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