Jeff Sharkey | 3e8b158 | 2012-07-13 16:37:13 -0700 | [diff] [blame] | 1 | /* $NetBSD: file.c,v 1.7 2011/04/18 22:46:48 joerg Exp $ */ |
| 2 | /* $FreeBSD: head/usr.bin/grep/file.c 211496 2010-08-19 09:28:59Z des $ */ |
| 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 | #if HAVE_NBTOOL_CONFIG_H |
| 34 | #include "nbtool_config.h" |
| 35 | #endif |
| 36 | |
| 37 | #include <sys/cdefs.h> |
| 38 | __RCSID("$NetBSD: file.c,v 1.7 2011/04/18 22:46:48 joerg Exp $"); |
| 39 | |
| 40 | #include <sys/param.h> |
| 41 | #include <sys/types.h> |
| 42 | #include <sys/stat.h> |
| 43 | |
| 44 | #ifndef ANDROID |
| 45 | #include <bzlib.h> |
| 46 | #endif |
| 47 | #include <err.h> |
| 48 | #include <errno.h> |
| 49 | #include <fcntl.h> |
| 50 | #include <stddef.h> |
| 51 | #include <stdlib.h> |
| 52 | #include <string.h> |
| 53 | #include <unistd.h> |
| 54 | #include <wchar.h> |
| 55 | #include <wctype.h> |
| 56 | #ifndef ANDROID |
| 57 | #include <zlib.h> |
| 58 | #endif |
| 59 | |
| 60 | #include "grep.h" |
| 61 | |
| 62 | #define MAXBUFSIZ (32 * 1024) |
| 63 | #define LNBUFBUMP 80 |
| 64 | |
| 65 | #ifndef ANDROID |
| 66 | static gzFile gzbufdesc; |
| 67 | static BZFILE* bzbufdesc; |
| 68 | #endif |
| 69 | |
| 70 | static unsigned char buffer[MAXBUFSIZ]; |
| 71 | static unsigned char *bufpos; |
| 72 | static size_t bufrem; |
| 73 | |
| 74 | static unsigned char *lnbuf; |
| 75 | static size_t lnbuflen; |
| 76 | |
| 77 | static inline int |
| 78 | grep_refill(struct file *f) |
| 79 | { |
| 80 | ssize_t nr; |
| 81 | int bzerr; |
| 82 | |
| 83 | bufpos = buffer; |
| 84 | bufrem = 0; |
| 85 | |
| 86 | #ifndef ANDROID |
| 87 | if (filebehave == FILE_GZIP) |
| 88 | nr = gzread(gzbufdesc, buffer, MAXBUFSIZ); |
| 89 | else if (filebehave == FILE_BZIP && bzbufdesc != NULL) { |
| 90 | nr = BZ2_bzRead(&bzerr, bzbufdesc, buffer, MAXBUFSIZ); |
| 91 | switch (bzerr) { |
| 92 | case BZ_OK: |
| 93 | case BZ_STREAM_END: |
| 94 | /* No problem, nr will be okay */ |
| 95 | break; |
| 96 | case BZ_DATA_ERROR_MAGIC: |
| 97 | /* |
| 98 | * As opposed to gzread(), which simply returns the |
| 99 | * plain file data, if it is not in the correct |
| 100 | * compressed format, BZ2_bzRead() instead aborts. |
| 101 | * |
| 102 | * So, just restart at the beginning of the file again, |
| 103 | * and use plain reads from now on. |
| 104 | */ |
| 105 | BZ2_bzReadClose(&bzerr, bzbufdesc); |
| 106 | bzbufdesc = NULL; |
| 107 | if (lseek(f->fd, 0, SEEK_SET) == -1) |
| 108 | return (-1); |
| 109 | nr = read(f->fd, buffer, MAXBUFSIZ); |
| 110 | break; |
| 111 | default: |
| 112 | /* Make sure we exit with an error */ |
| 113 | nr = -1; |
| 114 | } |
| 115 | } else |
| 116 | #endif |
| 117 | nr = read(f->fd, buffer, MAXBUFSIZ); |
| 118 | |
| 119 | if (nr < 0) |
| 120 | return (-1); |
| 121 | |
| 122 | bufrem = nr; |
| 123 | return (0); |
| 124 | } |
| 125 | |
| 126 | static inline int |
| 127 | grep_lnbufgrow(size_t newlen) |
| 128 | { |
| 129 | |
| 130 | if (lnbuflen < newlen) { |
| 131 | lnbuf = grep_realloc(lnbuf, newlen); |
| 132 | lnbuflen = newlen; |
| 133 | } |
| 134 | |
| 135 | return (0); |
| 136 | } |
| 137 | |
| 138 | char * |
| 139 | grep_fgetln(struct file *f, size_t *lenp) |
| 140 | { |
| 141 | unsigned char *p; |
| 142 | char *ret; |
| 143 | size_t len; |
| 144 | size_t off; |
| 145 | ptrdiff_t diff; |
| 146 | |
| 147 | /* Fill the buffer, if necessary */ |
| 148 | if (bufrem == 0 && grep_refill(f) != 0) |
| 149 | goto error; |
| 150 | |
| 151 | if (bufrem == 0) { |
| 152 | /* Return zero length to indicate EOF */ |
| 153 | *lenp = 0; |
| 154 | return ((char *)bufpos); |
| 155 | } |
| 156 | |
| 157 | /* Look for a newline in the remaining part of the buffer */ |
| 158 | if ((p = memchr(bufpos, line_sep, bufrem)) != NULL) { |
| 159 | ++p; /* advance over newline */ |
| 160 | ret = (char *)bufpos; |
| 161 | len = p - bufpos; |
| 162 | bufrem -= len; |
| 163 | bufpos = p; |
| 164 | *lenp = len; |
| 165 | return (ret); |
| 166 | } |
| 167 | |
| 168 | /* We have to copy the current buffered data to the line buffer */ |
| 169 | for (len = bufrem, off = 0; ; len += bufrem) { |
| 170 | /* Make sure there is room for more data */ |
| 171 | if (grep_lnbufgrow(len + LNBUFBUMP)) |
| 172 | goto error; |
| 173 | memcpy(lnbuf + off, bufpos, len - off); |
| 174 | off = len; |
| 175 | if (grep_refill(f) != 0) |
| 176 | goto error; |
| 177 | if (bufrem == 0) |
| 178 | /* EOF: return partial line */ |
| 179 | break; |
| 180 | if ((p = memchr(bufpos, line_sep, bufrem)) == NULL) |
| 181 | continue; |
| 182 | /* got it: finish up the line (like code above) */ |
| 183 | ++p; |
| 184 | diff = p - bufpos; |
| 185 | len += diff; |
| 186 | if (grep_lnbufgrow(len)) |
| 187 | goto error; |
| 188 | memcpy(lnbuf + off, bufpos, diff); |
| 189 | bufrem -= diff; |
| 190 | bufpos = p; |
| 191 | break; |
| 192 | } |
| 193 | *lenp = len; |
| 194 | return ((char *)lnbuf); |
| 195 | |
| 196 | error: |
| 197 | *lenp = 0; |
| 198 | return (NULL); |
| 199 | } |
| 200 | |
| 201 | static inline struct file * |
| 202 | grep_file_init(struct file *f) |
| 203 | { |
| 204 | |
| 205 | #ifndef ANDROID |
| 206 | if (filebehave == FILE_GZIP && |
| 207 | (gzbufdesc = gzdopen(f->fd, "r")) == NULL) |
| 208 | goto error; |
| 209 | |
| 210 | if (filebehave == FILE_BZIP && |
| 211 | (bzbufdesc = BZ2_bzdopen(f->fd, "r")) == NULL) |
| 212 | goto error; |
| 213 | #endif |
| 214 | |
| 215 | /* Fill read buffer, also catches errors early */ |
| 216 | if (grep_refill(f) != 0) |
| 217 | goto error; |
| 218 | |
| 219 | /* Check for binary stuff, if necessary */ |
| 220 | if (!nulldataflag && binbehave != BINFILE_TEXT && |
| 221 | memchr(bufpos, '\0', bufrem) != NULL) |
| 222 | f->binary = true; |
| 223 | |
| 224 | return (f); |
| 225 | error: |
| 226 | close(f->fd); |
| 227 | free(f); |
| 228 | return (NULL); |
| 229 | } |
| 230 | |
| 231 | /* |
| 232 | * Opens a file for processing. |
| 233 | */ |
| 234 | struct file * |
| 235 | grep_open(const char *path) |
| 236 | { |
| 237 | struct file *f; |
| 238 | |
| 239 | f = grep_malloc(sizeof *f); |
| 240 | memset(f, 0, sizeof *f); |
| 241 | if (path == NULL) { |
| 242 | /* Processing stdin implies --line-buffered. */ |
| 243 | lbflag = true; |
| 244 | f->fd = STDIN_FILENO; |
| 245 | } else if ((f->fd = open(path, O_RDONLY)) == -1) { |
| 246 | free(f); |
| 247 | return (NULL); |
| 248 | } |
| 249 | |
| 250 | return (grep_file_init(f)); |
| 251 | } |
| 252 | |
| 253 | /* |
| 254 | * Closes a file. |
| 255 | */ |
| 256 | void |
| 257 | grep_close(struct file *f) |
| 258 | { |
| 259 | |
| 260 | close(f->fd); |
| 261 | |
| 262 | /* Reset read buffer and line buffer */ |
| 263 | bufpos = buffer; |
| 264 | bufrem = 0; |
| 265 | |
| 266 | free(lnbuf); |
| 267 | lnbuf = NULL; |
| 268 | lnbuflen = 0; |
| 269 | } |