 0 | } |
289 | 0 | else { |
290 | 0 | ctx->first_read = n; |
291 | 0 | } |
292 | |
|
293 | 0 | ctx->algorithm->input(ctx->ud, ctx->buffer, (zip_uint64_t)n); |
294 | 0 | } |
295 | 0 | break; |
296 | | |
297 | 0 | case ZIP_COMPRESSION_ERROR: |
298 | | /* error set by algorithm */ |
299 | 0 | if (zip_error_code_zip(&ctx->error) == ZIP_ER_OK) { |
300 | 0 | zip_error_set(&ctx->error, ZIP_ER_INTERNAL, 0); |
301 | 0 | } |
302 | 0 | end = true; |
303 | 0 | break; |
304 | 0 | } |
305 | 0 | } |
306 | | |
307 | 0 | if (out_offset > 0) { |
308 | 0 | ctx->can_store = false; |
309 | 0 | ctx->size += out_offset; |
310 | 0 | return (zip_int64_t)out_offset; |
311 | 0 | } |
312 | | |
313 | 0 | return (zip_error_code_zip(&ctx->error) == ZIP_ER_OK) ? 0 : -1; |
314 | 0 | } |
315 | | |
316 | | |
317 | | static zip_int64_t |
318 | 0 | compress_callback(zip_source_t *src, void *ud, void *data, zip_uint64_t len, zip_source_cmd_t cmd) { |
319 | 0 | struct context *ctx; |
320 | |
|
321 | 0 | ctx = (struct context *)ud; |
322 | |
|
323 | 0 | switch (cmd) { |
324 | 0 | case ZIP_SOURCE_OPEN: { |
325 | 0 | zip_stat_t st; |
326 | 0 | zip_file_attributes_t attributes; |
327 | | |
328 | 0 | ctx->size = 0; |
329 | 0 | ctx->end_of_input = false; |
330 | 0 | ctx->end_of_stream = false; |
331 | 0 | ctx->is_stored = false; |
332 | 0 | ctx->first_read = -1; |
333 | | |
334 | 0 | if (zip_source_stat(src, &st) < 0 || zip_source_get_file_attributes(src, &attributes) < 0) { |
335 | 0 | zip_error_set_from_source(&ctx->error, src); |
336 | 0 | return -1; |
337 | 0 | } |
338 | | |
339 | 0 | if (!ctx->algorithm->start(ctx->ud, &st, &attributes)) { |
340 | 0 | return -1; |
341 | 0 | } |
342 | | |
343 | 0 | return 0; |
344 | 0 | } |
345 | | |
346 | 0 | case ZIP_SOURCE_READ: |
347 | 0 | return compress_read(src, ctx, data, len); |
348 | | |
349 | 0 | case ZIP_SOURCE_CLOSE: |
350 | 0 | if (!ctx->algorithm->end(ctx->ud)) { |
351 | 0 | return -1; |
352 | 0 | } |
353 | 0 | return 0; |
354 | | |
355 | 0 | case ZIP_SOURCE_STAT: { |
356 | 0 | zip_stat_t *st; |
357 | |
|
358 | 0 | st = (zip_stat_t *)data; |
359 | |
|
360 | 0 | if (ctx->compress) { |
361 | 0 | if (ctx->end_of_stream) { |
362 | 0 | st->comp_method = ctx->is_stored ? ZIP_CM_STORE : ZIP_CM_ACTUAL(ctx->method); |
363 | 0 | st->comp_size = ctx->size; |
364 | 0 | st->valid |= ZIP_STAT_COMP_SIZE | ZIP_STAT_COMP_METHOD; |
365 | 0 | } |
366 | 0 | else { |
367 | 0 | st->valid &= ~(ZIP_STAT_COMP_SIZE | ZIP_STAT_COMP_METHOD); |
368 | 0 | } |
369 | 0 | } |
370 | 0 | else { |
371 | 0 | st->comp_method = ZIP_CM_STORE; |
372 | 0 | st->valid |= ZIP_STAT_COMP_METHOD; |
373 | 0 | st->valid &= ~ZIP_STAT_COMP_SIZE; |
374 | 0 | if (ctx->end_of_stream) { |
375 | 0 | st->size = ctx->size; |
376 | 0 | st->valid |= ZIP_STAT_SIZE; |
377 | 0 | } |
378 | 0 | } |
379 | 0 | } |
380 | 0 | return 0; |
381 | | |
382 | 0 | case ZIP_SOURCE_ERROR: |
383 | 0 | return zip_error_to_data(&ctx->error, data, len); |
384 | | |
385 | 0 | case ZIP_SOURCE_FREE: |
386 | 0 | context_free(ctx); |
387 | 0 | return 0; |
388 | | |
389 | 0 | case ZIP_SOURCE_GET_FILE_ATTRIBUTES: { |
390 | 0 | zip_file_attributes_t *attributes = (zip_file_attributes_t *)data; |
391 | |
|
392 | 0 | if (len < sizeof(*attributes)) { |
393 | 0 | zip_error_set(&ctx->error, ZIP_ER_INVAL, 0); |
394 | 0 | return -1; |
395 | 0 | } |
396 | | |
397 | 0 | attributes->valid |= ZIP_FILE_ATTRIBUTES_VERSION_NEEDED | ZIP_FILE_ATTRIBUTES_GENERAL_PURPOSE_BIT_FLAGS; |
398 | 0 | attributes->version_needed = ctx->algorithm->version_needed; |
399 | 0 | attributes->general_purpose_bit_mask = ZIP_FILE_ATTRIBUTES_GENERAL_PURPOSE_BIT_FLAGS_ALLOWED_MASK; |
400 | 0 | attributes->general_purpose_bit_flags = (ctx->is_stored ? 0 : ctx->algorithm->general_purpose_bit_flags(ctx->ud)); |
401 | |
|
402 | 0 | return sizeof(*attributes); |
403 | 0 | } |
404 | | |
405 | 0 | case ZIP_SOURCE_SUPPORTS: |
406 | 0 | return ZIP_SOURCE_SUPPORTS_READABLE | zip_source_make_command_bitmap(ZIP_SOURCE_GET_FILE_ATTRIBUTES, ZIP_SOURCE_SUPPORTS_REOPEN, -1); |
407 | | |
408 | 0 | default: |
409 | 0 | return zip_source_pass_to_lower_layer(src, data, len, cmd); |
410 | 0 | } |
411 | 0 | } |