Annotation of src/usr.bin/compress/gzopen.c, Revision 1.20
1.20 ! henning 1: /* $OpenBSD: gzopen.c,v 1.19 2003/12/16 22:46:25 henning Exp $ */
1.1 mickey 2:
3: /*
4: * Copyright (c) 1997 Michael Shalayeff
5: * All rights reserved.
6: *
7: * Redistribution and use in source and binary forms, with or without
8: * modification, are permitted provided that the following conditions
9: * are met:
10: * 1. Redistributions of source code must retain the above copyright
11: * notice, this list of conditions and the following disclaimer.
12: * 2. Redistributions in binary form must reproduce the above copyright
13: * notice, this list of conditions and the following disclaimer in the
14: * documentation and/or other materials provided with the distribution.
15: *
1.5 mickey 16: * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
17: * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
1.1 mickey 18: * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
19: * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
20: * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
21: * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
22: * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
23: * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
24: * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
25: * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
26: * SUCH DAMAGE.
27: *
28: */
29: /* this is partially derived from the zlib's gzio.c file, so the notice: */
30: /*
31: zlib.h -- interface of the 'zlib' general purpose compression library
32: version 1.0.4, Jul 24th, 1996.
33:
34: Copyright (C) 1995-1996 Jean-loup Gailly and Mark Adler
35:
36: This software is provided 'as-is', without any express or implied
37: warranty. In no event will the authors be held liable for any damages
38: arising from the use of this software.
39:
40: Permission is granted to anyone to use this software for any purpose,
41: including commercial applications, and to alter it and redistribute it
42: freely, subject to the following restrictions:
43:
44: 1. The origin of this software must not be misrepresented; you must not
45: claim that you wrote the original software. If you use this software
46: in a product, an acknowledgment in the product documentation would be
47: appreciated but is not required.
48: 2. Altered source versions must be plainly marked as such, and must not be
49: misrepresented as being the original software.
50: 3. This notice may not be removed or altered from any source distribution.
51:
52: Jean-loup Gailly Mark Adler
53: gzip@prep.ai.mit.edu madler@alumni.caltech.edu
54:
55:
56: The data format used by the zlib library is described by RFCs (Request for
57: Comments) 1950 to 1952 in the files ftp://ds.internic.net/rfc/rfc1950.txt
58: (zlib format), rfc1951.txt (deflate format) and rfc1952.txt (gzip format).
59: */
60:
1.19 henning 61: #ifndef SMALL
1.5 mickey 62: const char gz_rcsid[] =
1.20 ! henning 63: "$OpenBSD: gzopen.c,v 1.19 2003/12/16 22:46:25 henning Exp $";
1.19 henning 64: #endif
1.5 mickey 65:
1.13 millert 66: #include <sys/param.h>
1.1 mickey 67: #include <sys/stat.h>
1.13 millert 68: #include <sys/uio.h>
1.1 mickey 69: #include <stdio.h>
70: #include <stdlib.h>
1.11 david 71: #include <string.h>
1.1 mickey 72: #include <errno.h>
73: #include <unistd.h>
74: #include <zlib.h>
75: #include "compress.h"
76:
77: /* gzip flag byte */
78: #define ASCII_FLAG 0x01 /* bit 0 set: file probably ascii text */
79: #define HEAD_CRC 0x02 /* bit 1 set: header CRC present */
80: #define EXTRA_FIELD 0x04 /* bit 2 set: extra field present */
81: #define ORIG_NAME 0x08 /* bit 3 set: original file name present */
82: #define COMMENT 0x10 /* bit 4 set: file comment present */
83: #define RESERVED 0xE0 /* bits 5..7: reserved */
84:
85: #define DEF_MEM_LEVEL 8
86: #define OS_CODE 0x03 /* unix */
87:
88: typedef
89: struct gz_stream {
90: int z_fd; /* .gz file */
91: z_stream z_stream; /* libz stream */
92: int z_eof; /* set if end of input file */
93: u_char z_buf[Z_BUFSIZE]; /* i/o buffer */
1.13 millert 94: u_int32_t z_time; /* timestamp (mtime) */
95: u_int32_t z_hlen; /* length of the gz header */
1.1 mickey 96: u_int32_t z_crc; /* crc32 of uncompressed data */
97: char z_mode; /* 'w' or 'r' */
98:
99: } gz_stream;
100:
1.5 mickey 101: static const u_char gz_magic[2] = {0x1f, 0x8b}; /* gzip magic header */
1.1 mickey 102:
1.3 millert 103: static int put_int32(gz_stream *, u_int32_t);
104: static u_int32_t get_int32(gz_stream *);
1.13 millert 105: static int get_header(gz_stream *, char *, int);
106: static int put_header(gz_stream *, char *, u_int32_t);
1.3 millert 107: static int get_byte(gz_stream *);
1.1 mickey 108:
109: void *
1.13 millert 110: gz_open(int fd, const char *mode, char *name, int bits,
111: u_int32_t mtime, int gotmagic)
1.1 mickey 112: {
113: gz_stream *s;
114:
115: if (fd < 0 || !mode)
116: return NULL;
117:
118: if ((mode[0] != 'r' && mode[0] != 'w') || mode[1] != '\0' ||
119: bits < 0 || bits > Z_BEST_COMPRESSION) {
120: errno = EINVAL;
121: return NULL;
122: }
1.4 aaron 123: if ((s = (gz_stream *)calloc(1, sizeof(gz_stream))) == NULL)
1.1 mickey 124: return NULL;
125:
126: s->z_stream.zalloc = (alloc_func)0;
127: s->z_stream.zfree = (free_func)0;
128: s->z_stream.opaque = (voidpf)0;
129: s->z_stream.next_in = Z_NULL;
130: s->z_stream.next_out = Z_NULL;
131: s->z_stream.avail_in = s->z_stream.avail_out = 0;
132: s->z_fd = 0;
133: s->z_eof = 0;
1.13 millert 134: s->z_time = 0;
135: s->z_hlen = 0;
1.1 mickey 136: s->z_crc = crc32(0L, Z_NULL, 0);
137: s->z_mode = mode[0];
138:
139: if (s->z_mode == 'w') {
1.20 ! henning 140: #ifndef SMALL
1.1 mickey 141: /* windowBits is passed < 0 to suppress zlib header */
142: if (deflateInit2(&(s->z_stream), bits, Z_DEFLATED,
143: -MAX_WBITS, DEF_MEM_LEVEL, 0) != Z_OK) {
144: free (s);
145: return NULL;
146: }
147: s->z_stream.next_out = s->z_buf;
1.20 ! henning 148: #else
! 149: return (NULL);
! 150: #endif
1.1 mickey 151: } else {
152: if (inflateInit2(&(s->z_stream), -MAX_WBITS) != Z_OK) {
153: free (s);
154: return NULL;
155: }
156: s->z_stream.next_in = s->z_buf;
157: }
158: s->z_stream.avail_out = Z_BUFSIZE;
159:
160: errno = 0;
161: s->z_fd = fd;
162:
163: if (s->z_mode == 'w') {
1.13 millert 164: /* write the .gz header */
165: if (put_header(s, name, mtime) != 0) {
166: gz_close(s, NULL);
1.1 mickey 167: s = NULL;
168: }
169: } else {
1.13 millert 170: /* read the .gz header */
171: if (get_header(s, name, gotmagic) != 0) {
172: gz_close(s, NULL);
1.1 mickey 173: s = NULL;
174: }
175: }
176:
177: return s;
178: }
179:
180: int
1.13 millert 181: gz_close(void *cookie, struct z_info *info)
1.1 mickey 182: {
1.2 mpech 183: gz_stream *s = (gz_stream*)cookie;
1.1 mickey 184: int err = 0;
185:
186: if (s == NULL)
187: return -1;
188:
1.20 ! henning 189: #ifndef SMALL
1.1 mickey 190: if (s->z_mode == 'w' && (err = gz_flush (s, Z_FINISH)) == Z_OK) {
1.13 millert 191: if ((err = put_int32 (s, s->z_crc)) == Z_OK) {
192: s->z_hlen += sizeof(int32_t);
193: if ((err = put_int32 (s, s->z_stream.total_in)) == Z_OK)
194: s->z_hlen += sizeof(int32_t);
195: }
1.1 mickey 196: }
1.20 ! henning 197: #endif
1.1 mickey 198: if (!err && s->z_stream.state != NULL) {
199: if (s->z_mode == 'w')
1.20 ! henning 200: #ifndef SMALL
1.1 mickey 201: err = deflateEnd(&s->z_stream);
1.20 ! henning 202: #else
! 203: err = -1;
! 204: #endif;
1.1 mickey 205: else if (s->z_mode == 'r')
206: err = inflateEnd(&s->z_stream);
207: }
1.10 mickey 208:
1.13 millert 209: if (info != NULL) {
210: info->mtime = s->z_time;
211: info->crc = s->z_crc;
212: info->hlen = s->z_hlen;
213: info->total_in = (off_t)s->z_stream.total_in;
214: info->total_out = (off_t)s->z_stream.total_out;
215: }
216:
1.10 mickey 217: if (!err)
218: err = close(s->z_fd);
219: else
220: (void)close(s->z_fd);
1.1 mickey 221:
222: free(s);
223:
224: return err;
225: }
226:
1.20 ! henning 227: #ifndef SMALL
1.1 mickey 228: int
1.7 deraadt 229: gz_flush(void *cookie, int flush)
1.1 mickey 230: {
1.2 mpech 231: gz_stream *s = (gz_stream*)cookie;
1.1 mickey 232: size_t len;
233: int done = 0;
234: int err;
235:
236: if (s == NULL || s->z_mode != 'w') {
237: errno = EBADF;
238: return Z_ERRNO;
239: }
240:
241: s->z_stream.avail_in = 0; /* should be zero already anyway */
242:
243: for (;;) {
244: len = Z_BUFSIZE - s->z_stream.avail_out;
245:
246: if (len != 0) {
247: if (write(s->z_fd, s->z_buf, len) != len)
248: return Z_ERRNO;
249: s->z_stream.next_out = s->z_buf;
250: s->z_stream.avail_out = Z_BUFSIZE;
251: }
252: if (done)
253: break;
254: if ((err = deflate(&(s->z_stream), flush)) != Z_OK &&
255: err != Z_STREAM_END)
256: return err;
257:
258: /* deflate has finished flushing only when it hasn't
259: * used up all the available space in the output buffer
260: */
261: done = (s->z_stream.avail_out != 0 || err == Z_STREAM_END);
262: }
263: return 0;
264: }
1.20 ! henning 265: #endif
1.1 mickey 266:
267: static int
1.7 deraadt 268: put_int32(gz_stream *s, u_int32_t x)
1.1 mickey 269: {
1.8 millert 270: u_int32_t y = htole32(x);
271:
272: if (write(s->z_fd, &y, sizeof(y)) != sizeof(y))
1.1 mickey 273: return Z_ERRNO;
274: return 0;
275: }
276:
277: static int
1.7 deraadt 278: get_byte(gz_stream *s)
1.1 mickey 279: {
280: if (s->z_eof)
281: return EOF;
1.5 mickey 282:
1.1 mickey 283: if (s->z_stream.avail_in == 0) {
284: errno = 0;
285: s->z_stream.avail_in = read(s->z_fd, s->z_buf, Z_BUFSIZE);
286: if (s->z_stream.avail_in <= 0) {
287: s->z_eof = 1;
288: return EOF;
289: }
290: s->z_stream.next_in = s->z_buf;
291: }
292: s->z_stream.avail_in--;
293: return *s->z_stream.next_in++;
294: }
295:
1.5 mickey 296: static u_int32_t
1.7 deraadt 297: get_int32(gz_stream *s)
1.1 mickey 298: {
1.2 mpech 299: u_int32_t x;
1.1 mickey 300:
301: x = ((u_int32_t)(get_byte(s) & 0xff));
302: x |= ((u_int32_t)(get_byte(s) & 0xff))<<8;
303: x |= ((u_int32_t)(get_byte(s) & 0xff))<<16;
304: x |= ((u_int32_t)(get_byte(s) & 0xff))<<24;
305: return x;
306: }
307:
308: static int
1.13 millert 309: get_header(gz_stream *s, char *name, int gotmagic)
1.1 mickey 310: {
311: int method; /* method byte */
312: int flags; /* flags byte */
1.13 millert 313: char *ep;
1.1 mickey 314: uInt len;
315: int c;
316:
317: /* Check the gzip magic header */
1.12 millert 318: if (!gotmagic) {
319: for (len = 0; len < 2; len++) {
320: c = get_byte(s);
321: if (c != gz_magic[len]) {
322: errno = EFTYPE;
323: return -1;
324: }
1.1 mickey 325: }
326: }
327:
328: method = get_byte(s);
329: flags = get_byte(s);
330: if (method != Z_DEFLATED || (flags & RESERVED) != 0) {
331: errno = EFTYPE;
332: return -1;
333: }
334:
1.13 millert 335: /* Stash timestamp (mtime) */
336: s->z_time = get_int32(s);
337:
338: /* Discard xflags and OS code */
339: (void)get_byte(s);
340: (void)get_byte(s);
1.1 mickey 341:
1.13 millert 342: s->z_hlen = 10; /* magic, method, flags, time, xflags, OS code */
1.1 mickey 343: if ((flags & EXTRA_FIELD) != 0) { /* skip the extra field */
344: len = (uInt)get_byte(s);
345: len += ((uInt)get_byte(s))<<8;
1.13 millert 346: s->z_hlen += 2;
1.1 mickey 347: /* len is garbage if EOF but the loop below will quit anyway */
348: while (len-- != 0 && get_byte(s) != EOF)
1.13 millert 349: s->z_hlen++;
1.1 mickey 350: }
351:
1.13 millert 352: if ((flags & ORIG_NAME) != 0) { /* read/save the original file name */
353: if ((ep = name) != NULL)
354: ep += MAXPATHLEN - 1;
355: while ((c = get_byte(s)) != EOF) {
356: s->z_hlen++;
357: if (c == '\0')
358: break;
359: if (name < ep)
360: *name++ = c;
361: }
362: if (name != NULL)
363: *name = '\0';
1.1 mickey 364: }
365:
366: if ((flags & COMMENT) != 0) { /* skip the .gz file comment */
1.13 millert 367: while ((c = get_byte(s)) != EOF) {
368: s->z_hlen++;
369: if (c == '\0')
370: break;
371: }
1.1 mickey 372: }
373:
374: if ((flags & HEAD_CRC) != 0) { /* skip the header crc */
1.14 mickey 375: (void)get_byte(s);
376: (void)get_byte(s);
1.13 millert 377: s->z_hlen += 2;
1.1 mickey 378: }
379:
380: if (s->z_eof) {
381: errno = EFTYPE;
382: return -1;
383: }
384:
385: return 0;
386: }
387:
1.13 millert 388: static int
389: put_header(gz_stream *s, char *name, u_int32_t mtime)
390: {
391: struct iovec iov[2];
392: u_char buf[10];
393:
394: buf[0] = gz_magic[0];
395: buf[1] = gz_magic[1];
396: buf[2] = Z_DEFLATED;
397: buf[3] = name ? ORIG_NAME : 0;
398: buf[4] = mtime & 0xff;
399: buf[5] = (mtime >> 8) & 0xff;
400: buf[6] = (mtime >> 16) & 0xff;
401: buf[7] = (mtime >> 24) & 0xff;
402: buf[8] = 0 /* xflags */;
403: buf[9] = OS_CODE;
404: iov[0].iov_base = buf;
405: iov[0].iov_len = sizeof(buf);
406: s->z_hlen = sizeof(buf);
407:
408: if (name != NULL) {
409: iov[1].iov_base = name;
410: iov[1].iov_len = strlen(name) + 1;
411: s->z_hlen += iov[1].iov_len;
412: }
413: if (writev(s->z_fd, iov, name ? 2 : 1) == -1)
414: return (-1);
415: return (0);
416: }
417:
1.1 mickey 418: int
1.7 deraadt 419: gz_read(void *cookie, char *buf, int len)
1.1 mickey 420: {
1.2 mpech 421: gz_stream *s = (gz_stream*)cookie;
1.1 mickey 422: u_char *start = buf; /* starting point for crc computation */
1.17 millert 423: int error = Z_OK;
1.1 mickey 424:
425: s->z_stream.next_out = buf;
426: s->z_stream.avail_out = len;
427:
1.17 millert 428: while (error == Z_OK && !s->z_eof && s->z_stream.avail_out != 0) {
1.1 mickey 429:
430: if (s->z_stream.avail_in == 0) {
431:
432: errno = 0;
433: if ((s->z_stream.avail_in =
1.7 deraadt 434: read(s->z_fd, s->z_buf, Z_BUFSIZE)) == 0)
1.1 mickey 435: s->z_eof = 1;
436: s->z_stream.next_in = s->z_buf;
437: }
438:
1.17 millert 439: error = inflate(&(s->z_stream), Z_NO_FLUSH);
440: if (error == Z_STREAM_END) {
1.1 mickey 441: /* Check CRC and original size */
442: s->z_crc = crc32(s->z_crc, start,
1.7 deraadt 443: (uInt)(s->z_stream.next_out - start));
1.1 mickey 444: start = s->z_stream.next_out;
445:
1.9 millert 446: if (get_int32(s) != s->z_crc) {
447: errno = EINVAL;
448: return -1;
449: }
1.18 henning 450: if (get_int32(s) != (u_int32_t)s->z_stream.total_out) {
1.7 deraadt 451: errno = EIO;
1.1 mickey 452: return -1;
453: }
1.13 millert 454: s->z_hlen += 2 * sizeof(int32_t);
1.16 millert 455: /* Check for the existence of an appended file. */
456: if (get_header(s, NULL, 0) != 0) {
457: s->z_eof = 1;
458: break;
459: }
460: inflateReset(&(s->z_stream));
461: s->z_crc = crc32(0L, Z_NULL, 0);
1.17 millert 462: error = Z_OK;
1.1 mickey 463: }
464: }
465: s->z_crc = crc32(s->z_crc, start,
1.7 deraadt 466: (uInt)(s->z_stream.next_out - start));
1.15 millert 467: len -= s->z_stream.avail_out;
468:
469: return (len);
1.1 mickey 470: }
471:
472: int
1.7 deraadt 473: gz_write(void *cookie, const char *buf, int len)
1.1 mickey 474: {
1.20 ! henning 475: #ifndef SMALL
1.2 mpech 476: gz_stream *s = (gz_stream*)cookie;
1.1 mickey 477:
478: s->z_stream.next_in = (char *)buf;
479: s->z_stream.avail_in = len;
480:
481: while (s->z_stream.avail_in != 0) {
482: if (s->z_stream.avail_out == 0) {
483: if (write(s->z_fd, s->z_buf, Z_BUFSIZE) != Z_BUFSIZE)
484: break;
485: s->z_stream.next_out = s->z_buf;
486: s->z_stream.avail_out = Z_BUFSIZE;
487: }
488: if (deflate(&(s->z_stream), Z_NO_FLUSH) != Z_OK)
489: break;
490: }
491: s->z_crc = crc32(s->z_crc, buf, len);
492:
493: return (int)(len - s->z_stream.avail_in);
1.20 ! henning 494: #endif
1.1 mickey 495: }