1 /*- 2 * Copyright (c) 2008 Joerg Sonnenberger 3 * All rights reserved. 4 * 5 * Redistribution and use in source and binary forms, with or without 6 * modification, are permitted provided that the following conditions 7 * are met: 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. 13 * 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. 24 */ 25 26 /*- 27 * Copyright (c) 1985, 1986, 1992, 1993 28 * The Regents of the University of California. All rights reserved. 29 * 30 * This code is derived from software contributed to Berkeley by 31 * Diomidis Spinellis and James A. Woods, derived from original 32 * work by Spencer Thomas and Joseph Orost. 33 * 34 * Redistribution and use in source and binary forms, with or without 35 * modification, are permitted provided that the following conditions 36 * are met: 37 * 1. Redistributions of source code must retain the above copyright 38 * notice, this list of conditions and the following disclaimer. 39 * 2. Redistributions in binary form must reproduce the above copyright 40 * notice, this list of conditions and the following disclaimer in the 41 * documentation and/or other materials provided with the distribution. 42 * 3. Neither the name of the University nor the names of its contributors 43 * may be used to endorse or promote products derived from this software 44 * without specific prior written permission. 45 * 46 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 47 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 48 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 49 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 50 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 51 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 52 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 53 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 54 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 55 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 56 * SUCH DAMAGE. 57 */ 58 59 #include "archive_platform.h" 60 61 __FBSDID("$FreeBSD: head/lib/libarchive/archive_write_set_compression_compress.c 201111 2009-12-28 03:33:05Z kientzle $"); 62 63 #ifdef HAVE_ERRNO_H 64 #include <errno.h> 65 #endif 66 #ifdef HAVE_STDLIB_H 67 #include <stdlib.h> 68 #endif 69 #ifdef HAVE_STRING_H 70 #include <string.h> 71 #endif 72 73 #include "archive.h" 74 #include "archive_private.h" 75 #include "archive_write_private.h" 76 77 #define HSIZE 69001 /* 95% occupancy */ 78 #define HSHIFT 8 /* 8 - trunc(log2(HSIZE / 65536)) */ 79 #define CHECK_GAP 10000 /* Ratio check interval. */ 80 81 #define MAXCODE(bits) ((1 << (bits)) - 1) 82 83 /* 84 * the next two codes should not be changed lightly, as they must not 85 * lie within the contiguous general code space. 86 */ 87 #define FIRST 257 /* First free entry. */ 88 #define CLEAR 256 /* Table clear output code. */ 89 90 struct private_data { 91 int64_t in_count, out_count, checkpoint; 92 93 int code_len; /* Number of bits/code. */ 94 int cur_maxcode; /* Maximum code, given n_bits. */ 95 int max_maxcode; /* Should NEVER generate this code. */ 96 int hashtab [HSIZE]; 97 unsigned short codetab [HSIZE]; 98 int first_free; /* First unused entry. */ 99 int compress_ratio; 100 101 int cur_code, cur_fcode; 102 103 int bit_offset; 104 unsigned char bit_buf; 105 106 unsigned char *compressed; 107 size_t compressed_buffer_size; 108 size_t compressed_offset; 109 }; 110 111 static int archive_compressor_compress_open(struct archive_write_filter *); 112 static int archive_compressor_compress_write(struct archive_write_filter *, 113 const void *, size_t); 114 static int archive_compressor_compress_close(struct archive_write_filter *); 115 static int archive_compressor_compress_free(struct archive_write_filter *); 116 117 #if ARCHIVE_VERSION_NUMBER < 4000000 118 int 119 archive_write_set_compression_compress(struct archive *a) 120 { 121 __archive_write_filters_free(a); 122 return (archive_write_add_filter_compress(a)); 123 } 124 #endif 125 126 /* 127 * Add a compress filter to this write handle. 128 */ 129 int 130 archive_write_add_filter_compress(struct archive *_a) 131 { 132 struct archive_write *a = (struct archive_write *)_a; 133 struct archive_write_filter *f = __archive_write_allocate_filter(_a); 134 135 archive_check_magic(&a->archive, ARCHIVE_WRITE_MAGIC, 136 ARCHIVE_STATE_NEW, "archive_write_add_filter_compress"); 137 f->open = &archive_compressor_compress_open; 138 f->code = ARCHIVE_FILTER_COMPRESS; 139 f->name = "compress"; 140 return (ARCHIVE_OK); 141 } 142 143 /* 144 * Setup callback. 145 */ 146 static int 147 archive_compressor_compress_open(struct archive_write_filter *f) 148 { 149 struct private_data *state; 150 size_t bs = 65536, bpb; 151 152 f->code = ARCHIVE_FILTER_COMPRESS; 153 f->name = "compress"; 154 155 state = (struct private_data *)calloc(1, sizeof(*state)); 156 if (state == NULL) { 157 archive_set_error(f->archive, ENOMEM, 158 "Can't allocate data for compression"); 159 return (ARCHIVE_FATAL); 160 } 161 162 if (f->archive->magic == ARCHIVE_WRITE_MAGIC) { 163 /* Buffer size should be a multiple number of the of bytes 164 * per block for performance. */ 165 bpb = archive_write_get_bytes_per_block(f->archive); 166 if (bpb > bs) 167 bs = bpb; 168 else if (bpb != 0) 169 bs -= bs % bpb; 170 } 171 state->compressed_buffer_size = bs; 172 state->compressed = malloc(state->compressed_buffer_size); 173 174 if (state->compressed == NULL) { 175 archive_set_error(f->archive, ENOMEM, 176 "Can't allocate data for compression buffer"); 177 free(state); 178 return (ARCHIVE_FATAL); 179 } 180 181 f->write = archive_compressor_compress_write; 182 f->close = archive_compressor_compress_close; 183 f->free = archive_compressor_compress_free; 184 185 state->max_maxcode = 0x10000; /* Should NEVER generate this code. */ 186 state->in_count = 0; /* Length of input. */ 187 state->bit_buf = 0; 188 state->bit_offset = 0; 189 state->out_count = 3; /* Includes 3-byte header mojo. */ 190 state->compress_ratio = 0; 191 state->checkpoint = CHECK_GAP; 192 state->code_len = 9; 193 state->cur_maxcode = MAXCODE(state->code_len); 194 state->first_free = FIRST; 195 196 memset(state->hashtab, 0xff, sizeof(state->hashtab)); 197 198 /* Prime output buffer with a gzip header. */ 199 state->compressed[0] = 0x1f; /* Compress */ 200 state->compressed[1] = 0x9d; 201 state->compressed[2] = 0x90; /* Block mode, 16bit max */ 202 state->compressed_offset = 3; 203 204 f->data = state; 205 return (0); 206 } 207 208 /*- 209 * Output the given code. 210 * Inputs: 211 * code: A n_bits-bit integer. If == -1, then EOF. This assumes 212 * that n_bits <= (long)wordsize - 1. 213 * Outputs: 214 * Outputs code to the file. 215 * Assumptions: 216 * Chars are 8 bits long. 217 * Algorithm: 218 * Maintain a BITS character long buffer (so that 8 codes will 219 * fit in it exactly). Use the VAX insv instruction to insert each 220 * code in turn. When the buffer fills up empty it and start over. 221 */ 222 223 static const unsigned char rmask[9] = 224 {0x00, 0x01, 0x03, 0x07, 0x0f, 0x1f, 0x3f, 0x7f, 0xff}; 225 226 static int 227 output_byte(struct archive_write_filter *f, unsigned char c) 228 { 229 struct private_data *state = f->data; 230 231 state->compressed[state->compressed_offset++] = c; 232 ++state->out_count; 233 234 if (state->compressed_buffer_size == state->compressed_offset) { 235 int ret = __archive_write_filter(f->next_filter, 236 state->compressed, state->compressed_buffer_size); 237 if (ret != ARCHIVE_OK) 238 return ARCHIVE_FATAL; 239 state->compressed_offset = 0; 240 } 241 242 return ARCHIVE_OK; 243 } 244 245 static int 246 output_code(struct archive_write_filter *f, int ocode) 247 { 248 struct private_data *state = f->data; 249 int bits, ret, clear_flg, bit_offset; 250 251 clear_flg = ocode == CLEAR; 252 253 /* 254 * Since ocode is always >= 8 bits, only need to mask the first 255 * hunk on the left. 256 */ 257 bit_offset = state->bit_offset % 8; 258 state->bit_buf |= (ocode << bit_offset) & 0xff; 259 output_byte(f, state->bit_buf); 260 261 bits = state->code_len - (8 - bit_offset); 262 ocode >>= 8 - bit_offset; 263 /* Get any 8 bit parts in the middle (<=1 for up to 16 bits). */ 264 if (bits >= 8) { 265 output_byte(f, ocode & 0xff); 266 ocode >>= 8; 267 bits -= 8; 268 } 269 /* Last bits. */ 270 state->bit_offset += state->code_len; 271 state->bit_buf = ocode & rmask[bits]; 272 if (state->bit_offset == state->code_len * 8) 273 state->bit_offset = 0; 274 275 /* 276 * If the next entry is going to be too big for the ocode size, 277 * then increase it, if possible. 278 */ 279 if (clear_flg || state->first_free > state->cur_maxcode) { 280 /* 281 * Write the whole buffer, because the input side won't 282 * discover the size increase until after it has read it. 283 */ 284 if (state->bit_offset > 0) { 285 while (state->bit_offset < state->code_len * 8) { 286 ret = output_byte(f, state->bit_buf); 287 if (ret != ARCHIVE_OK) 288 return ret; 289 state->bit_offset += 8; 290 state->bit_buf = 0; 291 } 292 } 293 state->bit_buf = 0; 294 state->bit_offset = 0; 295 296 if (clear_flg) { 297 state->code_len = 9; 298 state->cur_maxcode = MAXCODE(state->code_len); 299 } else { 300 state->code_len++; 301 if (state->code_len == 16) 302 state->cur_maxcode = state->max_maxcode; 303 else 304 state->cur_maxcode = MAXCODE(state->code_len); 305 } 306 } 307 308 return (ARCHIVE_OK); 309 } 310 311 static int 312 output_flush(struct archive_write_filter *f) 313 { 314 struct private_data *state = f->data; 315 int ret; 316 317 /* At EOF, write the rest of the buffer. */ 318 if (state->bit_offset % 8) { 319 state->code_len = (state->bit_offset % 8 + 7) / 8; 320 ret = output_byte(f, state->bit_buf); 321 if (ret != ARCHIVE_OK) 322 return ret; 323 } 324 325 return (ARCHIVE_OK); 326 } 327 328 /* 329 * Write data to the compressed stream. 330 */ 331 static int 332 archive_compressor_compress_write(struct archive_write_filter *f, 333 const void *buff, size_t length) 334 { 335 struct private_data *state = (struct private_data *)f->data; 336 int i; 337 int ratio; 338 int c, disp, ret; 339 const unsigned char *bp; 340 341 if (length == 0) 342 return ARCHIVE_OK; 343 344 bp = buff; 345 346 if (state->in_count == 0) { 347 state->cur_code = *bp++; 348 ++state->in_count; 349 --length; 350 } 351 352 while (length--) { 353 c = *bp++; 354 state->in_count++; 355 state->cur_fcode = (c << 16) + state->cur_code; 356 i = ((c << HSHIFT) ^ state->cur_code); /* Xor hashing. */ 357 358 if (state->hashtab[i] == state->cur_fcode) { 359 state->cur_code = state->codetab[i]; 360 continue; 361 } 362 if (state->hashtab[i] < 0) /* Empty slot. */ 363 goto nomatch; 364 /* Secondary hash (after G. Knott). */ 365 if (i == 0) 366 disp = 1; 367 else 368 disp = HSIZE - i; 369 probe: 370 if ((i -= disp) < 0) 371 i += HSIZE; 372 373 if (state->hashtab[i] == state->cur_fcode) { 374 state->cur_code = state->codetab[i]; 375 continue; 376 } 377 if (state->hashtab[i] >= 0) 378 goto probe; 379 nomatch: 380 ret = output_code(f, state->cur_code); 381 if (ret != ARCHIVE_OK) 382 return ret; 383 state->cur_code = c; 384 if (state->first_free < state->max_maxcode) { 385 state->codetab[i] = state->first_free++; /* code -> hashtable */ 386 state->hashtab[i] = state->cur_fcode; 387 continue; 388 } 389 if (state->in_count < state->checkpoint) 390 continue; 391 392 state->checkpoint = state->in_count + CHECK_GAP; 393 394 if (state->in_count <= 0x007fffff && state->out_count != 0) 395 ratio = (int)(state->in_count * 256 / state->out_count); 396 else if ((ratio = (int)(state->out_count / 256)) == 0) 397 ratio = 0x7fffffff; 398 else 399 ratio = (int)(state->in_count / ratio); 400 401 if (ratio > state->compress_ratio) 402 state->compress_ratio = ratio; 403 else { 404 state->compress_ratio = 0; 405 memset(state->hashtab, 0xff, sizeof(state->hashtab)); 406 state->first_free = FIRST; 407 ret = output_code(f, CLEAR); 408 if (ret != ARCHIVE_OK) 409 return ret; 410 } 411 } 412 413 return (ARCHIVE_OK); 414 } 415 416 417 /* 418 * Finish the compression... 419 */ 420 static int 421 archive_compressor_compress_close(struct archive_write_filter *f) 422 { 423 struct private_data *state = (struct private_data *)f->data; 424 int ret; 425 426 ret = output_code(f, state->cur_code); 427 if (ret != ARCHIVE_OK) 428 return ret; 429 ret = output_flush(f); 430 if (ret != ARCHIVE_OK) 431 return ret; 432 433 /* Write the last block */ 434 ret = __archive_write_filter(f->next_filter, 435 state->compressed, state->compressed_offset); 436 return (ret); 437 } 438 439 static int 440 archive_compressor_compress_free(struct archive_write_filter *f) 441 { 442 struct private_data *state = (struct private_data *)f->data; 443 444 free(state->compressed); 445 free(state); 446 return (ARCHIVE_OK); 447 } 448