|
1
|
/* zran.c -- example of deflate stream indexing and random access |
|
2
|
* Copyright (C) 2005, 2012, 2018, 2023, 2024, 2025 Mark Adler |
|
3
|
* For conditions of distribution and use, see copyright notice in zlib.h |
|
4
|
* Version 1.7 16 May 2025 Mark Adler */ |
|
5
|
|
|
6
|
/* Version History: |
|
7
|
1.0 29 May 2005 First version |
|
8
|
1.1 29 Sep 2012 Fix memory reallocation error |
|
9
|
1.2 14 Oct 2018 Handle gzip streams with multiple members |
|
10
|
Add a header file to facilitate usage in applications |
|
11
|
1.3 18 Feb 2023 Permit raw deflate streams as well as zlib and gzip |
|
12
|
Permit crossing gzip member boundaries when extracting |
|
13
|
Support a size_t size when extracting (was an int) |
|
14
|
Do a binary search over the index for an access point |
|
15
|
Expose the access point type to enable save and load |
|
16
|
1.4 13 Apr 2023 Add a NOPRIME define to not use inflatePrime() |
|
17
|
1.5 4 Feb 2024 Set returned index to NULL on an index build error |
|
18
|
Stop decoding once request is satisfied |
|
19
|
Provide a reusable inflate engine in the index |
|
20
|
Allocate the dictionaries to reduce memory usage |
|
21
|
1.6 2 Aug 2024 Remove unneeded dependency on limits.h |
|
22
|
1.7 16 May 2025 Remove redundant frees of point list on error |
|
23
|
Clean out point list structure when freed |
|
24
|
*/ |
|
25
|
|
|
26
|
// Illustrate the use of Z_BLOCK, inflatePrime(), and inflateSetDictionary() |
|
27
|
// for random access of a compressed file. A file containing a raw deflate |
|
28
|
// stream is provided on the command line. The compressed stream is decoded in |
|
29
|
// its entirety, and an index built with access points about every SPAN bytes |
|
30
|
// in the uncompressed output. The compressed file is left open, and can then |
|
31
|
// be read randomly, having to decompress on the average SPAN/2 uncompressed |
|
32
|
// bytes before getting to the desired block of data. |
|
33
|
// |
|
34
|
// An access point can be created at the start of any deflate block, by saving |
|
35
|
// the starting file offset and bit of that block, and the 32K bytes of |
|
36
|
// uncompressed data that precede that block. Also the uncompressed offset of |
|
37
|
// that block is saved to provide a reference for locating a desired starting |
|
38
|
// point in the uncompressed stream. deflate_index_build() decompresses the |
|
39
|
// input raw deflate stream a block at a time, and at the end of each block |
|
40
|
// decides if enough uncompressed data has gone by to justify the creation of a |
|
41
|
// new access point. If so, that point is saved in a data structure that grows |
|
42
|
// as needed to accommodate the points. |
|
43
|
// |
|
44
|
// To use the index, an offset in the uncompressed data is provided, for which |
|
45
|
// the latest access point at or preceding that offset is located in the index. |
|
46
|
// The input file is positioned to the specified location in the index, and if |
|
47
|
// necessary the first few bits of the compressed data is read from the file. |
|
48
|
// inflate is initialized with those bits and the 32K of uncompressed data, and |
|
49
|
// decompression then proceeds until the desired offset in the file is reached. |
|
50
|
// Then decompression continues to read the requested uncompressed data from |
|
51
|
// the file. |
|
52
|
// |
|
53
|
// There is some fair bit of overhead to starting inflation for the random |
|
54
|
// access, mainly copying the 32K byte dictionary. If small pieces of the file |
|
55
|
// are being accessed, it would make sense to implement a cache to hold some |
|
56
|
// lookahead to avoid many calls to deflate_index_extract() for small lengths. |
|
57
|
// |
|
58
|
// Another way to build an index would be to use inflateCopy(). That would not |
|
59
|
// be constrained to have access points at block boundaries, but would require |
|
60
|
// more memory per access point, and could not be saved to a file due to the |
|
61
|
// use of pointers in the state. The approach here allows for storage of the |
|
62
|
// index in a file. |
|
63
|
|
|
64
|
#include <stdio.h> |
|
65
|
#include <stdlib.h> |
|
66
|
#include <string.h> |
|
67
|
#include "zlib.h" |
|
68
|
#include "zran.h" |
|
69
|
|
|
70
|
#define WINSIZE 32768U // sliding window size |
|
71
|
#define CHUNK 16384 // file input buffer size |
|
72
|
|
|
73
|
// See comments in zran.h. |
|
74
|
void deflate_index_free(struct deflate_index *index) { |
|
75
|
if (index != NULL) { |
|
76
|
while (index->have) |
|
77
|
free(index->list[--index->have].window); |
|
78
|
free(index->list); |
|
79
|
index->list = NULL; |
|
80
|
inflateEnd(&index->strm); |
|
81
|
free(index); |
|
82
|
} |
|
83
|
} |
|
84
|
|
|
85
|
// Add an access point to the list. If out of memory, return NULL. index->mode |
|
86
|
// is temporarily the allocated number of access points, until it is time for |
|
87
|
// deflate_index_build() to return. Then index->mode is set to the mode of |
|
88
|
// inflation. |
|
89
|
static struct deflate_index *add_point(struct deflate_index *index, off_t in, |
|
90
|
off_t out, off_t beg, |
|
91
|
unsigned char *window) { |
|
92
|
if (index->have == index->mode) { |
|
93
|
// The list is full. Make it bigger. |
|
94
|
index->mode = index->mode ? index->mode << 1 : 8; |
|
95
|
point_t *next = realloc(index->list, sizeof(point_t) * index->mode); |
|
96
|
if (next == NULL) |
|
97
|
return NULL; |
|
98
|
index->list = next; |
|
99
|
} |
|
100
|
|
|
101
|
// Fill in the access point and increment how many we have. |
|
102
|
point_t *next = (point_t *)(index->list) + index->have++; |
|
103
|
if (index->have < 0) |
|
104
|
// Overflowed the int! |
|
105
|
return NULL; |
|
106
|
next->out = out; |
|
107
|
next->in = in; |
|
108
|
next->bits = index->strm.data_type & 7; |
|
109
|
next->dict = out - beg > WINSIZE ? WINSIZE : (unsigned)(out - beg); |
|
110
|
next->window = malloc(next->dict); |
|
111
|
if (next->window == NULL) |
|
112
|
return NULL; |
|
113
|
unsigned recent = WINSIZE - index->strm.avail_out; |
|
114
|
unsigned copy = recent > next->dict ? next->dict : recent; |
|
115
|
memcpy(next->window + next->dict - copy, window + recent - copy, copy); |
|
116
|
copy = next->dict - copy; |
|
117
|
memcpy(next->window, window + WINSIZE - copy, copy); |
|
118
|
|
|
119
|
// Return the index, which may have been newly allocated or destroyed. |
|
120
|
return index; |
|
121
|
} |
|
122
|
|
|
123
|
// Decompression modes. These are the inflateInit2() windowBits parameter. |
|
124
|
#define RAW -15 |
|
125
|
#define ZLIB 15 |
|
126
|
#define GZIP 31 |
|
127
|
|
|
128
|
// See comments in zran.h. |
|
129
|
int deflate_index_build(FILE *in, off_t span, struct deflate_index **built) { |
|
130
|
// If this returns with an error, any attempt to use the index will cleanly |
|
131
|
// return an error. |
|
132
|
*built = NULL; |
|
133
|
|
|
134
|
// Create and initialize the index list. |
|
135
|
struct deflate_index *index = malloc(sizeof(struct deflate_index)); |
|
136
|
if (index == NULL) |
|
137
|
return Z_MEM_ERROR; |
|
138
|
index->have = 0; |
|
139
|
index->mode = 0; // entries in index->list allocation |
|
140
|
index->list = NULL; |
|
141
|
index->strm.state = Z_NULL; // so inflateEnd() can work |
|
142
|
|
|
143
|
// Set up the inflation state. |
|
144
|
index->strm.avail_in = 0; |
|
145
|
index->strm.avail_out = 0; |
|
146
|
unsigned char buf[CHUNK]; // input buffer |
|
147
|
unsigned char win[WINSIZE] = {0}; // output sliding window |
|
148
|
off_t totin = 0; // total bytes read from input |
|
149
|
off_t totout = 0; // total bytes uncompressed |
|
150
|
off_t beg = 0; // starting offset of last history reset |
|
151
|
int mode = 0; // mode: RAW, ZLIB, or GZIP (0 => not set yet) |
|
152
|
|
|
153
|
// Decompress from in, generating access points along the way. |
|
154
|
int ret; // the return value from zlib, or Z_ERRNO |
|
155
|
off_t last; // last access point uncompressed offset |
|
156
|
do { |
|
157
|
// Assure available input, at least until reaching EOF. |
|
158
|
if (index->strm.avail_in == 0) { |
|
159
|
index->strm.avail_in = fread(buf, 1, sizeof(buf), in); |
|
160
|
totin += index->strm.avail_in; |
|
161
|
index->strm.next_in = buf; |
|
162
|
if (index->strm.avail_in < sizeof(buf) && ferror(in)) { |
|
163
|
ret = Z_ERRNO; |
|
164
|
break; |
|
165
|
} |
|
166
|
|
|
167
|
if (mode == 0) { |
|
168
|
// At the start of the input -- determine the type. Assume raw |
|
169
|
// if it is neither zlib nor gzip. This could in theory result |
|
170
|
// in a false positive for zlib, but in practice the fill bits |
|
171
|
// after a stored block are always zeros, so a raw stream won't |
|
172
|
// start with an 8 in the low nybble. |
|
173
|
mode = index->strm.avail_in == 0 ? RAW : // will fail |
|
174
|
(index->strm.next_in[0] & 0xf) == 8 ? ZLIB : |
|
175
|
index->strm.next_in[0] == 0x1f ? GZIP : |
|
176
|
/* else */ RAW; |
|
177
|
index->strm.zalloc = Z_NULL; |
|
178
|
index->strm.zfree = Z_NULL; |
|
179
|
index->strm.opaque = Z_NULL; |
|
180
|
ret = inflateInit2(&index->strm, mode); |
|
181
|
if (ret != Z_OK) |
|
182
|
break; |
|
183
|
} |
|
184
|
} |
|
185
|
|
|
186
|
// Assure available output. This rotates the output through, for use as |
|
187
|
// a sliding window on the uncompressed data. |
|
188
|
if (index->strm.avail_out == 0) { |
|
189
|
index->strm.avail_out = sizeof(win); |
|
190
|
index->strm.next_out = win; |
|
191
|
} |
|
192
|
|
|
193
|
if (mode == RAW && index->have == 0) |
|
194
|
// We skip the inflate() call at the start of raw deflate data in |
|
195
|
// order generate an access point there. Set data_type to imitate |
|
196
|
// the end of a header. |
|
197
|
index->strm.data_type = 0x80; |
|
198
|
else { |
|
199
|
// Inflate and update the number of uncompressed bytes. |
|
200
|
unsigned before = index->strm.avail_out; |
|
201
|
ret = inflate(&index->strm, Z_BLOCK); |
|
202
|
totout += before - index->strm.avail_out; |
|
203
|
} |
|
204
|
|
|
205
|
if ((index->strm.data_type & 0xc0) == 0x80 && |
|
206
|
(index->have == 0 || totout - last >= span)) { |
|
207
|
// We are at the end of a header or a non-last deflate block, so we |
|
208
|
// can add an access point here. Furthermore, we are either at the |
|
209
|
// very start for the first access point, or there has been span or |
|
210
|
// more uncompressed bytes since the last access point, so we want |
|
211
|
// to add an access point here. |
|
212
|
index = add_point(index, totin - index->strm.avail_in, totout, beg, |
|
213
|
win); |
|
214
|
if (index == NULL) { |
|
215
|
ret = Z_MEM_ERROR; |
|
216
|
break; |
|
217
|
} |
|
218
|
last = totout; |
|
219
|
} |
|
220
|
|
|
221
|
if (ret == Z_STREAM_END && mode == GZIP && |
|
222
|
(index->strm.avail_in || ungetc(getc(in), in) != EOF)) { |
|
223
|
// There is more input after the end of a gzip member. Reset the |
|
224
|
// inflate state to read another gzip member. On success, this will |
|
225
|
// set ret to Z_OK to continue decompressing. |
|
226
|
ret = inflateReset2(&index->strm, GZIP); |
|
227
|
beg = totout; // reset history |
|
228
|
} |
|
229
|
|
|
230
|
// Keep going until Z_STREAM_END or error. If the compressed data ends |
|
231
|
// prematurely without a file read error, Z_BUF_ERROR is returned. |
|
232
|
} while (ret == Z_OK); |
|
233
|
|
|
234
|
if (ret != Z_STREAM_END) { |
|
235
|
// An error was encountered. Discard the index and return a negative |
|
236
|
// error code. |
|
237
|
deflate_index_free(index); |
|
238
|
return ret == Z_NEED_DICT ? Z_DATA_ERROR : ret; |
|
239
|
} |
|
240
|
|
|
241
|
// Return the index. |
|
242
|
index->mode = mode; |
|
243
|
index->length = totout; |
|
244
|
*built = index; |
|
245
|
return index->have; |
|
246
|
} |
|
247
|
|
|
248
|
#ifdef NOPRIME |
|
249
|
// Support zlib versions before 1.2.3 (July 2005), or incomplete zlib clones |
|
250
|
// that do not have inflatePrime(). |
|
251
|
|
|
252
|
# define INFLATEPRIME inflatePreface |
|
253
|
|
|
254
|
// Append the low bits bits of value to in[] at bit position *have, updating |
|
255
|
// *have. value must be zero above its low bits bits. bits must be positive. |
|
256
|
// This assumes that any bits above the *have bits in the last byte are zeros. |
|
257
|
// That assumption is preserved on return, as any bits above *have + bits in |
|
258
|
// the last byte written will be set to zeros. |
|
259
|
static inline void append_bits(unsigned value, int bits, |
|
260
|
unsigned char *in, int *have) { |
|
261
|
in += *have >> 3; // where the first bits from value will go |
|
262
|
int k = *have & 7; // the number of bits already there |
|
263
|
*have += bits; |
|
264
|
if (k) |
|
265
|
*in |= value << k; // write value above the low k bits |
|
266
|
else |
|
267
|
*in = value; |
|
268
|
k = 8 - k; // the number of bits just appended |
|
269
|
while (bits > k) { |
|
270
|
value >>= k; // drop the bits appended |
|
271
|
bits -= k; |
|
272
|
k = 8; // now at a byte boundary |
|
273
|
*++in = value; |
|
274
|
} |
|
275
|
} |
|
276
|
|
|
277
|
// Insert enough bits in the form of empty deflate blocks in front of the |
|
278
|
// low bits bits of value, in order to bring the sequence to a byte boundary. |
|
279
|
// Then feed that to inflate(). This does what inflatePrime() does, except that |
|
280
|
// a negative value of bits is not supported. bits must be in 0..16. If the |
|
281
|
// arguments are invalid, Z_STREAM_ERROR is returned. Otherwise the return |
|
282
|
// value from inflate() is returned. |
|
283
|
static int inflatePreface(z_stream *strm, int bits, int value) { |
|
284
|
// Check input. |
|
285
|
if (strm == Z_NULL || bits < 0 || bits > 16) |
|
286
|
return Z_STREAM_ERROR; |
|
287
|
if (bits == 0) |
|
288
|
return Z_OK; |
|
289
|
value &= (2 << (bits - 1)) - 1; |
|
290
|
|
|
291
|
// An empty dynamic block with an odd number of bits (95). The high bit of |
|
292
|
// the last byte is unused. |
|
293
|
static const unsigned char dyn[] = { |
|
294
|
4, 0xe0, 0x81, 8, 0, 0, 0, 0, 0x20, 0xa8, 0xab, 0x1f |
|
295
|
}; |
|
296
|
const int dynlen = 95; // number of bits in the block |
|
297
|
|
|
298
|
// Build an input buffer for inflate that is a multiple of eight bits in |
|
299
|
// length, and that ends with the low bits bits of value. |
|
300
|
unsigned char in[(dynlen + 3 * 10 + 16 + 7) / 8]; |
|
301
|
int have = 0; |
|
302
|
if (bits & 1) { |
|
303
|
// Insert an empty dynamic block to get to an odd number of bits, so |
|
304
|
// when bits bits from value are appended, we are at an even number of |
|
305
|
// bits. |
|
306
|
memcpy(in, dyn, sizeof(dyn)); |
|
307
|
have = dynlen; |
|
308
|
} |
|
309
|
while ((have + bits) & 7) |
|
310
|
// Insert empty fixed blocks until appending bits bits would put us on |
|
311
|
// a byte boundary. This will insert at most three fixed blocks. |
|
312
|
append_bits(2, 10, in, &have); |
|
313
|
|
|
314
|
// Append the bits bits from value, which takes us to a byte boundary. |
|
315
|
append_bits(value, bits, in, &have); |
|
316
|
|
|
317
|
// Deliver the input to inflate(). There is no output space provided, but |
|
318
|
// inflate() can't get stuck waiting on output not ingesting all of the |
|
319
|
// provided input. The reason is that there will be at most 16 bits of |
|
320
|
// input from value after the empty deflate blocks (which themselves |
|
321
|
// generate no output). At least ten bits are needed to generate the first |
|
322
|
// output byte from a fixed block. The last two bytes of the buffer have to |
|
323
|
// be ingested in order to get ten bits, which is the most that value can |
|
324
|
// occupy. |
|
325
|
strm->avail_in = have >> 3; |
|
326
|
strm->next_in = in; |
|
327
|
strm->avail_out = 0; |
|
328
|
strm->next_out = in; // not used, but can't be NULL |
|
329
|
return inflate(strm, Z_NO_FLUSH); |
|
330
|
} |
|
331
|
|
|
332
|
#else |
|
333
|
# define INFLATEPRIME inflatePrime |
|
334
|
#endif |
|
335
|
|
|
336
|
// See comments in zran.h. |
|
337
|
ptrdiff_t deflate_index_extract(FILE *in, struct deflate_index *index, |
|
338
|
off_t offset, unsigned char *buf, size_t len) { |
|
339
|
// Do a quick sanity check on the index. |
|
340
|
if (index == NULL || index->have < 1 || index->list[0].out != 0 || |
|
341
|
index->strm.state == Z_NULL) |
|
342
|
return Z_STREAM_ERROR; |
|
343
|
|
|
344
|
// If nothing to extract, return zero bytes extracted. |
|
345
|
if (len == 0 || offset < 0 || offset >= index->length) |
|
346
|
return 0; |
|
347
|
|
|
348
|
// Find the access point closest to but not after offset. |
|
349
|
int lo = -1, hi = index->have; |
|
350
|
point_t *point = index->list; |
|
351
|
while (hi - lo > 1) { |
|
352
|
int mid = (lo + hi) >> 1; |
|
353
|
if (offset < point[mid].out) |
|
354
|
hi = mid; |
|
355
|
else |
|
356
|
lo = mid; |
|
357
|
} |
|
358
|
point += lo; |
|
359
|
|
|
360
|
// Initialize the input file and prime the inflate engine to start there. |
|
361
|
int ret = fseeko(in, point->in - (point->bits ? 1 : 0), SEEK_SET); |
|
362
|
if (ret == -1) |
|
363
|
return Z_ERRNO; |
|
364
|
int ch = 0; |
|
365
|
if (point->bits && (ch = getc(in)) == EOF) |
|
366
|
return ferror(in) ? Z_ERRNO : Z_BUF_ERROR; |
|
367
|
index->strm.avail_in = 0; |
|
368
|
ret = inflateReset2(&index->strm, RAW); |
|
369
|
if (ret != Z_OK) |
|
370
|
return ret; |
|
371
|
if (point->bits) |
|
372
|
INFLATEPRIME(&index->strm, point->bits, ch >> (8 - point->bits)); |
|
373
|
inflateSetDictionary(&index->strm, point->window, point->dict); |
|
374
|
|
|
375
|
// Skip uncompressed bytes until offset reached, then satisfy request. |
|
376
|
unsigned char input[CHUNK]; |
|
377
|
unsigned char discard[WINSIZE]; |
|
378
|
offset -= point->out; // number of bytes to skip to get to offset |
|
379
|
size_t left = len; // number of bytes left to read after offset |
|
380
|
do { |
|
381
|
if (offset) { |
|
382
|
// Discard up to offset uncompressed bytes. |
|
383
|
index->strm.avail_out = offset < WINSIZE ? (unsigned)offset : |
|
384
|
WINSIZE; |
|
385
|
index->strm.next_out = discard; |
|
386
|
} |
|
387
|
else { |
|
388
|
// Uncompress up to left bytes into buf. |
|
389
|
index->strm.avail_out = left < (unsigned)-1 ? (unsigned)left : |
|
390
|
(unsigned)-1; |
|
391
|
index->strm.next_out = buf + len - left; |
|
392
|
} |
|
393
|
|
|
394
|
// Uncompress, setting got to the number of bytes uncompressed. |
|
395
|
if (index->strm.avail_in == 0) { |
|
396
|
// Assure available input. |
|
397
|
index->strm.avail_in = fread(input, 1, CHUNK, in); |
|
398
|
if (index->strm.avail_in < CHUNK && ferror(in)) { |
|
399
|
ret = Z_ERRNO; |
|
400
|
break; |
|
401
|
} |
|
402
|
index->strm.next_in = input; |
|
403
|
} |
|
404
|
unsigned got = index->strm.avail_out; |
|
405
|
ret = inflate(&index->strm, Z_NO_FLUSH); |
|
406
|
got -= index->strm.avail_out; |
|
407
|
|
|
408
|
// Update the appropriate count. |
|
409
|
if (offset) |
|
410
|
offset -= got; |
|
411
|
else { |
|
412
|
left -= got; |
|
413
|
if (left == 0) |
|
414
|
// Request satisfied. |
|
415
|
break; |
|
416
|
} |
|
417
|
|
|
418
|
// If we're at the end of a gzip member and there's more to read, |
|
419
|
// continue to the next gzip member. |
|
420
|
if (ret == Z_STREAM_END && index->mode == GZIP) { |
|
421
|
// Discard the gzip trailer. |
|
422
|
unsigned drop = 8; // length of gzip trailer |
|
423
|
if (index->strm.avail_in >= drop) { |
|
424
|
index->strm.avail_in -= drop; |
|
425
|
index->strm.next_in += drop; |
|
426
|
} |
|
427
|
else { |
|
428
|
// Read and discard the remainder of the gzip trailer. |
|
429
|
drop -= index->strm.avail_in; |
|
430
|
index->strm.avail_in = 0; |
|
431
|
do { |
|
432
|
if (getc(in) == EOF) |
|
433
|
// The input does not have a complete trailer. |
|
434
|
return ferror(in) ? Z_ERRNO : Z_BUF_ERROR; |
|
435
|
} while (--drop); |
|
436
|
} |
|
437
|
|
|
438
|
if (index->strm.avail_in || ungetc(getc(in), in) != EOF) { |
|
439
|
// There's more after the gzip trailer. Use inflate to skip the |
|
440
|
// gzip header and resume the raw inflate there. |
|
441
|
inflateReset2(&index->strm, GZIP); |
|
442
|
do { |
|
443
|
if (index->strm.avail_in == 0) { |
|
444
|
index->strm.avail_in = fread(input, 1, CHUNK, in); |
|
445
|
if (index->strm.avail_in < CHUNK && ferror(in)) { |
|
446
|
ret = Z_ERRNO; |
|
447
|
break; |
|
448
|
} |
|
449
|
index->strm.next_in = input; |
|
450
|
} |
|
451
|
index->strm.avail_out = WINSIZE; |
|
452
|
index->strm.next_out = discard; |
|
453
|
ret = inflate(&index->strm, Z_BLOCK); // stop after header |
|
454
|
} while (ret == Z_OK && (index->strm.data_type & 0x80) == 0); |
|
455
|
if (ret != Z_OK) |
|
456
|
break; |
|
457
|
inflateReset2(&index->strm, RAW); |
|
458
|
} |
|
459
|
} |
|
460
|
|
|
461
|
// Continue until we have the requested data, the deflate data has |
|
462
|
// ended, or an error is encountered. |
|
463
|
} while (ret == Z_OK); |
|
464
|
|
|
465
|
// Return the number of uncompressed bytes read into buf, or the error. |
|
466
|
return ret == Z_OK || ret == Z_STREAM_END ? len - left : ret; |
|
467
|
} |
|
468
|
|
|
469
|
#ifdef TEST |
|
470
|
|
|
471
|
#define SPAN 1048576L // desired distance between access points |
|
472
|
#define LEN 16384 // number of bytes to extract |
|
473
|
|
|
474
|
// Demonstrate the use of deflate_index_build() and deflate_index_extract() by |
|
475
|
// processing the file provided on the command line, and extracting LEN bytes |
|
476
|
// from 2/3rds of the way through the uncompressed output, writing that to |
|
477
|
// stdout. An offset can be provided as the second argument, in which case the |
|
478
|
// data is extracted from there instead. |
|
479
|
int main(int argc, char **argv) { |
|
480
|
// Open the input file. |
|
481
|
if (argc < 2 || argc > 3) { |
|
482
|
fprintf(stderr, "usage: zran file.raw [offset]\n"); |
|
483
|
return 1; |
|
484
|
} |
|
485
|
FILE *in = fopen(argv[1], "rb"); |
|
486
|
if (in == NULL) { |
|
487
|
fprintf(stderr, "zran: could not open %s for reading\n", argv[1]); |
|
488
|
return 1; |
|
489
|
} |
|
490
|
|
|
491
|
// Get optional offset. |
|
492
|
off_t offset = -1; |
|
493
|
if (argc == 3) { |
|
494
|
char *end; |
|
495
|
offset = strtoll(argv[2], &end, 10); |
|
496
|
if (*end || offset < 0) { |
|
497
|
fprintf(stderr, "zran: %s is not a valid offset\n", argv[2]); |
|
498
|
return 1; |
|
499
|
} |
|
500
|
} |
|
501
|
|
|
502
|
// Build index. |
|
503
|
struct deflate_index *index = NULL; |
|
504
|
int len = deflate_index_build(in, SPAN, &index); |
|
505
|
if (len < 0) { |
|
506
|
fclose(in); |
|
507
|
switch (len) { |
|
508
|
case Z_MEM_ERROR: |
|
509
|
fprintf(stderr, "zran: out of memory\n"); |
|
510
|
break; |
|
511
|
case Z_BUF_ERROR: |
|
512
|
fprintf(stderr, "zran: %s ended prematurely\n", argv[1]); |
|
513
|
break; |
|
514
|
case Z_DATA_ERROR: |
|
515
|
fprintf(stderr, "zran: compressed data error in %s\n", argv[1]); |
|
516
|
break; |
|
517
|
case Z_ERRNO: |
|
518
|
fprintf(stderr, "zran: read error on %s\n", argv[1]); |
|
519
|
break; |
|
520
|
default: |
|
521
|
fprintf(stderr, "zran: error %d while building index\n", len); |
|
522
|
} |
|
523
|
return 1; |
|
524
|
} |
|
525
|
fprintf(stderr, "zran: built index with %d access points\n", len); |
|
526
|
|
|
527
|
// Use index by reading some bytes from an arbitrary offset. |
|
528
|
unsigned char buf[LEN]; |
|
529
|
if (offset == -1) |
|
530
|
offset = ((index->length + 1) << 1) / 3; |
|
531
|
ptrdiff_t got = deflate_index_extract(in, index, offset, buf, LEN); |
|
532
|
if (got < 0) |
|
533
|
fprintf(stderr, "zran: extraction failed: %s error\n", |
|
534
|
got == Z_MEM_ERROR ? "out of memory" : "input corrupted"); |
|
535
|
else { |
|
536
|
fwrite(buf, 1, got, stdout); |
|
537
|
fprintf(stderr, "zran: extracted %ld bytes at %lld\n", got, offset); |
|
538
|
} |
|
539
|
|
|
540
|
// Clean up and exit. |
|
541
|
deflate_index_free(index); |
|
542
|
fclose(in); |
|
543
|
return 0; |
|
544
|
} |
|
545
|
|
|
546
|
#endif |
|
547
|
|