1 /* $NetBSD: zuncompress.c,v 1.11 2011/08/16 13:55:02 joerg Exp $ */
4 * Copyright (c) 1985, 1986, 1992, 1993
5 * The Regents of the University of California. All rights reserved.
7 * This code is derived from software contributed to Berkeley by
8 * Diomidis Spinellis and James A. Woods, derived from original
9 * work by Spencer Thomas and Joseph Orost.
11 * Redistribution and use in source and binary forms, with or without
12 * modification, are permitted provided that the following conditions
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 * 3. Neither the name of the University nor the names of its contributors
20 * may be used to endorse or promote products derived from this software
21 * without specific prior written permission.
23 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
24 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
25 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
26 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
27 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
28 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
29 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
30 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
31 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
32 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
35 * from: NetBSD: zopen.c,v 1.8 2003/08/07 11:13:29 agc Exp
38 /* This file is #included by gzip.c */
40 static int zread(void *, char *, int);
42 #define tab_prefixof(i) (zs->zs_codetab[i])
43 #define tab_suffixof(i) ((char_type *)(zs->zs_htab))[i]
44 #define de_stack ((char_type *)&tab_suffixof(1 << BITS))
46 #define BITS 16 /* Default bits. */
47 #define HSIZE 69001 /* 95% occupancy */ /* XXX may not need HSIZE */
48 #define BIT_MASK 0x1f /* Defines for third byte of header. */
49 #define BLOCK_MASK 0x80
50 #define CHECK_GAP 10000 /* Ratio check interval. */
51 #define BUFSIZE (64 * 1024)
54 * Masks 0x40 and 0x20 are free. I think 0x20 should mean that there is
55 * a fourth header byte (for expansion).
57 #define INIT_BITS 9 /* Initial number of bits/code. */
60 * the next two codes should not be changed lightly, as they must not
61 * lie within the contiguous general code space.
63 #define FIRST 257 /* First free entry. */
64 #define CLEAR 256 /* Table clear output code. */
67 #define MAXCODE(n_bits) ((1 << (n_bits)) - 1)
69 typedef long code_int
;
70 typedef long count_int
;
71 typedef u_char char_type
;
73 static char_type magic_header
[] =
74 {'\037', '\235'}; /* 1F 9D */
76 static char_type rmask
[9] =
77 {0x00, 0x01, 0x03, 0x07, 0x0f, 0x1f, 0x3f, 0x7f, 0xff};
79 /* XXX zuncompress global */
80 off_t total_compressed_bytes
;
81 size_t compressed_prelen
;
85 FILE *zs_fp
; /* File stream for I/O */
86 char zs_mode
; /* r or w */
88 S_START
, S_MIDDLE
, S_EOF
89 } zs_state
; /* State of computation */
90 int zs_n_bits
; /* Number of bits/code. */
91 int zs_maxbits
; /* User settable max # bits/code. */
92 code_int zs_maxcode
; /* Maximum code, given n_bits. */
93 code_int zs_maxmaxcode
; /* Should NEVER generate this code. */
94 count_int zs_htab
[HSIZE
];
95 u_short zs_codetab
[HSIZE
];
96 code_int zs_hsize
; /* For dynamic table sizing. */
97 code_int zs_free_ent
; /* First unused entry. */
99 * Block compression parameters -- after all codes are used up,
100 * and compression rate changes, start over.
102 int zs_block_compress
;
105 count_int zs_checkpoint
;
107 long zs_in_count
; /* Length of input. */
108 long zs_bytes_out
; /* Length of compressed output. */
109 long zs_out_count
; /* # of codes output (for debugging). */
110 char_type zs_buf
[BITS
];
115 code_int zs_hsize_reg
;
117 } w
; /* Write parameters */
119 char_type
*zs_stackp
;
121 code_int zs_code
, zs_oldcode
, zs_incode
;
122 int zs_roffset
, zs_size
;
123 char_type zs_gbuf
[BITS
];
124 } r
; /* Read parameters */
128 static code_int
getcode(struct s_zstate
*zs
);
131 zuncompress(FILE *in
, FILE *out
, char *pre
, size_t prelen
,
132 off_t
*compressed_bytes
)
137 buf
= malloc(BUFSIZE
);
142 compressed_prelen
= prelen
;
144 compressed_pre
= pre
;
146 compressed_pre
= NULL
;
148 while ((bin
= fread(buf
, 1, sizeof(buf
), in
)) != 0) {
149 if (tflag
== 0 && (off_t
)fwrite(buf
, 1, bin
, out
) != bin
) {
156 if (compressed_bytes
)
157 *compressed_bytes
= total_compressed_bytes
;
167 /* We leave the caller to close the fd passed to zdopen() */
176 if ((zs
= calloc(1, sizeof(struct s_zstate
))) == NULL
)
179 zs
->zs_state
= S_START
;
181 /* XXX we can get rid of some of these */
182 zs
->zs_hsize
= HSIZE
; /* For dynamic table sizing. */
183 zs
->zs_free_ent
= 0; /* First unused entry. */
184 zs
->zs_block_compress
= BLOCK_MASK
;
185 zs
->zs_clear_flg
= 0; /* XXX we calloc()'d this structure why = 0? */
187 zs
->zs_checkpoint
= CHECK_GAP
;
188 zs
->zs_in_count
= 1; /* Length of input. */
189 zs
->zs_out_count
= 0; /* # of codes output (for debugging). */
190 zs
->u
.r
.zs_roffset
= 0;
194 * Layering compress on top of stdio in order to provide buffering,
195 * and ensure that reads and write work with the data specified.
197 if ((zs
->zs_fp
= fdopen(fd
, "r")) == NULL
) {
202 return funopen(zs
, zread
, NULL
, NULL
, zclose
);
206 * Decompress read. This routine adapts to the codes in the file building
207 * the "string" table on-the-fly; requiring no table to be stored in the
208 * compressed file. The tables used herein are shared with those of the
209 * compress() routine. See the definitions above.
212 zread(void *cookie
, char *rbp
, int num
)
216 u_char
*bp
, header
[3];
224 switch (zs
->zs_state
) {
226 zs
->zs_state
= S_MIDDLE
;
234 /* Check the magic number */
235 for (i
= 0; i
< 3 && compressed_prelen
; i
++, compressed_prelen
--)
236 header
[i
] = *compressed_pre
++;
238 if (fread(header
+ i
, 1, sizeof(header
) - i
, zs
->zs_fp
) !=
239 sizeof(header
) - i
||
240 memcmp(header
, magic_header
, sizeof(magic_header
)) != 0) {
244 total_compressed_bytes
= 0;
245 zs
->zs_maxbits
= header
[2]; /* Set -b from file. */
246 zs
->zs_block_compress
= zs
->zs_maxbits
& BLOCK_MASK
;
247 zs
->zs_maxbits
&= BIT_MASK
;
248 zs
->zs_maxmaxcode
= 1L << zs
->zs_maxbits
;
249 if (zs
->zs_maxbits
> BITS
|| zs
->zs_maxbits
< 12) {
253 /* As above, initialize the first 256 entries in the table. */
254 zs
->zs_maxcode
= MAXCODE(zs
->zs_n_bits
= INIT_BITS
);
255 for (zs
->u
.r
.zs_code
= 255; zs
->u
.r
.zs_code
>= 0; zs
->u
.r
.zs_code
--) {
256 tab_prefixof(zs
->u
.r
.zs_code
) = 0;
257 tab_suffixof(zs
->u
.r
.zs_code
) = (char_type
) zs
->u
.r
.zs_code
;
259 zs
->zs_free_ent
= zs
->zs_block_compress
? FIRST
: 256;
261 zs
->u
.r
.zs_oldcode
= -1;
262 zs
->u
.r
.zs_stackp
= de_stack
;
264 while ((zs
->u
.r
.zs_code
= getcode(zs
)) > -1) {
266 if ((zs
->u
.r
.zs_code
== CLEAR
) && zs
->zs_block_compress
) {
267 for (zs
->u
.r
.zs_code
= 255; zs
->u
.r
.zs_code
>= 0;
269 tab_prefixof(zs
->u
.r
.zs_code
) = 0;
270 zs
->zs_clear_flg
= 1;
271 zs
->zs_free_ent
= FIRST
;
272 zs
->u
.r
.zs_oldcode
= -1;
275 zs
->u
.r
.zs_incode
= zs
->u
.r
.zs_code
;
277 /* Special case for KwKwK string. */
278 if (zs
->u
.r
.zs_code
>= zs
->zs_free_ent
) {
279 if (zs
->u
.r
.zs_code
> zs
->zs_free_ent
||
280 zs
->u
.r
.zs_oldcode
== -1) {
285 *zs
->u
.r
.zs_stackp
++ = zs
->u
.r
.zs_finchar
;
286 zs
->u
.r
.zs_code
= zs
->u
.r
.zs_oldcode
;
289 * The above condition ensures that code < free_ent.
290 * The construction of tab_prefixof in turn guarantees that
291 * each iteration decreases code and therefore stack usage is
292 * bound by 1 << BITS - 256.
295 /* Generate output characters in reverse order. */
296 while (zs
->u
.r
.zs_code
>= 256) {
297 *zs
->u
.r
.zs_stackp
++ = tab_suffixof(zs
->u
.r
.zs_code
);
298 zs
->u
.r
.zs_code
= tab_prefixof(zs
->u
.r
.zs_code
);
300 *zs
->u
.r
.zs_stackp
++ = zs
->u
.r
.zs_finchar
= tab_suffixof(zs
->u
.r
.zs_code
);
302 /* And put them out in forward order. */
306 *bp
++ = *--zs
->u
.r
.zs_stackp
;
307 } while (zs
->u
.r
.zs_stackp
> de_stack
);
309 /* Generate the new entry. */
310 if ((zs
->u
.r
.zs_code
= zs
->zs_free_ent
) < zs
->zs_maxmaxcode
&&
311 zs
->u
.r
.zs_oldcode
!= -1) {
312 tab_prefixof(zs
->u
.r
.zs_code
) = (u_short
) zs
->u
.r
.zs_oldcode
;
313 tab_suffixof(zs
->u
.r
.zs_code
) = zs
->u
.r
.zs_finchar
;
314 zs
->zs_free_ent
= zs
->u
.r
.zs_code
+ 1;
317 /* Remember previous code. */
318 zs
->u
.r
.zs_oldcode
= zs
->u
.r
.zs_incode
;
320 zs
->zs_state
= S_EOF
;
321 eof
: return (num
- count
);
325 * Read one code from the standard input. If EOF, return -1.
329 * code or -1 is returned.
332 getcode(struct s_zstate
*zs
)
338 bp
= zs
->u
.r
.zs_gbuf
;
339 if (zs
->zs_clear_flg
> 0 || zs
->u
.r
.zs_roffset
>= zs
->u
.r
.zs_size
||
340 zs
->zs_free_ent
> zs
->zs_maxcode
) {
342 * If the next entry will be too big for the current gcode
343 * size, then we must increase the size. This implies reading
344 * a new buffer full, too.
346 if (zs
->zs_free_ent
> zs
->zs_maxcode
) {
348 if (zs
->zs_n_bits
== zs
->zs_maxbits
) /* Won't get any bigger now. */
349 zs
->zs_maxcode
= zs
->zs_maxmaxcode
;
351 zs
->zs_maxcode
= MAXCODE(zs
->zs_n_bits
);
353 if (zs
->zs_clear_flg
> 0) {
354 zs
->zs_maxcode
= MAXCODE(zs
->zs_n_bits
= INIT_BITS
);
355 zs
->zs_clear_flg
= 0;
358 for (i
= 0; i
< zs
->zs_n_bits
&& compressed_prelen
; i
++, compressed_prelen
--)
359 zs
->u
.r
.zs_gbuf
[i
] = *compressed_pre
++;
360 zs
->u
.r
.zs_size
= fread(zs
->u
.r
.zs_gbuf
+ i
, 1, zs
->zs_n_bits
- i
, zs
->zs_fp
);
361 zs
->u
.r
.zs_size
+= i
;
362 if (zs
->u
.r
.zs_size
<= 0) /* End of file. */
364 zs
->u
.r
.zs_roffset
= 0;
366 total_compressed_bytes
+= zs
->u
.r
.zs_size
;
368 /* Round size down to integral number of codes. */
369 zs
->u
.r
.zs_size
= (zs
->u
.r
.zs_size
<< 3) - (zs
->zs_n_bits
- 1);
371 r_off
= zs
->u
.r
.zs_roffset
;
372 bits
= zs
->zs_n_bits
;
374 /* Get to the first byte. */
378 /* Get first part (low order bits). */
379 gcode
= (*bp
++ >> r_off
);
381 r_off
= 8 - r_off
; /* Now, roffset into gcode word. */
383 /* Get any 8 bit parts in the middle (<=1 for up to 16 bits). */
385 gcode
|= *bp
++ << r_off
;
390 /* High order bits. */
391 gcode
|= (*bp
& rmask
[bits
]) << r_off
;
392 zs
->u
.r
.zs_roffset
+= zs
->zs_n_bits
;