libsa/mkext.c

   1 /*
   2  * Copyright (c) 2000-2004 Apple Computer, Inc. All rights reserved.
   3  *
   4  * @APPLE_LICENSE_OSREFERENCE_HEADER_START@
   5  *
   6  * This file contains Original Code and/or Modifications of Original Code
   7  * as defined in and that are subject to the Apple Public Source License
   8  * Version 2.0 (the 'License'). You may not use this file except in
   9  * compliance with the License.  The rights granted to you under the
  10  * License may not be used to create, or enable the creation or
  11  * redistribution of, unlawful or unlicensed copies of an Apple operating
  12  * system, or to circumvent, violate, or enable the circumvention or
  13  * violation of, any terms of an Apple operating system software license
  14  * agreement.
  15  *
  16  * Please obtain a copy of the License at
  17  * http://www.opensource.apple.com/apsl/ and read it before using this
  18  * file.
  19  *
  20  * The Original Code and all software distributed under the License are
  21  * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
  22  * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
  23  * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
  24  * FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT.
  25  * Please see the License for the specific language governing rights and
  26  * limitations under the License.
  27  *
  28  * @APPLE_LICENSE_OSREFERENCE_HEADER_END@
  29  */
  30 #include <string.h>
  31 #if KERNEL
  32 #include <libsa/mkext.h>
  33 #include <libsa/stdlib.h>
  34 #else
  35 #include <Kernel/libsa/mkext.h>
  36 #include <stdlib.h>
  37 #endif /* KERNEL */
  38
  39 #define BASE 65521L /* largest prime smaller than 65536 */
  40 #define NMAX 5000
  41 // NMAX (was 5521) the largest n such that 255n(n+1)/2 + (n+1)(BASE-1) <= 2^32-1
  42
  43 #define DO1(buf,i)  {s1 += buf[i]; s2 += s1;}
  44 #define DO2(buf,i)  DO1(buf,i); DO1(buf,i+1);
  45 #define DO4(buf,i)  DO2(buf,i); DO2(buf,i+2);
  46 #define DO8(buf,i)  DO4(buf,i); DO4(buf,i+4);
  47 #define DO16(buf)   DO8(buf,0); DO8(buf,8);
  48
  49 __private_extern__ u_int32_t
  50 adler32(uint8_t *buf, int32_t len)
  51 {
  52     unsigned long s1 = 1; // adler & 0xffff;
  53     unsigned long s2 = 0; // (adler >> 16) & 0xffff;
  54     int k;
  55
  56     while (len > 0) {
  57         k = len < NMAX ? len : NMAX;
  58         len -= k;
  59         while (k >= 16) {
  60             DO16(buf);
  61             buf += 16;
  62             k -= 16;
  63         }
  64         if (k != 0) do {
  65             s1 += *buf++;
  66             s2 += s1;
  67         } while (--k);
  68         s1 %= BASE;
  69         s2 %= BASE;
  70     }
  71     return (s2 << 16) | s1;
  72 }
  73
  74
  75 /**************************************************************
  76  LZSS.C -- A Data Compression Program
  77 ***************************************************************
  78     4/6/1989 Haruhiko Okumura
  79     Use, distribute, and modify this program freely.
  80     Please send me your improved versions.
  81         PC-VAN      SCIENCE
  82         NIFTY-Serve PAF01022
  83         CompuServe  74050,1022
  84
  85 **************************************************************/
  86
  87 #define N         4096  /* size of ring buffer - must be power of 2 */
  88 #define F         18    /* upper limit for match_length */
  89 #define THRESHOLD 2     /* encode string into position and length
  90                            if match_length is greater than this */
  91 #define NIL       N     /* index for root of binary search trees */
  92
  93 struct encode_state {
  94     /*
  95      * left & right children & parent. These constitute binary search trees.
  96      */
  97     int lchild[N + 1], rchild[N + 257], parent[N + 1];
  98
  99     /* ring buffer of size N, with extra F-1 bytes to aid string comparison */
 100     u_int8_t text_buf[N + F - 1];
 101
 102     /*
 103      * match_length of longest match.
 104      * These are set by the insert_node() procedure.
 105      */
 106     int match_position, match_length;
 107 };
 108
 109
 110 __private_extern__ int
 111 decompress_lzss(u_int8_t *dst, u_int8_t *src, u_int32_t srclen)
 112 {
 113     /* ring buffer of size N, with extra F-1 bytes to aid string comparison */
 114     u_int8_t text_buf[N + F - 1];
 115     u_int8_t *dststart = dst;
 116     u_int8_t *srcend = src + srclen;
 117     int  i, j, k, r, c;
 118     unsigned int flags;
 119
 120     dst = dststart;
 121     srcend = src + srclen;
 122     for (i = 0; i < N - F; i++)
 123         text_buf[i] = ' ';
 124     r = N - F;
 125     flags = 0;
 126     for ( ; ; ) {
 127         if (((flags >>= 1) & 0x100) == 0) {
 128             if (src < srcend) c = *src++; else break;
 129             flags = c | 0xFF00;  /* uses higher byte cleverly */
 130         }   /* to count eight */
 131         if (flags & 1) {
 132             if (src < srcend) c = *src++; else break;
 133             *dst++ = c;
 134             text_buf[r++] = c;
 135             r &= (N - 1);
 136         } else {
 137             if (src < srcend) i = *src++; else break;
 138             if (src < srcend) j = *src++; else break;
 139             i |= ((j & 0xF0) << 4);
 140             j  =  (j & 0x0F) + THRESHOLD;
 141             for (k = 0; k <= j; k++) {
 142                 c = text_buf[(i + k) & (N - 1)];
 143                 *dst++ = c;
 144                 text_buf[r++] = c;
 145                 r &= (N - 1);
 146             }
 147         }
 148     }
 149
 150     return dst - dststart;
 151 }
 152
 153 #if !KERNEL
 154
 155 /*
 156  * initialize state, mostly the trees
 157  *
 158  * For i = 0 to N - 1, rchild[i] and lchild[i] will be the right and left
 159  * children of node i.  These nodes need not be initialized.  Also, parent[i]
 160  * is the parent of node i.  These are initialized to NIL (= N), which stands
 161  * for 'not used.'  For i = 0 to 255, rchild[N + i + 1] is the root of the
 162  * tree for strings that begin with character i.  These are initialized to NIL.
 163  * Note there are 256 trees. */
 164 static void init_state(struct encode_state *sp)
 165 {
 166     int  i;
 167
 168     bzero(sp, sizeof(*sp));
 169
 170     for (i = 0; i < N - F; i++)
 171         sp->text_buf[i] = ' ';
 172     for (i = N + 1; i <= N + 256; i++)
 173         sp->rchild[i] = NIL;
 174     for (i = 0; i < N; i++)
 175         sp->parent[i] = NIL;
 176 }
 177
 178 /*
 179  * Inserts string of length F, text_buf[r..r+F-1], into one of the trees
 180  * (text_buf[r]'th tree) and returns the longest-match position and length
 181  * via the global variables match_position and match_length.
 182  * If match_length = F, then removes the old node in favor of the new one,
 183  * because the old one will be deleted sooner. Note r plays double role,
 184  * as tree node and position in buffer.
 185  */
 186 static void insert_node(struct encode_state *sp, int r)
 187 {
 188     int  i, p, cmp;
 189     u_int8_t  *key;
 190
 191     cmp = 1;
 192     key = &sp->text_buf[r];
 193     p = N + 1 + key[0];
 194     sp->rchild[r] = sp->lchild[r] = NIL;
 195     sp->match_length = 0;
 196     for ( ; ; ) {
 197         if (cmp >= 0) {
 198             if (sp->rchild[p] != NIL)
 199                 p = sp->rchild[p];
 200             else {
 201                 sp->rchild[p] = r;
 202                 sp->parent[r] = p;
 203                 return;
 204             }
 205         } else {
 206             if (sp->lchild[p] != NIL)
 207                 p = sp->lchild[p];
 208             else {
 209                 sp->lchild[p] = r;
 210                 sp->parent[r] = p;
 211                 return;
 212             }
 213         }
 214         for (i = 1; i < F; i++) {
 215             if ((cmp = key[i] - sp->text_buf[p + i]) != 0)
 216                 break;
 217         }
 218         if (i > sp->match_length) {
 219             sp->match_position = p;
 220             if ((sp->match_length = i) >= F)
 221                 break;
 222         }
 223     }
 224     sp->parent[r] = sp->parent[p];
 225     sp->lchild[r] = sp->lchild[p];
 226     sp->rchild[r] = sp->rchild[p];
 227     sp->parent[sp->lchild[p]] = r;
 228     sp->parent[sp->rchild[p]] = r;
 229     if (sp->rchild[sp->parent[p]] == p)
 230         sp->rchild[sp->parent[p]] = r;
 231     else
 232         sp->lchild[sp->parent[p]] = r;
 233     sp->parent[p] = NIL;  /* remove p */
 234 }
 235
 236 /* deletes node p from tree */
 237 static void delete_node(struct encode_state *sp, int p)
 238 {
 239     int  q;
 240
 241     if (sp->parent[p] == NIL)
 242         return;  /* not in tree */
 243     if (sp->rchild[p] == NIL)
 244         q = sp->lchild[p];
 245     else if (sp->lchild[p] == NIL)
 246         q = sp->rchild[p];
 247     else {
 248         q = sp->lchild[p];
 249         if (sp->rchild[q] != NIL) {
 250             do {
 251                 q = sp->rchild[q];
 252             } while (sp->rchild[q] != NIL);
 253             sp->rchild[sp->parent[q]] = sp->lchild[q];
 254             sp->parent[sp->lchild[q]] = sp->parent[q];
 255             sp->lchild[q] = sp->lchild[p];
 256             sp->parent[sp->lchild[p]] = q;
 257         }
 258         sp->rchild[q] = sp->rchild[p];
 259         sp->parent[sp->rchild[p]] = q;
 260     }
 261     sp->parent[q] = sp->parent[p];
 262     if (sp->rchild[sp->parent[p]] == p)
 263         sp->rchild[sp->parent[p]] = q;
 264     else
 265         sp->lchild[sp->parent[p]] = q;
 266     sp->parent[p] = NIL;
 267 }
 268
 269 __private_extern__ u_int8_t *
 270 compress_lzss(u_int8_t *dst, u_int32_t dstlen, u_int8_t *src, u_int32_t srcLen)
 271 {
 272     /* Encoding state, mostly tree but some current match stuff */
 273     struct encode_state *sp;
 274
 275     int  i, c, len, r, s, last_match_length, code_buf_ptr;
 276     u_int8_t code_buf[17], mask;
 277     u_int8_t *srcend = src + srcLen;
 278     u_int8_t *dstend = dst + dstlen;
 279
 280     /* initialize trees */
 281     sp = (struct encode_state *) malloc(sizeof(*sp));
 282     init_state(sp);
 283
 284     /*
 285      * code_buf[1..16] saves eight units of code, and code_buf[0] works
 286      * as eight flags, "1" representing that the unit is an unencoded
 287      * letter (1 byte), "0" a position-and-length pair (2 bytes).
 288      * Thus, eight units require at most 16 bytes of code.
 289      */
 290     code_buf[0] = 0;
 291     code_buf_ptr = mask = 1;
 292
 293     /* Clear the buffer with any character that will appear often. */
 294     s = 0;  r = N - F;
 295
 296     /* Read F bytes into the last F bytes of the buffer */
 297     for (len = 0; len < F && src < srcend; len++)
 298         sp->text_buf[r + len] = *src++;
 299     if (!len)
 300         return (void *) 0;  /* text of size zero */
 301
 302     /*
 303      * Insert the F strings, each of which begins with one or more
 304      * 'space' characters.  Note the order in which these strings are
 305      * inserted.  This way, degenerate trees will be less likely to occur.
 306      */
 307     for (i = 1; i <= F; i++)
 308         insert_node(sp, r - i);
 309
 310     /*
 311      * Finally, insert the whole string just read.
 312      * The global variables match_length and match_position are set.
 313      */
 314     insert_node(sp, r);
 315     do {
 316         /* match_length may be spuriously long near the end of text. */
 317         if (sp->match_length > len)
 318             sp->match_length = len;
 319         if (sp->match_length <= THRESHOLD) {
 320             sp->match_length = 1;  /* Not long enough match.  Send one byte. */
 321             code_buf[0] |= mask;  /* 'send one byte' flag */
 322             code_buf[code_buf_ptr++] = sp->text_buf[r];  /* Send uncoded. */
 323         } else {
 324             /* Send position and length pair. Note match_length > THRESHOLD. */
 325             code_buf[code_buf_ptr++] = (u_int8_t) sp->match_position;
 326             code_buf[code_buf_ptr++] = (u_int8_t)
 327                 ( ((sp->match_position >> 4) & 0xF0)
 328                 |  (sp->match_length - (THRESHOLD + 1)) );
 329         }
 330         if ((mask <<= 1) == 0) {  /* Shift mask left one bit. */
 331                 /* Send at most 8 units of code together */
 332             for (i = 0; i < code_buf_ptr; i++)
 333                 if (dst < dstend)
 334                     *dst++ = code_buf[i];
 335                 else
 336                     return (void *) 0;
 337             code_buf[0] = 0;
 338             code_buf_ptr = mask = 1;
 339         }
 340         last_match_length = sp->match_length;
 341         for (i = 0; i < last_match_length && src < srcend; i++) {
 342             delete_node(sp, s);    /* Delete old strings and */
 343             c = *src++;
 344             sp->text_buf[s] = c;    /* read new bytes */
 345
 346             /*
 347              * If the position is near the end of buffer, extend the buffer
 348              * to make string comparison easier.
 349              */
 350             if (s < F - 1)
 351                 sp->text_buf[s + N] = c;
 352
 353             /* Since this is a ring buffer, increment the position modulo N. */
 354             s = (s + 1) & (N - 1);
 355             r = (r + 1) & (N - 1);
 356
 357             /* Register the string in text_buf[r..r+F-1] */
 358             insert_node(sp, r);
 359         }
 360         while (i++ < last_match_length) {
 361         delete_node(sp, s);
 362
 363             /* After the end of text, no need to read, */
 364             s = (s + 1) & (N - 1);
 365             r = (r + 1) & (N - 1);
 366             /* but buffer may not be empty. */
 367             if (--len)
 368                 insert_node(sp, r);
 369         }
 370     } while (len > 0);   /* until length of string to be processed is zero */
 371
 372     if (code_buf_ptr > 1) {    /* Send remaining code. */
 373         for (i = 0; i < code_buf_ptr; i++)
 374             if (dst < dstend)
 375                 *dst++ = code_buf[i];
 376             else
 377                 return (void *) 0;
 378     }
 379
 380     return dst;
 381 }
 382
 383 #endif /* !KERNEL */
 384