1 /* 2 * Copyright (c) 1998 Michael Smith. 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 AND CONTRIBUTORS ``AS IS'' AND 15 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 16 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 17 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE 18 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 19 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 20 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 21 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 22 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 23 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 24 * SUCH DAMAGE. 25 */ 26 27 #include <sys/cdefs.h> 28 29 #include "stand.h" 30 31 #include <sys/stat.h> 32 #include <string.h> 33 #include <zlib.h> 34 35 #define Z_BUFSIZE 16384 /* match NFSREAD_MAX_SIZE */ 36 37 struct z_file 38 { 39 int zf_rawfd; 40 off_t zf_dataoffset; 41 z_stream zf_zstream; 42 int zf_endseen; 43 unsigned char zf_buf[Z_BUFSIZE]; 44 }; 45 46 static int zf_fill(struct z_file *z); 47 static int zf_open(const char *path, struct open_file *f); 48 static int zf_close(struct open_file *f); 49 static int zf_read(struct open_file *f, void *buf, size_t size, size_t *resid); 50 static off_t zf_seek(struct open_file *f, off_t offset, int where); 51 static int zf_stat(struct open_file *f, struct stat *sb); 52 53 struct fs_ops gzipfs_fsops = { 54 .fs_name = "zip", 55 .fo_open = zf_open, 56 .fo_close = zf_close, 57 .fo_read = zf_read, 58 .fo_write = null_write, 59 .fo_seek = zf_seek, 60 .fo_stat = zf_stat, 61 .fo_readdir = null_readdir 62 }; 63 64 static int 65 zf_fill(struct z_file *zf) 66 { 67 int result; 68 int req; 69 70 req = Z_BUFSIZE - zf->zf_zstream.avail_in; 71 result = 0; 72 73 /* If we need more */ 74 if (req > 0) { 75 /* move old data to bottom of buffer */ 76 if (req < Z_BUFSIZE) 77 bcopy(zf->zf_buf + req, zf->zf_buf, Z_BUFSIZE - req); 78 79 /* read to fill buffer and update availibility data */ 80 result = read(zf->zf_rawfd, 81 zf->zf_buf + zf->zf_zstream.avail_in, req); 82 zf->zf_zstream.next_in = zf->zf_buf; 83 if (result >= 0) 84 zf->zf_zstream.avail_in += result; 85 } 86 return (result); 87 } 88 89 /* 90 * Adapted from get_byte/check_header in libz 91 * 92 * Returns 0 if the header is OK, nonzero if not. 93 */ 94 static int 95 get_byte(struct z_file *zf, off_t *curoffp) 96 { 97 if ((zf->zf_zstream.avail_in == 0) && (zf_fill(zf) == -1)) 98 return (-1); 99 zf->zf_zstream.avail_in--; 100 ++*curoffp; 101 return (*(zf->zf_zstream.next_in)++); 102 } 103 104 static int gz_magic[2] = {0x1f, 0x8b}; /* gzip magic header */ 105 106 /* gzip flag byte */ 107 #define ASCII_FLAG 0x01 /* bit 0 set: file probably ascii text */ 108 #define HEAD_CRC 0x02 /* bit 1 set: header CRC present */ 109 #define EXTRA_FIELD 0x04 /* bit 2 set: extra field present */ 110 #define ORIG_NAME 0x08 /* bit 3 set: original file name present */ 111 #define COMMENT 0x10 /* bit 4 set: file comment present */ 112 #define RESERVED 0xE0 /* bits 5..7: reserved */ 113 114 static int 115 check_header(struct z_file *zf) 116 { 117 int method; /* method byte */ 118 int flags; /* flags byte */ 119 uInt len; 120 int c; 121 122 zf->zf_dataoffset = 0; 123 /* Check the gzip magic header */ 124 for (len = 0; len < 2; len++) { 125 c = get_byte(zf, &zf->zf_dataoffset); 126 if (c != gz_magic[len]) { 127 return (1); 128 } 129 } 130 method = get_byte(zf, &zf->zf_dataoffset); 131 flags = get_byte(zf, &zf->zf_dataoffset); 132 if (method != Z_DEFLATED || (flags & RESERVED) != 0) { 133 return (1); 134 } 135 136 /* Discard time, xflags and OS code: */ 137 for (len = 0; len < 6; len++) 138 (void) get_byte(zf, &zf->zf_dataoffset); 139 140 if ((flags & EXTRA_FIELD) != 0) { /* skip the extra field */ 141 len = (uInt)get_byte(zf, &zf->zf_dataoffset); 142 len += ((uInt)get_byte(zf, &zf->zf_dataoffset))<<8; 143 /* len is garbage if EOF but the loop below will quit anyway */ 144 while (len-- != 0 && get_byte(zf, &zf->zf_dataoffset) != -1) 145 ; 146 } 147 if ((flags & ORIG_NAME) != 0) { /* skip the original file name */ 148 while ((c = get_byte(zf, &zf->zf_dataoffset)) != 0 && c != -1) 149 ; 150 } 151 if ((flags & COMMENT) != 0) { /* skip the .gz file comment */ 152 while ((c = get_byte(zf, &zf->zf_dataoffset)) != 0 && c != -1) 153 ; 154 } 155 if ((flags & HEAD_CRC) != 0) { /* skip the header crc */ 156 for (len = 0; len < 2; len++) 157 c = get_byte(zf, &zf->zf_dataoffset); 158 } 159 /* if there's data left, we're in business */ 160 return ((c == -1) ? 1 : 0); 161 } 162 163 static int 164 zf_open(const char *fname, struct open_file *f) 165 { 166 char *zfname; 167 int rawfd; 168 struct z_file *zf; 169 char *cp; 170 int error; 171 struct stat sb; 172 size_t len; 173 174 /* Have to be in "just read it" mode */ 175 if (f->f_flags != F_READ) 176 return (EPERM); 177 178 /* directory names are not compressed. */ 179 len = strlen(fname); 180 if (len == 0 || fname[len - 1] == '/') 181 return (ENOENT); 182 183 /* If the name already ends in .gz or .bz2, ignore it */ 184 if ((cp = strrchr(fname, '.')) && (strcmp(cp, ".gz") == 0 || 185 strcmp(cp, ".bz2") == 0 || strcmp(cp, ".split") == 0)) 186 return (ENOENT); 187 188 /* Try to open the compressed datafile */ 189 rawfd = open(fname, O_RDONLY | F_GZIP); 190 if (rawfd == -1) { 191 /* add .gz sufix and try again */ 192 zfname = malloc(strlen(fname) + 4); 193 if (zfname == NULL) 194 return (ENOMEM); 195 sprintf(zfname, "%s.gz", fname); 196 rawfd = open(zfname, O_RDONLY); 197 free(zfname); 198 if (rawfd == -1) 199 return (ENOENT); 200 } 201 202 if (fstat(rawfd, &sb) < 0) { 203 printf("zf_open: stat failed\n"); 204 close(rawfd); 205 return (ENOENT); 206 } 207 if (!S_ISREG(sb.st_mode)) { 208 close(rawfd); 209 return (EISDIR); /* best guess */ 210 } 211 212 /* Allocate a z_file structure, populate it */ 213 zf = malloc(sizeof (struct z_file)); 214 if (zf == NULL) 215 return (ENOMEM); 216 bzero(zf, sizeof (struct z_file)); 217 zf->zf_rawfd = rawfd; 218 219 /* Verify that the file is gzipped */ 220 if (check_header(zf)) { 221 close(zf->zf_rawfd); 222 free(zf); 223 return (EFTYPE); 224 } 225 226 /* Initialise the inflation engine */ 227 if ((error = inflateInit2(&(zf->zf_zstream), -15)) != Z_OK) { 228 printf("zf_open: inflateInit returned %d : %s\n", error, 229 zf->zf_zstream.msg); 230 close(zf->zf_rawfd); 231 free(zf); 232 return (EIO); 233 } 234 235 /* Looks OK, we'll take it */ 236 f->f_fsdata = zf; 237 return (0); 238 } 239 240 static int 241 zf_close(struct open_file *f) 242 { 243 struct z_file *zf = (struct z_file *)f->f_fsdata; 244 245 inflateEnd(&(zf->zf_zstream)); 246 close(zf->zf_rawfd); 247 free(zf); 248 return (0); 249 } 250 251 static int 252 zf_read(struct open_file *f, void *buf, size_t size, size_t *resid) 253 { 254 struct z_file *zf = (struct z_file *)f->f_fsdata; 255 int error; 256 257 zf->zf_zstream.next_out = buf; /* where and how much */ 258 zf->zf_zstream.avail_out = size; 259 260 while (zf->zf_zstream.avail_out && zf->zf_endseen == 0) { 261 if ((zf->zf_zstream.avail_in == 0) && (zf_fill(zf) == -1)) { 262 printf("zf_read: fill error\n"); 263 return (EIO); 264 } 265 if (zf->zf_zstream.avail_in == 0) { /* oops, unexpected EOF */ 266 printf("zf_read: unexpected EOF\n"); 267 if (zf->zf_zstream.avail_out == size) 268 return (EIO); 269 break; 270 } 271 272 /* decompression pass */ 273 error = inflate(&zf->zf_zstream, Z_SYNC_FLUSH); 274 if (error == Z_STREAM_END) { /* EOF, all done */ 275 zf->zf_endseen = 1; 276 break; 277 } 278 if (error != Z_OK) { /* argh, decompression error */ 279 printf("inflate: %s\n", zf->zf_zstream.msg); 280 return (EIO); 281 } 282 } 283 if (resid != NULL) 284 *resid = zf->zf_zstream.avail_out; 285 return (0); 286 } 287 288 static int 289 zf_rewind(struct open_file *f) 290 { 291 struct z_file *zf = (struct z_file *)f->f_fsdata; 292 293 if (lseek(zf->zf_rawfd, zf->zf_dataoffset, SEEK_SET) == -1) 294 return (-1); 295 zf->zf_zstream.avail_in = 0; 296 zf->zf_zstream.next_in = NULL; 297 zf->zf_endseen = 0; 298 (void) inflateReset(&zf->zf_zstream); 299 300 return (0); 301 } 302 303 static off_t 304 zf_seek(struct open_file *f, off_t offset, int where) 305 { 306 struct z_file *zf = (struct z_file *)f->f_fsdata; 307 off_t target; 308 char discard[16]; 309 310 switch (where) { 311 case SEEK_SET: 312 target = offset; 313 break; 314 case SEEK_CUR: 315 target = offset + zf->zf_zstream.total_out; 316 break; 317 default: 318 errno = EINVAL; 319 return (-1); 320 } 321 322 /* rewind if required */ 323 if (target < zf->zf_zstream.total_out && zf_rewind(f) != 0) 324 return (-1); 325 326 /* skip forwards if required */ 327 while (target > zf->zf_zstream.total_out) { 328 errno = zf_read(f, discard, min(sizeof (discard), 329 target - zf->zf_zstream.total_out), NULL); 330 if (errno != 0) 331 return (-1); 332 } 333 /* This is where we are (be honest if we overshot) */ 334 return (zf->zf_zstream.total_out); 335 } 336 337 338 static int 339 zf_stat(struct open_file *f, struct stat *sb) 340 { 341 struct z_file *zf = (struct z_file *)f->f_fsdata; 342 int result; 343 off_t pos1, pos2; 344 uint32_t size; 345 346 /* stat as normal, but indicate that size is unknown */ 347 if ((result = fstat(zf->zf_rawfd, sb)) == 0) { 348 if (sb->st_size == -1) 349 return (result); 350 pos1 = lseek(zf->zf_rawfd, 0, SEEK_CUR); 351 pos2 = lseek(zf->zf_rawfd, sb->st_size - 4, SEEK_SET); 352 if (pos2 != -1) { 353 if (read(zf->zf_rawfd, &size, 4) == 4) 354 sb->st_size = (off_t)size; 355 else 356 sb->st_size = -1; 357 } else 358 sb->st_size = -1; 359 360 pos1 = lseek(zf->zf_rawfd, pos1, SEEK_SET); 361 } 362 return (result); 363 } 364