bin/grep/file.c

   1 /*      $NetBSD: file.c,v 1.5 2011/02/16 18:35:39 joerg Exp $   */
   2 /*      $FreeBSD$       */
   3 /*      $OpenBSD: file.c,v 1.11 2010/07/02 20:48:48 nicm Exp $  */
   4
   5 /*-
   6  * Copyright (c) 1999 James Howard and Dag-Erling Coïdan Smørgrav
   7  * Copyright (C) 2008-2010 Gabor Kovesdan <gabor@FreeBSD.org>
   8  * Copyright (C) 2010 Dimitry Andric <dimitry@andric.com>
   9  * All rights reserved.
  10  *
  11  * Redistribution and use in source and binary forms, with or without
  12  * modification, are permitted provided that the following conditions
  13  * are met:
  14  * 1. Redistributions of source code must retain the above copyright
  15  *    notice, this list of conditions and the following disclaimer.
  16  * 2. Redistributions in binary form must reproduce the above copyright
  17  *    notice, this list of conditions and the following disclaimer in the
  18  *    documentation and/or other materials provided with the distribution.
  19  *
  20  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
  21  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  22  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  23  * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
  24  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  25  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  26  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  27  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  28  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  29  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  30  * SUCH DAMAGE.
  31  */
  32
  33 #include <sys/param.h>
  34 #include <sys/mman.h>
  35 #include <sys/stat.h>
  36 #include <sys/types.h>
  37
  38 #include <err.h>
  39 #include <errno.h>
  40 #include <fcntl.h>
  41 #include <stddef.h>
  42 #include <stdlib.h>
  43 #include <string.h>
  44 #include <unistd.h>
  45 #include <wchar.h>
  46 #include <wctype.h>
  47 #include <zlib.h>
  48
  49 #ifndef WITHOUT_LZMA
  50 #include <lzma.h>
  51 #endif
  52
  53 #ifndef WITHOUT_BZIP2
  54 #include <bzlib.h>
  55 #endif
  56
  57 #include "grep.h"
  58 #include "freebsd-compat.h"
  59
  60 #define MAXBUFSIZ       (32 * 1024)
  61 #define LNBUFBUMP       80
  62
  63 static gzFile gzbufdesc;
  64 #ifndef WITHOUT_LZMA
  65 static lzma_stream lstrm = LZMA_STREAM_INIT;
  66 static lzma_action laction;
  67 static uint8_t lin_buf[MAXBUFSIZ];
  68 #endif
  69 #ifndef WITHOUT_BZIP2
  70 static BZFILE* bzbufdesc;
  71 #endif
  72
  73 static unsigned char *buffer;
  74 static unsigned char *bufpos;
  75 static size_t bufrem;
  76 static size_t fsiz;
  77
  78 static unsigned char *lnbuf;
  79 static size_t lnbuflen;
  80
  81 static inline int
  82 grep_refill(struct file *f)
  83 {
  84         ssize_t nr;
  85
  86         if (filebehave == FILE_MMAP)
  87                 return (0);
  88
  89         bufpos = buffer;
  90         bufrem = 0;
  91
  92         if (filebehave == FILE_GZIP) {
  93                 nr = gzread(gzbufdesc, buffer, MAXBUFSIZ);
  94 #ifndef WITHOUT_BZIP2
  95         } else if (filebehave == FILE_BZIP && bzbufdesc != NULL) {
  96                 int bzerr;
  97
  98                 nr = BZ2_bzRead(&bzerr, bzbufdesc, buffer, MAXBUFSIZ);
  99                 switch (bzerr) {
 100                 case BZ_OK:
 101                 case BZ_STREAM_END:
 102                         /* No problem, nr will be okay */
 103                         break;
 104                 case BZ_DATA_ERROR_MAGIC:
 105                         /*
 106                          * As opposed to gzread(), which simply returns the
 107                          * plain file data, if it is not in the correct
 108                          * compressed format, BZ2_bzRead() instead aborts.
 109                          *
 110                          * So, just restart at the beginning of the file again,
 111                          * and use plain reads from now on.
 112                          */
 113                         BZ2_bzReadClose(&bzerr, bzbufdesc);
 114                         bzbufdesc = NULL;
 115                         if (lseek(f->fd, 0, SEEK_SET) == -1)
 116                                 return (-1);
 117                         nr = read(f->fd, buffer, MAXBUFSIZ);
 118                         break;
 119                 default:
 120                         /* Make sure we exit with an error */
 121                         nr = -1;
 122                 }
 123 #endif
 124 #ifndef WITHOUT_LZMA
 125         } else if ((filebehave == FILE_XZ) || (filebehave == FILE_LZMA)) {
 126                 lzma_ret ret;
 127                 lstrm.next_out = buffer;
 128
 129                 do {
 130                         if (lstrm.avail_in == 0) {
 131                                 lstrm.next_in = lin_buf;
 132                                 nr = read(f->fd, lin_buf, MAXBUFSIZ);
 133
 134                                 if (nr < 0)
 135                                         return (-1);
 136                                 else if (nr == 0)
 137                                         laction = LZMA_FINISH;
 138
 139                                 lstrm.avail_in = nr;
 140                         }
 141
 142                         ret = lzma_code(&lstrm, laction);
 143
 144                         if (ret != LZMA_OK && ret != LZMA_STREAM_END)
 145                                 return (-1);
 146
 147                         if (lstrm.avail_out == 0 || ret == LZMA_STREAM_END) {
 148                                 bufrem = MAXBUFSIZ - lstrm.avail_out;
 149                                 lstrm.next_out = buffer;
 150                                 lstrm.avail_out = MAXBUFSIZ;
 151                         }
 152                 } while (bufrem == 0 && ret != LZMA_STREAM_END);
 153
 154                 return (0);
 155 #endif  /* WIHTOUT_LZMA */
 156         } else
 157                 nr = read(f->fd, buffer, MAXBUFSIZ);
 158
 159         if (nr < 0)
 160                 return (-1);
 161
 162         bufrem = nr;
 163         return (0);
 164 }
 165
 166 static inline int
 167 grep_lnbufgrow(size_t newlen)
 168 {
 169
 170         if (lnbuflen < newlen) {
 171                 lnbuf = grep_realloc(lnbuf, newlen);
 172                 lnbuflen = newlen;
 173         }
 174
 175         return (0);
 176 }
 177
 178 char *
 179 grep_fgetln(struct file *f, size_t *lenp)
 180 {
 181         unsigned char *p;
 182         char *ret;
 183         size_t len;
 184         size_t off;
 185         ptrdiff_t diff;
 186
 187         /* Fill the buffer, if necessary */
 188         if (bufrem == 0 && grep_refill(f) != 0)
 189                 goto error;
 190
 191         if (bufrem == 0) {
 192                 /* Return zero length to indicate EOF */
 193                 *lenp = 0;
 194                 return (bufpos);
 195         }
 196
 197         /* Look for a newline in the remaining part of the buffer */
 198         if ((p = memchr(bufpos, fileeol, bufrem)) != NULL) {
 199                 ++p; /* advance over newline */
 200                 ret = bufpos;
 201                 len = p - bufpos;
 202                 bufrem -= len;
 203                 bufpos = p;
 204                 *lenp = len;
 205                 return (ret);
 206         }
 207
 208         /* We have to copy the current buffered data to the line buffer */
 209         for (len = bufrem, off = 0; ; len += bufrem) {
 210                 /* Make sure there is room for more data */
 211                 if (grep_lnbufgrow(len + LNBUFBUMP))
 212                         goto error;
 213                 memcpy(lnbuf + off, bufpos, len - off);
 214                 /* With FILE_MMAP, this is EOF; there's no more to refill */
 215                 if (filebehave == FILE_MMAP) {
 216                         bufrem -= len;
 217                         break;
 218                 }
 219                 off = len;
 220                 /* Fetch more to try and find EOL/EOF */
 221                 if (grep_refill(f) != 0)
 222                         goto error;
 223                 if (bufrem == 0)
 224                         /* EOF: return partial line */
 225                         break;
 226                 if ((p = memchr(bufpos, fileeol, bufrem)) == NULL)
 227                         continue;
 228                 /* got it: finish up the line (like code above) */
 229                 ++p;
 230                 diff = p - bufpos;
 231                 len += diff;
 232                 if (grep_lnbufgrow(len))
 233                     goto error;
 234                 memcpy(lnbuf + off, bufpos, diff);
 235                 bufrem -= diff;
 236                 bufpos = p;
 237                 break;
 238         }
 239         *lenp = len;
 240         return (lnbuf);
 241
 242 error:
 243         *lenp = 0;
 244         return (NULL);
 245 }
 246
 247 /*
 248  * Opens a file for processing.
 249  */
 250 struct file *
 251 grep_open(const char *path)
 252 {
 253         struct file *f;
 254
 255         f = grep_malloc(sizeof *f);
 256         memset(f, 0, sizeof *f);
 257         if (path == NULL) {
 258                 /* Processing stdin implies --line-buffered. */
 259                 lbflag = true;
 260                 f->fd = STDIN_FILENO;
 261         } else if ((f->fd = open(path, O_RDONLY)) == -1)
 262                 goto error1;
 263
 264         if (filebehave == FILE_MMAP) {
 265                 struct stat st;
 266
 267                 if ((fstat(f->fd, &st) == -1) || (st.st_size > OFF_MAX) ||
 268                     (!S_ISREG(st.st_mode)))
 269                         filebehave = FILE_STDIO;
 270                 else {
 271                         int flags = MAP_PRIVATE | MAP_NOCORE | MAP_NOSYNC;
 272 #ifdef MAP_PREFAULT_READ
 273                         flags |= MAP_PREFAULT_READ;
 274 #endif
 275                         fsiz = st.st_size;
 276                         buffer = mmap(NULL, fsiz, PROT_READ, flags,
 277                              f->fd, (off_t)0);
 278                         if (buffer == MAP_FAILED)
 279                                 filebehave = FILE_STDIO;
 280                         else {
 281                                 bufrem = st.st_size;
 282                                 bufpos = buffer;
 283                                 madvise(buffer, st.st_size, MADV_SEQUENTIAL);
 284                         }
 285                 }
 286         }
 287
 288         if ((buffer == NULL) || (buffer == MAP_FAILED))
 289                 buffer = grep_malloc(MAXBUFSIZ);
 290
 291         if (filebehave == FILE_GZIP &&
 292             (gzbufdesc = gzdopen(f->fd, "r")) == NULL)
 293                 goto error2;
 294
 295 #ifndef WITHOUT_BZIP2
 296         if (filebehave == FILE_BZIP &&
 297             (bzbufdesc = BZ2_bzdopen(f->fd, "r")) == NULL)
 298                 goto error2;
 299 #endif
 300 #ifndef WITHOUT_LZMA
 301         else if ((filebehave == FILE_XZ) || (filebehave == FILE_LZMA)) {
 302                 lzma_ret ret;
 303
 304                 ret = (filebehave == FILE_XZ) ?
 305                         lzma_stream_decoder(&lstrm, UINT64_MAX,
 306                                         LZMA_CONCATENATED) :
 307                         lzma_alone_decoder(&lstrm, UINT64_MAX);
 308
 309                 if (ret != LZMA_OK)
 310                         goto error2;
 311
 312                 lstrm.avail_in = 0;
 313                 lstrm.avail_out = MAXBUFSIZ;
 314                 laction = LZMA_RUN;
 315         }
 316 #endif
 317
 318         /* Fill read buffer, also catches errors early */
 319         if (bufrem == 0 && grep_refill(f) != 0)
 320                 goto error2;
 321
 322         /* Check for binary stuff, if necessary */
 323         if (binbehave != BINFILE_TEXT && fileeol != '\0' &&
 324             memchr(bufpos, '\0', bufrem) != NULL)
 325         f->binary = true;
 326
 327         return (f);
 328
 329 error2:
 330         close(f->fd);
 331 error1:
 332         free(f);
 333         return (NULL);
 334 }
 335
 336 /*
 337  * Closes a file.
 338  */
 339 void
 340 grep_close(struct file *f)
 341 {
 342
 343         close(f->fd);
 344
 345         /* Reset read buffer and line buffer */
 346         if (filebehave == FILE_MMAP) {
 347                 munmap(buffer, fsiz);
 348                 buffer = NULL;
 349         }
 350         bufpos = buffer;
 351         bufrem = 0;
 352
 353         free(lnbuf);
 354         lnbuf = NULL;
 355         lnbuflen = 0;
 356 }