Annotation of src/usr.bin/compress/gzopen.c, Revision 1.19
1.19 ! henning 1: /* $OpenBSD: gzopen.c,v 1.18 2003/12/16 22:38:40 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[] =
1.19 ! henning 63: "$OpenBSD: gzopen.c,v 1.18 2003/12/16 22:38:40 henning Exp $";
! 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') {
140: /* windowBits is passed < 0 to suppress zlib header */
141: if (deflateInit2(&(s->z_stream), bits, Z_DEFLATED,
142: -MAX_WBITS, DEF_MEM_LEVEL, 0) != Z_OK) {
143: free (s);
144: return NULL;
145: }
146: s->z_stream.next_out = s->z_buf;
147: } else {
148: if (inflateInit2(&(s->z_stream), -MAX_WBITS) != Z_OK) {
149: free (s);
150: return NULL;
151: }
152: s->z_stream.next_in = s->z_buf;
153: }
154: s->z_stream.avail_out = Z_BUFSIZE;
155:
156: errno = 0;
157: s->z_fd = fd;
158:
159: if (s->z_mode == 'w') {
1.13 millert 160: /* write the .gz header */
161: if (put_header(s, name, mtime) != 0) {
162: gz_close(s, NULL);
1.1 mickey 163: s = NULL;
164: }
165: } else {
1.13 millert 166: /* read the .gz header */
167: if (get_header(s, name, gotmagic) != 0) {
168: gz_close(s, NULL);
1.1 mickey 169: s = NULL;
170: }
171: }
172:
173: return s;
174: }
175:
176: int
1.13 millert 177: gz_close(void *cookie, struct z_info *info)
1.1 mickey 178: {
1.2 mpech 179: gz_stream *s = (gz_stream*)cookie;
1.1 mickey 180: int err = 0;
181:
182: if (s == NULL)
183: return -1;
184:
185: if (s->z_mode == 'w' && (err = gz_flush (s, Z_FINISH)) == Z_OK) {
1.13 millert 186: if ((err = put_int32 (s, s->z_crc)) == Z_OK) {
187: s->z_hlen += sizeof(int32_t);
188: if ((err = put_int32 (s, s->z_stream.total_in)) == Z_OK)
189: s->z_hlen += sizeof(int32_t);
190: }
1.1 mickey 191: }
192:
193: if (!err && s->z_stream.state != NULL) {
194: if (s->z_mode == 'w')
195: err = deflateEnd(&s->z_stream);
196: else if (s->z_mode == 'r')
197: err = inflateEnd(&s->z_stream);
198: }
1.10 mickey 199:
1.13 millert 200: if (info != NULL) {
201: info->mtime = s->z_time;
202: info->crc = s->z_crc;
203: info->hlen = s->z_hlen;
204: info->total_in = (off_t)s->z_stream.total_in;
205: info->total_out = (off_t)s->z_stream.total_out;
206: }
207:
1.10 mickey 208: if (!err)
209: err = close(s->z_fd);
210: else
211: (void)close(s->z_fd);
1.1 mickey 212:
213: free(s);
214:
215: return err;
216: }
217:
218: int
1.7 deraadt 219: gz_flush(void *cookie, int flush)
1.1 mickey 220: {
1.2 mpech 221: gz_stream *s = (gz_stream*)cookie;
1.1 mickey 222: size_t len;
223: int done = 0;
224: int err;
225:
226: if (s == NULL || s->z_mode != 'w') {
227: errno = EBADF;
228: return Z_ERRNO;
229: }
230:
231: s->z_stream.avail_in = 0; /* should be zero already anyway */
232:
233: for (;;) {
234: len = Z_BUFSIZE - s->z_stream.avail_out;
235:
236: if (len != 0) {
237: if (write(s->z_fd, s->z_buf, len) != len)
238: return Z_ERRNO;
239: s->z_stream.next_out = s->z_buf;
240: s->z_stream.avail_out = Z_BUFSIZE;
241: }
242: if (done)
243: break;
244: if ((err = deflate(&(s->z_stream), flush)) != Z_OK &&
245: err != Z_STREAM_END)
246: return err;
247:
248: /* deflate has finished flushing only when it hasn't
249: * used up all the available space in the output buffer
250: */
251: done = (s->z_stream.avail_out != 0 || err == Z_STREAM_END);
252: }
253: return 0;
254: }
255:
256: static int
1.7 deraadt 257: put_int32(gz_stream *s, u_int32_t x)
1.1 mickey 258: {
1.8 millert 259: u_int32_t y = htole32(x);
260:
261: if (write(s->z_fd, &y, sizeof(y)) != sizeof(y))
1.1 mickey 262: return Z_ERRNO;
263: return 0;
264: }
265:
266: static int
1.7 deraadt 267: get_byte(gz_stream *s)
1.1 mickey 268: {
269: if (s->z_eof)
270: return EOF;
1.5 mickey 271:
1.1 mickey 272: if (s->z_stream.avail_in == 0) {
273: errno = 0;
274: s->z_stream.avail_in = read(s->z_fd, s->z_buf, Z_BUFSIZE);
275: if (s->z_stream.avail_in <= 0) {
276: s->z_eof = 1;
277: return EOF;
278: }
279: s->z_stream.next_in = s->z_buf;
280: }
281: s->z_stream.avail_in--;
282: return *s->z_stream.next_in++;
283: }
284:
1.5 mickey 285: static u_int32_t
1.7 deraadt 286: get_int32(gz_stream *s)
1.1 mickey 287: {
1.2 mpech 288: u_int32_t x;
1.1 mickey 289:
290: x = ((u_int32_t)(get_byte(s) & 0xff));
291: x |= ((u_int32_t)(get_byte(s) & 0xff))<<8;
292: x |= ((u_int32_t)(get_byte(s) & 0xff))<<16;
293: x |= ((u_int32_t)(get_byte(s) & 0xff))<<24;
294: return x;
295: }
296:
297: static int
1.13 millert 298: get_header(gz_stream *s, char *name, int gotmagic)
1.1 mickey 299: {
300: int method; /* method byte */
301: int flags; /* flags byte */
1.13 millert 302: char *ep;
1.1 mickey 303: uInt len;
304: int c;
305:
306: /* Check the gzip magic header */
1.12 millert 307: if (!gotmagic) {
308: for (len = 0; len < 2; len++) {
309: c = get_byte(s);
310: if (c != gz_magic[len]) {
311: errno = EFTYPE;
312: return -1;
313: }
1.1 mickey 314: }
315: }
316:
317: method = get_byte(s);
318: flags = get_byte(s);
319: if (method != Z_DEFLATED || (flags & RESERVED) != 0) {
320: errno = EFTYPE;
321: return -1;
322: }
323:
1.13 millert 324: /* Stash timestamp (mtime) */
325: s->z_time = get_int32(s);
326:
327: /* Discard xflags and OS code */
328: (void)get_byte(s);
329: (void)get_byte(s);
1.1 mickey 330:
1.13 millert 331: s->z_hlen = 10; /* magic, method, flags, time, xflags, OS code */
1.1 mickey 332: if ((flags & EXTRA_FIELD) != 0) { /* skip the extra field */
333: len = (uInt)get_byte(s);
334: len += ((uInt)get_byte(s))<<8;
1.13 millert 335: s->z_hlen += 2;
1.1 mickey 336: /* len is garbage if EOF but the loop below will quit anyway */
337: while (len-- != 0 && get_byte(s) != EOF)
1.13 millert 338: s->z_hlen++;
1.1 mickey 339: }
340:
1.13 millert 341: if ((flags & ORIG_NAME) != 0) { /* read/save the original file name */
342: if ((ep = name) != NULL)
343: ep += MAXPATHLEN - 1;
344: while ((c = get_byte(s)) != EOF) {
345: s->z_hlen++;
346: if (c == '\0')
347: break;
348: if (name < ep)
349: *name++ = c;
350: }
351: if (name != NULL)
352: *name = '\0';
1.1 mickey 353: }
354:
355: if ((flags & COMMENT) != 0) { /* skip the .gz file comment */
1.13 millert 356: while ((c = get_byte(s)) != EOF) {
357: s->z_hlen++;
358: if (c == '\0')
359: break;
360: }
1.1 mickey 361: }
362:
363: if ((flags & HEAD_CRC) != 0) { /* skip the header crc */
1.14 mickey 364: (void)get_byte(s);
365: (void)get_byte(s);
1.13 millert 366: s->z_hlen += 2;
1.1 mickey 367: }
368:
369: if (s->z_eof) {
370: errno = EFTYPE;
371: return -1;
372: }
373:
374: return 0;
375: }
376:
1.13 millert 377: static int
378: put_header(gz_stream *s, char *name, u_int32_t mtime)
379: {
380: struct iovec iov[2];
381: u_char buf[10];
382:
383: buf[0] = gz_magic[0];
384: buf[1] = gz_magic[1];
385: buf[2] = Z_DEFLATED;
386: buf[3] = name ? ORIG_NAME : 0;
387: buf[4] = mtime & 0xff;
388: buf[5] = (mtime >> 8) & 0xff;
389: buf[6] = (mtime >> 16) & 0xff;
390: buf[7] = (mtime >> 24) & 0xff;
391: buf[8] = 0 /* xflags */;
392: buf[9] = OS_CODE;
393: iov[0].iov_base = buf;
394: iov[0].iov_len = sizeof(buf);
395: s->z_hlen = sizeof(buf);
396:
397: if (name != NULL) {
398: iov[1].iov_base = name;
399: iov[1].iov_len = strlen(name) + 1;
400: s->z_hlen += iov[1].iov_len;
401: }
402: if (writev(s->z_fd, iov, name ? 2 : 1) == -1)
403: return (-1);
404: return (0);
405: }
406:
1.1 mickey 407: int
1.7 deraadt 408: gz_read(void *cookie, char *buf, int len)
1.1 mickey 409: {
1.2 mpech 410: gz_stream *s = (gz_stream*)cookie;
1.1 mickey 411: u_char *start = buf; /* starting point for crc computation */
1.17 millert 412: int error = Z_OK;
1.1 mickey 413:
414: s->z_stream.next_out = buf;
415: s->z_stream.avail_out = len;
416:
1.17 millert 417: while (error == Z_OK && !s->z_eof && s->z_stream.avail_out != 0) {
1.1 mickey 418:
419: if (s->z_stream.avail_in == 0) {
420:
421: errno = 0;
422: if ((s->z_stream.avail_in =
1.7 deraadt 423: read(s->z_fd, s->z_buf, Z_BUFSIZE)) == 0)
1.1 mickey 424: s->z_eof = 1;
425: s->z_stream.next_in = s->z_buf;
426: }
427:
1.17 millert 428: error = inflate(&(s->z_stream), Z_NO_FLUSH);
429: if (error == Z_STREAM_END) {
1.1 mickey 430: /* Check CRC and original size */
431: s->z_crc = crc32(s->z_crc, start,
1.7 deraadt 432: (uInt)(s->z_stream.next_out - start));
1.1 mickey 433: start = s->z_stream.next_out;
434:
1.9 millert 435: if (get_int32(s) != s->z_crc) {
436: errno = EINVAL;
437: return -1;
438: }
1.18 henning 439: if (get_int32(s) != (u_int32_t)s->z_stream.total_out) {
1.7 deraadt 440: errno = EIO;
1.1 mickey 441: return -1;
442: }
1.13 millert 443: s->z_hlen += 2 * sizeof(int32_t);
1.16 millert 444: /* Check for the existence of an appended file. */
445: if (get_header(s, NULL, 0) != 0) {
446: s->z_eof = 1;
447: break;
448: }
449: inflateReset(&(s->z_stream));
450: s->z_crc = crc32(0L, Z_NULL, 0);
1.17 millert 451: error = Z_OK;
1.1 mickey 452: }
453: }
454: s->z_crc = crc32(s->z_crc, start,
1.7 deraadt 455: (uInt)(s->z_stream.next_out - start));
1.15 millert 456: len -= s->z_stream.avail_out;
457:
458: return (len);
1.1 mickey 459: }
460:
461: int
1.7 deraadt 462: gz_write(void *cookie, const char *buf, int len)
1.1 mickey 463: {
1.2 mpech 464: gz_stream *s = (gz_stream*)cookie;
1.1 mickey 465:
466: s->z_stream.next_in = (char *)buf;
467: s->z_stream.avail_in = len;
468:
469: while (s->z_stream.avail_in != 0) {
470: if (s->z_stream.avail_out == 0) {
471: if (write(s->z_fd, s->z_buf, Z_BUFSIZE) != Z_BUFSIZE)
472: break;
473: s->z_stream.next_out = s->z_buf;
474: s->z_stream.avail_out = Z_BUFSIZE;
475: }
476: if (deflate(&(s->z_stream), Z_NO_FLUSH) != Z_OK)
477: break;
478: }
479: s->z_crc = crc32(s->z_crc, buf, len);
480:
481: return (int)(len - s->z_stream.avail_in);
482: }