contrib/libarchive-2/libarchive/archive_read_support_compression_bzip2.c

   1 /*-
   2  * Copyright (c) 2003-2007 Tim Kientzle
   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 #include "archive_platform.h"
  27
  28 __FBSDID("$FreeBSD: src/lib/libarchive/archive_read_support_compression_bzip2.c,v 1.17 2008/02/19 05:44:59 kientzle Exp $");
  29
  30 #ifdef HAVE_ERRNO_H
  31 #include <errno.h>
  32 #endif
  33 #include <stdio.h>
  34 #ifdef HAVE_STDLIB_H
  35 #include <stdlib.h>
  36 #endif
  37 #ifdef HAVE_STRING_H
  38 #include <string.h>
  39 #endif
  40 #ifdef HAVE_UNISTD_H
  41 #include <unistd.h>
  42 #endif
  43 #ifdef HAVE_BZLIB_H
  44 #include <bzlib.h>
  45 #endif
  46
  47 #include "archive.h"
  48 #include "archive_private.h"
  49 #include "archive_read_private.h"
  50
  51 #if HAVE_BZLIB_H
  52 struct private_data {
  53         bz_stream        stream;
  54         char            *uncompressed_buffer;
  55         size_t           uncompressed_buffer_size;
  56         char            *read_next;
  57         int64_t          total_out;
  58         char             eof; /* True = found end of compressed data. */
  59 };
  60
  61 static int      finish(struct archive_read *);
  62 static ssize_t  read_ahead(struct archive_read *, const void **, size_t);
  63 static ssize_t  read_consume(struct archive_read *, size_t);
  64 static int      drive_decompressor(struct archive_read *a, struct private_data *);
  65 #endif
  66
  67 /* These two functions are defined even if we lack the library.  See below. */
  68 static int      bid(const void *, size_t);
  69 static int      init(struct archive_read *, const void *, size_t);
  70
  71 int
  72 archive_read_support_compression_bzip2(struct archive *_a)
  73 {
  74         struct archive_read *a = (struct archive_read *)_a;
  75         if (__archive_read_register_compression(a, bid, init) != NULL)
  76                 return (ARCHIVE_OK);
  77         return (ARCHIVE_FATAL);
  78 }
  79
  80 /*
  81  * Test whether we can handle this data.
  82  *
  83  * This logic returns zero if any part of the signature fails.  It
  84  * also tries to Do The Right Thing if a very short buffer prevents us
  85  * from verifying as much as we would like.
  86  */
  87 static int
  88 bid(const void *buff, size_t len)
  89 {
  90         const unsigned char *buffer;
  91         int bits_checked;
  92
  93         if (len < 1)
  94                 return (0);
  95
  96         buffer = (const unsigned char *)buff;
  97         bits_checked = 0;
  98         if (buffer[0] != 'B')   /* Verify first ID byte. */
  99                 return (0);
 100         bits_checked += 8;
 101         if (len < 2)
 102                 return (bits_checked);
 103
 104         if (buffer[1] != 'Z')   /* Verify second ID byte. */
 105                 return (0);
 106         bits_checked += 8;
 107         if (len < 3)
 108                 return (bits_checked);
 109
 110         if (buffer[2] != 'h')   /* Verify third ID byte. */
 111                 return (0);
 112         bits_checked += 8;
 113         if (len < 4)
 114                 return (bits_checked);
 115
 116         if (buffer[3] < '1' || buffer[3] > '9')
 117                 return (0);
 118         bits_checked += 5;
 119         if (len < 5)
 120                 return (bits_checked);
 121
 122         /* After BZh[1-9], there must be either a data block
 123          * which begins with 0x314159265359 or an end-of-data
 124          * marker of 0x177245385090. */
 125
 126         if (buffer[4] == 0x31) {
 127                 /* Verify the data block signature. */
 128                 size_t s = len;
 129                 if (s > 10) s = 10;
 130                 if (memcmp(buffer + 4, "\x31\x41\x59\x26\x53\x59", s - 4) != 0)
 131                         return (0);
 132                 bits_checked += 8 * (s - 4);
 133         } else if (buffer[4] == 0x17) {
 134                 /* Verify the end-of-data marker. */
 135                 size_t s = len;
 136                 if (s > 10) s = 10;
 137                 if (memcmp(buffer + 4, "\x17\x72\x45\x38\x50\x90", s - 4) != 0)
 138                         return (0);
 139                 bits_checked += 8 * (s - 4);
 140         } else
 141                 return (0);
 142
 143         return (bits_checked);
 144 }
 145
 146 #ifndef HAVE_BZLIB_H
 147
 148 /*
 149  * If we don't have the library on this system, we can't actually do the
 150  * decompression.  We can, however, still detect compressed archives
 151  * and emit a useful message.
 152  */
 153 static int
 154 init(struct archive_read *a, const void *buff, size_t n)
 155 {
 156         (void)a;        /* UNUSED */
 157         (void)buff;     /* UNUSED */
 158         (void)n;        /* UNUSED */
 159
 160         archive_set_error(&a->archive, -1,
 161             "This version of libarchive was compiled without bzip2 support");
 162         return (ARCHIVE_FATAL);
 163 }
 164
 165
 166 #else
 167
 168 /*
 169  * Setup the callbacks.
 170  */
 171 static int
 172 init(struct archive_read *a, const void *buff, size_t n)
 173 {
 174         struct private_data *state;
 175         int ret;
 176
 177         a->archive.compression_code = ARCHIVE_COMPRESSION_BZIP2;
 178         a->archive.compression_name = "bzip2";
 179
 180         state = (struct private_data *)malloc(sizeof(*state));
 181         if (state == NULL) {
 182                 archive_set_error(&a->archive, ENOMEM,
 183                     "Can't allocate data for %s decompression",
 184                     a->archive.compression_name);
 185                 return (ARCHIVE_FATAL);
 186         }
 187         memset(state, 0, sizeof(*state));
 188
 189         state->uncompressed_buffer_size = 64 * 1024;
 190         state->uncompressed_buffer = (char *)malloc(state->uncompressed_buffer_size);
 191         state->stream.next_out = state->uncompressed_buffer;
 192         state->read_next = state->uncompressed_buffer;
 193         state->stream.avail_out = state->uncompressed_buffer_size;
 194
 195         if (state->uncompressed_buffer == NULL) {
 196                 archive_set_error(&a->archive, ENOMEM,
 197                     "Can't allocate %s decompression buffers",
 198                     a->archive.compression_name);
 199                 free(state);
 200                 return (ARCHIVE_FATAL);
 201         }
 202
 203         /*
 204          * A bug in bzlib.h: stream.next_in should be marked 'const'
 205          * but isn't (the library never alters data through the
 206          * next_in pointer, only reads it).  The result: this ugly
 207          * cast to remove 'const'.
 208          */
 209         state->stream.next_in = (char *)(uintptr_t)(const void *)buff;
 210         state->stream.avail_in = n;
 211
 212         a->decompressor->read_ahead = read_ahead;
 213         a->decompressor->consume = read_consume;
 214         a->decompressor->skip = NULL; /* not supported */
 215         a->decompressor->finish = finish;
 216
 217         /* Initialize compression library. */
 218         ret = BZ2_bzDecompressInit(&(state->stream),
 219             0 /* library verbosity */,
 220             0 /* don't use slow low-mem algorithm */);
 221
 222         /* If init fails, try using low-memory algorithm instead. */
 223         if (ret == BZ_MEM_ERROR) {
 224                 ret = BZ2_bzDecompressInit(&(state->stream),
 225                     0 /* library verbosity */,
 226                     1 /* do use slow low-mem algorithm */);
 227         }
 228
 229         if (ret == BZ_OK) {
 230                 a->decompressor->data = state;
 231                 return (ARCHIVE_OK);
 232         }
 233
 234         /* Library setup failed: Clean up. */
 235         archive_set_error(&a->archive, ARCHIVE_ERRNO_MISC,
 236             "Internal error initializing %s library",
 237             a->archive.compression_name);
 238         free(state->uncompressed_buffer);
 239         free(state);
 240
 241         /* Override the error message if we know what really went wrong. */
 242         switch (ret) {
 243         case BZ_PARAM_ERROR:
 244                 archive_set_error(&a->archive, ARCHIVE_ERRNO_MISC,
 245                     "Internal error initializing compression library: "
 246                     "invalid setup parameter");
 247                 break;
 248         case BZ_MEM_ERROR:
 249                 archive_set_error(&a->archive, ENOMEM,
 250                     "Internal error initializing compression library: "
 251                     "out of memory");
 252                 break;
 253         case BZ_CONFIG_ERROR:
 254                 archive_set_error(&a->archive, ARCHIVE_ERRNO_MISC,
 255                     "Internal error initializing compression library: "
 256                     "mis-compiled library");
 257                 break;
 258         }
 259
 260         return (ARCHIVE_FATAL);
 261 }
 262
 263 /*
 264  * Return a block of data from the decompression buffer.  Decompress more
 265  * as necessary.
 266  */
 267 static ssize_t
 268 read_ahead(struct archive_read *a, const void **p, size_t min)
 269 {
 270         struct private_data *state;
 271         size_t read_avail, was_avail;
 272         int ret;
 273
 274         state = (struct private_data *)a->decompressor->data;
 275         if (!a->client_reader) {
 276                 archive_set_error(&a->archive, ARCHIVE_ERRNO_PROGRAMMER,
 277                     "No read callback is registered?  "
 278                     "This is probably an internal programming error.");
 279                 return (ARCHIVE_FATAL);
 280         }
 281
 282         read_avail = state->stream.next_out - state->read_next;
 283
 284         if (read_avail + state->stream.avail_out < min) {
 285                 memmove(state->uncompressed_buffer, state->read_next,
 286                     read_avail);
 287                 state->read_next = state->uncompressed_buffer;
 288                 state->stream.next_out = state->read_next + read_avail;
 289                 state->stream.avail_out
 290                     = state->uncompressed_buffer_size - read_avail;
 291         }
 292
 293         while (read_avail < min &&              /* Haven't satisfied min. */
 294             read_avail < state->uncompressed_buffer_size) { /* !full */
 295                 was_avail = read_avail;
 296                 if ((ret = drive_decompressor(a, state)) < ARCHIVE_OK)
 297                         return (ret);
 298                 if (ret == ARCHIVE_EOF)
 299                         break; /* Break on EOF even if we haven't met min. */
 300                 read_avail = state->stream.next_out - state->read_next;
 301                 if (was_avail == read_avail) /* No progress? */
 302                         break;
 303         }
 304
 305         *p = state->read_next;
 306         return (read_avail);
 307 }
 308
 309 /*
 310  * Mark a previously-returned block of data as read.
 311  */
 312 static ssize_t
 313 read_consume(struct archive_read *a, size_t n)
 314 {
 315         struct private_data *state;
 316
 317         state = (struct private_data *)a->decompressor->data;
 318         a->archive.file_position += n;
 319         state->read_next += n;
 320         if (state->read_next > state->stream.next_out)
 321                 __archive_errx(1, "Request to consume too many "
 322                     "bytes from bzip2 decompressor");
 323         return (n);
 324 }
 325
 326 /*
 327  * Clean up the decompressor.
 328  */
 329 static int
 330 finish(struct archive_read *a)
 331 {
 332         struct private_data *state;
 333         int ret;
 334
 335         state = (struct private_data *)a->decompressor->data;
 336         ret = ARCHIVE_OK;
 337         switch (BZ2_bzDecompressEnd(&(state->stream))) {
 338         case BZ_OK:
 339                 break;
 340         default:
 341                 archive_set_error(&a->archive, ARCHIVE_ERRNO_MISC,
 342                     "Failed to clean up %s compressor",
 343                     a->archive.compression_name);
 344                 ret = ARCHIVE_FATAL;
 345         }
 346
 347         free(state->uncompressed_buffer);
 348         free(state);
 349
 350         a->decompressor->data = NULL;
 351         return (ret);
 352 }
 353
 354 /*
 355  * Utility function to pull data through decompressor, reading input
 356  * blocks as necessary.
 357  */
 358 static int
 359 drive_decompressor(struct archive_read *a, struct private_data *state)
 360 {
 361         ssize_t ret;
 362         int decompressed, total_decompressed;
 363         char *output;
 364         const void *read_buf;
 365
 366         if (state->eof)
 367                 return (ARCHIVE_EOF);
 368         total_decompressed = 0;
 369         for (;;) {
 370                 if (state->stream.avail_in == 0) {
 371                         read_buf = state->stream.next_in;
 372                         ret = (a->client_reader)(&a->archive, a->client_data,
 373                             &read_buf);
 374                         state->stream.next_in = (void *)(uintptr_t)read_buf;
 375                         if (ret < 0) {
 376                                 /*
 377                                  * TODO: Find a better way to handle
 378                                  * this read failure.
 379                                  */
 380                                 goto fatal;
 381                         }
 382                         if (ret == 0  &&  total_decompressed == 0) {
 383                                 archive_set_error(&a->archive, EIO,
 384                                     "Premature end of %s compressed data",
 385                                     a->archive.compression_name);
 386                                 return (ARCHIVE_FATAL);
 387                         }
 388                         a->archive.raw_position += ret;
 389                         state->stream.avail_in = ret;
 390                 }
 391
 392                 {
 393                         output = state->stream.next_out;
 394
 395                         /* Decompress some data. */
 396                         ret = BZ2_bzDecompress(&(state->stream));
 397                         decompressed = state->stream.next_out - output;
 398
 399                         /* Accumulate the total bytes of output. */
 400                         state->total_out += decompressed;
 401                         total_decompressed += decompressed;
 402
 403                         switch (ret) {
 404                         case BZ_OK: /* Decompressor made some progress. */
 405                                 if (decompressed > 0)
 406                                         return (ARCHIVE_OK);
 407                                 break;
 408                         case BZ_STREAM_END:     /* Found end of stream. */
 409                                 state->eof = 1;
 410                                 return (ARCHIVE_OK);
 411                         default:
 412                                 /* Any other return value is an error. */
 413                                 goto fatal;
 414                         }
 415                 }
 416         }
 417         return (ARCHIVE_OK);
 418
 419         /* Return a fatal error. */
 420 fatal:
 421         archive_set_error(&a->archive, ARCHIVE_ERRNO_MISC,
 422             "%s decompression failed", a->archive.compression_name);
 423         return (ARCHIVE_FATAL);
 424 }
 425
 426 #endif /* HAVE_BZLIB_H */