1 /* cabextract 0.6 - a program to extract Microsoft Cabinet files
2 * (C) 2000-2002 Stuart Caie <kyzer@4u.net>
3 * Modifications for Captive project by:
4 * Copyright (C) 2003 Jan Kratochvil <project-captive@jankratochvil.net>
6 * This program is free software; you can redistribute it and/or modify
7 * it under the terms of the GNU General Public License as published by
8 * the Free Software Foundation; either version 2 of the License, or
9 * (at your option) any later version.
11 * This program is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 * GNU General Public License for more details.
16 * You should have received a copy of the GNU General Public License
17 * along with this program; if not, write to the Free Software
18 * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
21 /* This is NOT a general purpose cabinet library with a front end tacked
22 * on. If you want a comprehensive library to read and write cabinet
23 * files, please get "libcabinet". If you want to create CAB files on UNIX
24 * systems, get "Cablinux".
26 * Get the official Microsoft CAB SDK from here:
27 * http://msdn.microsoft.com/workshop/management/cab/cab-sdk.exe
28 * You can use cabextract on this file to extract the contents.
30 * Many thanks to Dirk Stoecker, Matthew Russoto and Dave Tritscher and,
31 * of course, Microsoft for the documentation they _did_ provide wholly
32 * and accurately. MSZIP is a one-byte adaption of the deflate and inflate
33 * methods created by Phil Katz. Quantum is based on the Quantum archiver,
34 * created by David Stafford. LZX is an adaption of the LZX method created
35 * by Jonathan Forbes and Tomi Poutanen.
37 * Furthermore, thanks to Jae Jung, for single-handedly fixing all the
38 * bugs with LZX decompression in cabextract 0.1, and Eric Sharkey for the
39 * original manual page.
42 /* CAB files are 'cabinets'. 'Folders' store compressed data, and may span
43 * several cabinets. 'Files' live as data inside a folder when
44 * uncompressed. EOR checksums are used instead of CRCs. Four compression
45 * formats are known - NONE, MSZIP, QUANTUM and LZX. NONE is obviously
46 * uncompressed data. MSZIP is simply PKZIP's deflate/inflate algorithims
47 * with 'CK' as a signature instead of 'PK'. QUANTUM is an LZ77 +
48 * arithmetic coding method. LZX is a much loved LZH based archiver in the
49 * Amiga world, the algorithm taken (bought?) by Microsoft and tweaked for
53 #include "config.h" /* CAPTIVE */
58 #include <stdio.h> /* everyone has this! */
60 #ifdef HAVE_SYS_TYPES_H
61 # include <sys/types.h>
84 #ifdef HAVE_SYS_STAT_H
85 # include <sys/stat.h>
88 #if TIME_WITH_SYS_TIME
89 # include <sys/time.h>
93 # include <sys/time.h>
106 # define dirent direct
108 # include <sys/ndir.h>
111 # include <sys/dir.h>
120 # define strchr index
121 # define strrchr rindex
123 # if !HAVE_STRCASECMP
124 # define strcasecmp strcmpi
127 # define memcpy(d,,n) bcopy((s),(d),(n)
132 extern time_t mktime(struct tm *tp);
137 #else /* !HAVE_CONFIG_H */
143 #include <sys/stat.h>
144 #include <sys/types.h>
150 #define VERSION "x.x"
154 #include <glib/gmessages.h>
156 #include "../cabinet.h"
157 #include "cabextract.h"
160 # define D(x) printf x ;
166 /* number of bits in a ULONG */
168 # define CHAR_BIT (8)
170 #define ULONG_BITS (sizeof(ULONG) * CHAR_BIT)
172 /* endian-neutral reading of little-endian data */
173 #define EndGetI32(a) ((((a)[3])<<24)|(((a)[2])<<16)|(((a)[1])<<8)|((a)[0]))
174 #define EndGetI16(a) ((((a)[1])<<8)|((a)[0]))
176 /* maximum number of cabinets any one folder can be split across */
177 #define CAB_SPLITMAX (10)
181 struct cabinet *cab[CAB_SPLITMAX]; /* cabinet(s) this folder spans */
182 off_t offset[CAB_SPLITMAX]; /* offset to data blocks */
183 UWORD comp_type; /* compression format/window size */
184 ULONG comp_size; /* compressed size of folder */
185 UBYTE num_splits; /* number of split blocks + 1 */
186 UWORD num_blocks; /* total number of blocks */
187 struct file *contfile; /* the first split file */
191 /* structure offsets */
192 #define cfhead_Signature (0x00)
193 #define cfhead_CabinetSize (0x08)
194 #define cfhead_FileOffset (0x10)
195 #define cfhead_MinorVersion (0x18)
196 #define cfhead_MajorVersion (0x19)
197 #define cfhead_NumFolders (0x1A)
198 #define cfhead_NumFiles (0x1C)
199 #define cfhead_Flags (0x1E)
200 #define cfhead_SetID (0x20)
201 #define cfhead_CabinetIndex (0x22)
202 #define cfhead_SIZEOF (0x24)
203 #define cfheadext_HeaderReserved (0x00)
204 #define cfheadext_FolderReserved (0x02)
205 #define cfheadext_DataReserved (0x03)
206 #define cfheadext_SIZEOF (0x04)
207 #define cffold_DataOffset (0x00)
208 #define cffold_NumBlocks (0x04)
209 #define cffold_CompType (0x06)
210 #define cffold_SIZEOF (0x08)
211 #define cffile_UncompressedSize (0x00)
212 #define cffile_FolderOffset (0x04)
213 #define cffile_FolderIndex (0x08)
214 #define cffile_Date (0x0A)
215 #define cffile_Time (0x0C)
216 #define cffile_Attribs (0x0E)
217 #define cffile_SIZEOF (0x10)
218 #define cfdata_CheckSum (0x00)
219 #define cfdata_CompressedSize (0x04)
220 #define cfdata_UncompressedSize (0x06)
221 #define cfdata_SIZEOF (0x08)
224 #define cffoldCOMPTYPE_MASK (0x000f)
225 #define cffoldCOMPTYPE_NONE (0x0000)
226 #define cffoldCOMPTYPE_MSZIP (0x0001)
227 #define cffoldCOMPTYPE_QUANTUM (0x0002)
228 #define cffoldCOMPTYPE_LZX (0x0003)
229 #define cfheadPREV_CABINET (0x0001)
230 #define cfheadNEXT_CABINET (0x0002)
231 #define cfheadRESERVE_PRESENT (0x0004)
232 #define cffileCONTINUED_FROM_PREV (0xFFFD)
233 #define cffileCONTINUED_TO_NEXT (0xFFFE)
234 #define cffileCONTINUED_PREV_AND_NEXT (0xFFFF)
235 #define cffile_A_RDONLY (0x01)
236 #define cffile_A_HIDDEN (0x02)
237 #define cffile_A_SYSTEM (0x04)
238 #define cffile_A_ARCH (0x20)
239 #define cffile_A_EXEC (0x40)
240 #define cffile_A_NAME_IS_UTF (0x80)
243 /*--------------------------------------------------------------------------*/
244 /* our archiver information / state */
247 #define ZIPWSIZE 0x8000 /* window size */
248 #define ZIPLBITS 9 /* bits in base literal/length lookup table */
249 #define ZIPDBITS 6 /* bits in base distance lookup table */
250 #define ZIPBMAX 16 /* maximum bit length of any code */
251 #define ZIPN_MAX 288 /* maximum number of codes in any set */
254 UBYTE e; /* number of extra bits or operation */
255 UBYTE b; /* number of bits in this code or subcode */
257 UWORD n; /* literal, length base, or distance base */
258 struct Ziphuft *t; /* pointer to next level of table */
263 ULONG window_posn; /* current offset within the window */
264 ULONG bb; /* bit buffer */
265 ULONG bk; /* bits in bit buffer */
266 ULONG ll[288+32]; /* literal/length and distance code lengths */
267 ULONG c[ZIPBMAX+1]; /* bit length count table */
268 LONG lx[ZIPBMAX+1]; /* memory for l[-1..ZIPBMAX-1] */
269 struct Ziphuft *u[ZIPBMAX]; /* table stack */
270 ULONG v[ZIPN_MAX]; /* values in order of bit length */
271 ULONG x[ZIPBMAX+1]; /* bit offsets, then code stack */
282 int shiftsleft, entries;
283 struct QTMmodelsym *syms;
288 UBYTE *window; /* the actual decoding window */
289 ULONG window_size; /* window size (1Kb through 2Mb) */
290 ULONG actual_size; /* window size when it was first allocated */
291 ULONG window_posn; /* current offset within the window */
293 struct QTMmodel model7;
294 struct QTMmodelsym m7sym[7+1];
296 struct QTMmodel model4, model5, model6pos, model6len;
297 struct QTMmodelsym m4sym[0x18 + 1];
298 struct QTMmodelsym m5sym[0x24 + 1];
299 struct QTMmodelsym m6psym[0x2a + 1], m6lsym[0x1b + 1];
301 struct QTMmodel model00, model40, model80, modelC0;
302 struct QTMmodelsym m00sym[0x40 + 1], m40sym[0x40 + 1];
303 struct QTMmodelsym m80sym[0x40 + 1], mC0sym[0x40 + 1];
308 /* some constants defined by the LZX specification */
309 #define LZX_MIN_MATCH (2)
310 #define LZX_MAX_MATCH (257)
311 #define LZX_NUM_CHARS (256)
312 #define LZX_BLOCKTYPE_INVALID (0) /* also blocktypes 4-7 invalid */
313 #define LZX_BLOCKTYPE_VERBATIM (1)
314 #define LZX_BLOCKTYPE_ALIGNED (2)
315 #define LZX_BLOCKTYPE_UNCOMPRESSED (3)
316 #define LZX_PRETREE_NUM_ELEMENTS (20)
317 #define LZX_ALIGNED_NUM_ELEMENTS (8) /* aligned offset tree #elements */
318 #define LZX_NUM_PRIMARY_LENGTHS (7) /* this one missing from spec! */
319 #define LZX_NUM_SECONDARY_LENGTHS (249) /* length tree #elements */
321 /* LZX huffman defines: tweak tablebits as desired */
322 #define LZX_PRETREE_MAXSYMBOLS (LZX_PRETREE_NUM_ELEMENTS)
323 #define LZX_PRETREE_TABLEBITS (6)
324 #define LZX_MAINTREE_MAXSYMBOLS (LZX_NUM_CHARS + 50*8)
325 #define LZX_MAINTREE_TABLEBITS (12)
326 #define LZX_LENGTH_MAXSYMBOLS (LZX_NUM_SECONDARY_LENGTHS+1)
327 #define LZX_LENGTH_TABLEBITS (12)
328 #define LZX_ALIGNED_MAXSYMBOLS (LZX_ALIGNED_NUM_ELEMENTS)
329 #define LZX_ALIGNED_TABLEBITS (7)
331 #define LZX_LENTABLE_SAFETY (64) /* we allow length table decoding overruns */
333 #define LZX_DECLARE_TABLE(tbl) \
334 UWORD tbl##_table[(1<<LZX_##tbl##_TABLEBITS) + (LZX_##tbl##_MAXSYMBOLS<<1)];\
335 UBYTE tbl##_len [LZX_##tbl##_MAXSYMBOLS + LZX_LENTABLE_SAFETY]
338 UBYTE *window; /* the actual decoding window */
339 ULONG window_size; /* window size (32Kb through 2Mb) */
340 ULONG actual_size; /* window size when it was first allocated */
341 ULONG window_posn; /* current offset within the window */
342 ULONG R0, R1, R2; /* for the LRU offset system */
343 UWORD main_elements; /* number of main tree elements */
344 int header_read; /* have we started decoding at all yet? */
345 UWORD block_type; /* type of this block */
346 ULONG block_length; /* uncompressed length of this block */
347 ULONG block_remaining; /* uncompressed bytes still left to decode */
348 ULONG frames_read; /* the number of CFDATA blocks processed */
349 LONG intel_filesize; /* magic header value used for transform */
350 LONG intel_curpos; /* current offset in transform space */
351 int intel_started; /* have we seen any translatable data yet? */
353 LZX_DECLARE_TABLE(PRETREE);
354 LZX_DECLARE_TABLE(MAINTREE);
355 LZX_DECLARE_TABLE(LENGTH);
356 LZX_DECLARE_TABLE(ALIGNED);
361 #define CAB(x) (decomp_state.x)
362 #define ZIP(x) (decomp_state.methods.zip.x)
363 #define QTM(x) (decomp_state.methods.qtm.x)
364 #define LZX(x) (decomp_state.methods.lzx.x)
366 #define DECR_DATAFORMAT (1)
367 #define DECR_ILLEGALDATA (2)
368 #define DECR_NOMEMORY (3)
369 #define DECR_CHECKSUM (4)
370 #define DECR_INPUT (5)
371 #define DECR_OUTPUT (6)
373 /* CAB data blocks are <= 32768 bytes in uncompressed form. Uncompressed
374 * blocks have zero growth. MSZIP guarantees that it won't grow above
375 * uncompressed size by more than 12 bytes. LZX guarantees it won't grow
376 * more than 6144 bytes.
378 #define CAB_BLOCKMAX (32768)
379 #define CAB_INPUTMAX (CAB_BLOCKMAX+6144)
382 struct folder *current; /* current folder we're extracting from */
383 ULONG offset; /* uncompressed offset within folder */
384 UBYTE *outpos; /* (high level) start of data to use up */
385 UWORD outlen; /* (high level) amount of data to use up */
386 UWORD split; /* at which split in current folder? */
387 int (*decompress)(int, int); /* the chosen compression func */
388 UBYTE inbuf[CAB_INPUTMAX+2]; /* +2 for lzx bitbuffer overflows! */
389 UBYTE outbuf[CAB_BLOCKMAX];
398 /* MSZIP decruncher */
400 /* Dirk Stoecker wrote the ZIP decoder, based on the InfoZip deflate code */
402 /* Tables for deflate from PKZIP's appnote.txt. */
403 static const UBYTE Zipborder[] = /* Order of the bit length code lengths */
404 { 16, 17, 18, 0, 8, 7, 9, 6, 10, 5, 11, 4, 12, 3, 13, 2, 14, 1, 15};
405 static const UWORD Zipcplens[] = /* Copy lengths for literal codes 257..285 */
406 { 3, 4, 5, 6, 7, 8, 9, 10, 11, 13, 15, 17, 19, 23, 27, 31, 35, 43, 51,
407 59, 67, 83, 99, 115, 131, 163, 195, 227, 258, 0, 0};
408 static const UWORD Zipcplext[] = /* Extra bits for literal codes 257..285 */
409 { 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 2, 2, 2, 2, 3, 3, 3, 3, 4, 4, 4,
410 4, 5, 5, 5, 5, 0, 99, 99}; /* 99==invalid */
411 static const UWORD Zipcpdist[] = /* Copy offsets for distance codes 0..29 */
412 { 1, 2, 3, 4, 5, 7, 9, 13, 17, 25, 33, 49, 65, 97, 129, 193, 257, 385,
413 513, 769, 1025, 1537, 2049, 3073, 4097, 6145, 8193, 12289, 16385, 24577};
414 static const UWORD Zipcpdext[] = /* Extra bits for distance codes */
415 { 0, 0, 0, 0, 1, 1, 2, 2, 3, 3, 4, 4, 5, 5, 6, 6, 7, 7, 8, 8, 9, 9, 10,
416 10, 11, 11, 12, 12, 13, 13};
418 /* And'ing with Zipmask[n] masks the lower n bits */
419 static const UWORD Zipmask[17] = {
420 0x0000, 0x0001, 0x0003, 0x0007, 0x000f, 0x001f, 0x003f, 0x007f, 0x00ff,
421 0x01ff, 0x03ff, 0x07ff, 0x0fff, 0x1fff, 0x3fff, 0x7fff, 0xffff
424 #define ZIPNEEDBITS(n) {while(k<(n)){LONG c=*(ZIP(inpos)++);\
425 b|=((ULONG)c)<<k;k+=8;}}
426 #define ZIPDUMPBITS(n) {b>>=(n);k-=(n);}
428 static void Ziphuft_free(struct Ziphuft *t)
430 register struct Ziphuft *p, *q;
432 /* Go through linked list, freeing from the allocated (t[-1]) address. */
434 while (p != (struct Ziphuft *)NULL)
442 static LONG Ziphuft_build(ULONG *b, ULONG n, ULONG s, UWORD *d, UWORD *e,
443 struct Ziphuft **t, LONG *m)
445 ULONG a; /* counter for codes of length k */
446 ULONG el; /* length of EOB code (value 256) */
447 ULONG f; /* i repeats in table every f entries */
448 LONG g; /* maximum code length */
449 LONG h; /* table level */
450 register ULONG i; /* counter, current code */
451 register ULONG j; /* counter */
452 register LONG k; /* number of bits in current code */
453 LONG *l; /* stack of bits per table */
454 register ULONG *p; /* pointer into ZIP(c)[],ZIP(b)[],ZIP(v)[] */
455 register struct Ziphuft *q; /* points to current table */
456 struct Ziphuft r; /* table entry for structure assignment */
457 register LONG w; /* bits before this table == (l * h) */
458 ULONG *xp; /* pointer into x */
459 LONG y; /* number of dummy codes added */
460 ULONG z; /* number of entries in current table */
464 /* Generate counts for each bit length */
465 el = n > 256 ? b[256] : ZIPBMAX; /* set length of EOB code, if any */
467 for(i = 0; i < ZIPBMAX+1; ++i)
472 ZIP(c)[*p]++; p++; /* assume all entries <= ZIPBMAX */
474 if (ZIP(c)[0] == n) /* null input--all zero length codes */
476 *t = (struct Ziphuft *)NULL;
481 /* Find minimum and maximum length, bound *m by those */
482 for (j = 1; j <= ZIPBMAX; j++)
485 k = j; /* minimum code length */
488 for (i = ZIPBMAX; i; i--)
491 g = i; /* maximum code length */
495 /* Adjust last length count to fill out codes, if needed */
496 for (y = 1 << j; j < i; j++, y <<= 1)
497 if ((y -= ZIP(c)[j]) < 0)
498 return 2; /* bad input: more codes than bits */
499 if ((y -= ZIP(c)[i]) < 0)
503 /* Generate starting offsets LONGo the value table for each length */
505 p = ZIP(c) + 1; xp = ZIP(x) + 2;
507 { /* note that i == g from above */
511 /* Make a table of values in order of bit lengths */
515 ZIP(v)[ZIP(x)[j]++] = i;
519 /* Generate the Huffman codes and for each, make the table entries */
520 ZIP(x)[0] = i = 0; /* first Huffman code is zero */
521 p = ZIP(v); /* grab values in bit order */
522 h = -1; /* no tables yet--level -1 */
523 w = l[-1] = 0; /* no bits decoded yet */
524 ZIP(u)[0] = (struct Ziphuft *)NULL; /* just to keep compilers happy */
525 q = (struct Ziphuft *)NULL; /* ditto */
528 /* go through the bit lengths (k already is bits in shortest code) */
534 /* here i is the Huffman code of length k bits for value *p */
535 /* make tables up to required level */
538 w += l[h++]; /* add bits already decoded */
540 /* compute minimum size table less than or equal to *m bits */
541 z = (z = g - w) > (ULONG)*m ? (ULONG)*m : z; /* upper limit */
542 if ((f = 1 << (j = k - w)) > a + 1) /* try a k-w bit table */
543 { /* too few codes for k-w bit table */
544 f -= a + 1; /* deduct codes from patterns left */
546 while (++j < z) /* try smaller tables up to z bits */
548 if ((f <<= 1) <= *++xp)
549 break; /* enough codes to use up j bits */
550 f -= *xp; /* else deduct codes from patterns */
553 if ((ULONG)w + j > el && (ULONG)w < el)
554 j = el - w; /* make EOB code end at table */
555 z = 1 << j; /* table entries for j-bit table */
556 l[h] = j; /* set table size in stack */
558 /* allocate and link in new table */
559 if (!(q = (struct Ziphuft *) malloc((z + 1)*sizeof(struct Ziphuft))))
562 Ziphuft_free(ZIP(u)[0]);
563 return 3; /* not enough memory */
565 *t = q + 1; /* link to list for Ziphuft_free() */
566 *(t = &(q->v.t)) = (struct Ziphuft *)NULL;
567 ZIP(u)[h] = ++q; /* table starts after link */
569 /* connect to last table, if there is one */
572 ZIP(x)[h] = i; /* save pattern for backing up */
573 r.b = (UBYTE)l[h-1]; /* bits to dump before this table */
574 r.e = (UBYTE)(16 + j); /* bits in this table */
575 r.v.t = q; /* pointer to this table */
576 j = (i & ((1 << w) - 1)) >> (w - l[h-1]);
577 ZIP(u)[h-1][j] = r; /* connect to last table */
581 /* set up table entry in r */
582 r.b = (UBYTE)(k - w);
584 r.e = 99; /* out of values--invalid code */
587 r.e = (UBYTE)(*p < 256 ? 16 : 15); /* 256 is end-of-block code */
588 r.v.n = *p++; /* simple code is just the value */
592 r.e = (UBYTE)e[*p - s]; /* non-simple--look up in lists */
596 /* fill code-like entries with r */
598 for (j = i >> w; j < z; j += f)
601 /* backwards increment the k-bit code i */
602 for (j = 1 << (k - 1); i & j; j >>= 1)
606 /* backup over finished tables */
607 while ((i & ((1 << w) - 1)) != ZIP(x)[h])
608 w -= l[--h]; /* don't need to update q */
612 /* return actual size of base table */
615 /* Return true (1) if we were given an incomplete table */
616 return y != 0 && g != 1;
619 static LONG Zipinflate_codes(struct Ziphuft *tl, struct Ziphuft *td,
622 register ULONG e; /* table entry flag/number of extra bits */
623 ULONG n, d; /* length and index for copy */
624 ULONG w; /* current window position */
625 struct Ziphuft *t; /* pointer to table entry */
626 ULONG ml, md; /* masks for bl and bd bits */
627 register ULONG b; /* bit buffer */
628 register ULONG k; /* number of bits in bit buffer */
630 /* make local copies of globals */
631 b = ZIP(bb); /* initialize bit buffer */
633 w = ZIP(window_posn); /* initialize window position */
635 /* inflate the coded data */
636 ml = Zipmask[bl]; /* precompute masks for speed */
641 ZIPNEEDBITS((ULONG)bl)
642 if((e = (t = tl + ((ULONG)b & ml))->e) > 16)
650 } while ((e = (t = t->v.t + ((ULONG)b & Zipmask[e]))->e) > 16);
652 if (e == 16) /* then it's a literal */
653 CAB(outbuf)[w++] = (UBYTE)t->v.n;
654 else /* it's an EOB or a length */
656 /* exit if end of block */
660 /* get length of block to copy */
662 n = t->v.n + ((ULONG)b & Zipmask[e]);
665 /* decode distance of block to copy */
666 ZIPNEEDBITS((ULONG)bd)
667 if ((e = (t = td + ((ULONG)b & md))->e) > 16)
674 } while ((e = (t = t->v.t + ((ULONG)b & Zipmask[e]))->e) > 16);
677 d = w - t->v.n - ((ULONG)b & Zipmask[e]);
681 n -= (e = (e = ZIPWSIZE - ((d &= ZIPWSIZE-1) > w ? d : w)) > n ?n:e);
684 CAB(outbuf)[w++] = CAB(outbuf)[d++];
690 /* restore the globals from the locals */
691 ZIP(window_posn) = w; /* restore global window pointer */
692 ZIP(bb) = b; /* restore global bit buffer */
699 static LONG Zipinflate_stored(void)
700 /* "decompress" an inflated type 0 (stored) block. */
702 ULONG n; /* number of bytes in block */
703 ULONG w; /* current window position */
704 register ULONG b; /* bit buffer */
705 register ULONG k; /* number of bits in bit buffer */
707 /* make local copies of globals */
708 b = ZIP(bb); /* initialize bit buffer */
710 w = ZIP(window_posn); /* initialize window position */
712 /* go to byte boundary */
716 /* get the length and its complement */
718 n = ((ULONG)b & 0xffff);
721 if (n != (ULONG)((~b) & 0xffff))
722 return 1; /* error in compressed data */
725 /* read and output the compressed data */
729 CAB(outbuf)[w++] = (UBYTE)b;
733 /* restore the globals from the locals */
734 ZIP(window_posn) = w; /* restore global window pointer */
735 ZIP(bb) = b; /* restore global bit buffer */
740 static LONG Zipinflate_fixed(void)
742 struct Ziphuft *fixed_tl;
743 struct Ziphuft *fixed_td;
744 LONG fixed_bl, fixed_bd;
745 LONG i; /* temporary variable */
751 for(i = 0; i < 144; i++)
757 for(; i < 288; i++) /* make a complete, but wrong code set */
760 if((i = Ziphuft_build(l, 288, 257, (UWORD *) Zipcplens,
761 (UWORD *) Zipcplext, &fixed_tl, &fixed_bl)))
765 for(i = 0; i < 30; i++) /* make an incomplete code set */
768 if((i = Ziphuft_build(l, 30, 0, (UWORD *) Zipcpdist, (UWORD *) Zipcpdext,
769 &fixed_td, &fixed_bd)) > 1)
771 Ziphuft_free(fixed_tl);
775 /* decompress until an end-of-block code */
776 i = Zipinflate_codes(fixed_tl, fixed_td, fixed_bl, fixed_bd);
778 Ziphuft_free(fixed_td);
779 Ziphuft_free(fixed_tl);
783 static LONG Zipinflate_dynamic(void)
784 /* decompress an inflated type 2 (dynamic Huffman codes) block. */
786 LONG i; /* temporary variables */
789 ULONG l; /* last length */
790 ULONG m; /* mask for bit lengths table */
791 ULONG n; /* number of lengths to get */
792 struct Ziphuft *tl; /* literal/length code table */
793 struct Ziphuft *td; /* distance code table */
794 LONG bl; /* lookup bits for tl */
795 LONG bd; /* lookup bits for td */
796 ULONG nb; /* number of bit length codes */
797 ULONG nl; /* number of literal/length codes */
798 ULONG nd; /* number of distance codes */
799 register ULONG b; /* bit buffer */
800 register ULONG k; /* number of bits in bit buffer */
802 /* make local bit buffer */
807 /* read in table lengths */
809 nl = 257 + ((ULONG)b & 0x1f); /* number of literal/length codes */
812 nd = 1 + ((ULONG)b & 0x1f); /* number of distance codes */
815 nb = 4 + ((ULONG)b & 0xf); /* number of bit length codes */
817 if(nl > 288 || nd > 32)
818 return 1; /* bad lengths */
820 /* read in bit-length-code lengths */
821 for(j = 0; j < nb; j++)
824 ll[Zipborder[j]] = (ULONG)b & 7;
828 ll[Zipborder[j]] = 0;
830 /* build decoding table for trees--single level, 7 bit lookup */
832 if((i = Ziphuft_build(ll, 19, 19, NULL, NULL, &tl, &bl)) != 0)
836 return i; /* incomplete code set */
839 /* read in literal and distance code lengths */
845 ZIPNEEDBITS((ULONG)bl)
846 j = (td = tl + ((ULONG)b & m))->b;
849 if (j < 16) /* length of code in bits (0..15) */
850 ll[i++] = l = j; /* save last length in l */
851 else if (j == 16) /* repeat last length 3 to 6 times */
854 j = 3 + ((ULONG)b & 3);
861 else if (j == 17) /* 3 to 10 zero length codes */
864 j = 3 + ((ULONG)b & 7);
866 if ((ULONG)i + j > n)
872 else /* j == 18: 11 to 138 zero length codes */
875 j = 11 + ((ULONG)b & 0x7f);
877 if ((ULONG)i + j > n)
885 /* free decoding table for trees */
888 /* restore the global bit buffer */
892 /* build the decoding tables for literal/length and distance codes */
894 if((i = Ziphuft_build(ll, nl, 257, (UWORD *) Zipcplens, (UWORD *) Zipcplext, &tl, &bl)) != 0)
898 return i; /* incomplete code set */
901 Ziphuft_build(ll + nl, nd, 0, (UWORD *) Zipcpdist, (UWORD *) Zipcpdext, &td, &bd);
903 /* decompress until an end-of-block code */
904 if(Zipinflate_codes(tl, td, bl, bd))
907 /* free the decoding tables, return */
913 static LONG Zipinflate_block(LONG *e) /* e == last block flag */
914 { /* decompress an inflated block */
915 ULONG t; /* block type */
916 register ULONG b; /* bit buffer */
917 register ULONG k; /* number of bits in bit buffer */
919 /* make local bit buffer */
923 /* read in last block bit */
928 /* read in block type */
933 /* restore the global bit buffer */
937 /* inflate that block type */
939 return Zipinflate_dynamic();
941 return Zipinflate_stored();
943 return Zipinflate_fixed();
948 static int ZIPdecompress(int inlen, int outlen)
950 LONG e; /* last block flag */
952 ZIP(inpos) = CAB(inbuf);
953 ZIP(bb) = ZIP(bk) = ZIP(window_posn) = 0;
954 if(outlen > ZIPWSIZE)
955 return DECR_DATAFORMAT;
957 /* CK = Chris Kirmse, official Microsoft purloiner */
958 if(ZIP(inpos)[0] != 0x43 || ZIP(inpos)[1] != 0x4B)
959 return DECR_ILLEGALDATA;
964 if(Zipinflate_block(&e))
965 return DECR_ILLEGALDATA;
972 /* Quantum decruncher */
974 /* This decruncher was researched and implemented by Matthew Russoto. */
975 /* It has since been tidied up by Stuart Caie */
977 static UBYTE q_length_base[27], q_length_extra[27], q_extra_bits[42];
978 static ULONG q_position_base[42];
980 /* Initialise a model which decodes symbols from [s] to [s]+[n]-1 */
981 static void QTMinitmodel(struct QTMmodel *m, struct QTMmodelsym *sym, int n, int s) {
986 memset(m->tabloc, 0xFF, sizeof(m->tabloc)); /* clear out look-up table */
987 for (i = 0; i < n; i++) {
988 m->tabloc[i+s] = i; /* set up a look-up entry for symbol */
989 m->syms[i].sym = i+s; /* actual symbol */
990 m->syms[i].cumfreq = n-i; /* current frequency of that symbol */
992 m->syms[n].cumfreq = 0;
995 static int QTMinit(int window, int level) {
996 int wndsize = 1 << window, msz = window * 2, i;
999 /* QTM supports window sizes of 2^10 (1Kb) through 2^21 (2Mb) */
1000 /* if a previously allocated window is big enough, keep it */
1001 if (window < 10 || window > 21) return DECR_DATAFORMAT;
1002 if (QTM(actual_size) < (ULONG)wndsize) {
1003 if (QTM(window)) free(QTM(window));
1007 if (!(QTM(window) = malloc(wndsize))) return DECR_NOMEMORY;
1008 QTM(actual_size) = wndsize;
1010 QTM(window_size) = wndsize;
1011 QTM(window_posn) = 0;
1013 /* initialise static slot/extrabits tables */
1014 for (i = 0, j = 0; i < 27; i++) {
1015 q_length_extra[i] = (i == 26) ? 0 : (i < 2 ? 0 : i - 2) >> 2;
1016 q_length_base[i] = j; j += 1 << ((i == 26) ? 5 : q_length_extra[i]);
1018 for (i = 0, j = 0; i < 42; i++) {
1019 q_extra_bits[i] = (i < 2 ? 0 : i-2) >> 1;
1020 q_position_base[i] = j; j += 1 << q_extra_bits[i];
1023 /* initialise arithmetic coding models */
1025 QTMinitmodel(&QTM(model7), &QTM(m7sym)[0], 7, 0);
1027 QTMinitmodel(&QTM(model00), &QTM(m00sym)[0], 0x40, 0x00);
1028 QTMinitmodel(&QTM(model40), &QTM(m40sym)[0], 0x40, 0x40);
1029 QTMinitmodel(&QTM(model80), &QTM(m80sym)[0], 0x40, 0x80);
1030 QTMinitmodel(&QTM(modelC0), &QTM(mC0sym)[0], 0x40, 0xC0);
1032 /* model 4 depends on table size, ranges from 20 to 24 */
1033 QTMinitmodel(&QTM(model4), &QTM(m4sym)[0], (msz < 24) ? msz : 24, 0);
1034 /* model 5 depends on table size, ranges from 20 to 36 */
1035 QTMinitmodel(&QTM(model5), &QTM(m5sym)[0], (msz < 36) ? msz : 36, 0);
1036 /* model 6pos depends on table size, ranges from 20 to 42 */
1037 QTMinitmodel(&QTM(model6pos), &QTM(m6psym)[0], msz, 0);
1038 QTMinitmodel(&QTM(model6len), &QTM(m6lsym)[0], 27, 0);
1044 static void QTMupdatemodel(struct QTMmodel *model, int sym) {
1045 struct QTMmodelsym temp;
1048 for (i = 0; i < sym; i++) model->syms[i].cumfreq += 8;
1050 if (model->syms[0].cumfreq > 3800) {
1051 if (--model->shiftsleft) {
1052 for (i = model->entries - 1; i >= 0; i--) {
1053 /* -1, not -2; the 0 entry saves this */
1054 model->syms[i].cumfreq >>= 1;
1055 if (model->syms[i].cumfreq <= model->syms[i+1].cumfreq) {
1056 model->syms[i].cumfreq = model->syms[i+1].cumfreq + 1;
1061 model->shiftsleft = 50;
1062 for (i = 0; i < model->entries ; i++) {
1063 /* no -1, want to include the 0 entry */
1064 /* this converts cumfreqs into frequencies, then shifts right */
1065 model->syms[i].cumfreq -= model->syms[i+1].cumfreq;
1066 model->syms[i].cumfreq++; /* avoid losing things entirely */
1067 model->syms[i].cumfreq >>= 1;
1070 /* now sort by frequencies, decreasing order -- this must be an
1071 * inplace selection sort, or a sort with the same (in)stability
1074 for (i = 0; i < model->entries - 1; i++) {
1075 for (j = i + 1; j < model->entries; j++) {
1076 if (model->syms[i].cumfreq < model->syms[j].cumfreq) {
1077 temp = model->syms[i];
1078 model->syms[i] = model->syms[j];
1079 model->syms[j] = temp;
1084 /* then convert frequencies back to cumfreq */
1085 for (i = model->entries - 1; i >= 0; i--) {
1086 model->syms[i].cumfreq += model->syms[i+1].cumfreq;
1088 /* then update the other part of the table */
1089 for (i = 0; i < model->entries; i++) {
1090 model->tabloc[model->syms[i].sym] = i;
1096 /* Bitstream reading macros (Quantum / normal byte order)
1098 * Q_INIT_BITSTREAM should be used first to set up the system
1099 * Q_READ_BITS(var,n) takes N bits from the buffer and puts them in var.
1100 * unlike LZX, this can loop several times to get the
1101 * requisite number of bits.
1102 * Q_FILL_BUFFER adds more data to the bit buffer, if there is room
1103 * for another 16 bits.
1104 * Q_PEEK_BITS(n) extracts (without removing) N bits from the bit
1106 * Q_REMOVE_BITS(n) removes N bits from the bit buffer
1108 * These bit access routines work by using the area beyond the MSB and the
1109 * LSB as a free source of zeroes. This avoids having to mask any bits.
1110 * So we have to know the bit width of the bitbuffer variable. This is
1111 * defined as ULONG_BITS.
1113 * ULONG_BITS should be at least 16 bits. Unlike LZX's Huffman decoding,
1114 * Quantum's arithmetic decoding only needs 1 bit at a time, it doesn't
1115 * need an assured number. Retrieving larger bitstrings can be done with
1116 * multiple reads and fills of the bitbuffer. The code should work fine
1117 * for machines where ULONG >= 32 bits.
1119 * Also note that Quantum reads bytes in normal order; LZX is in
1120 * little-endian order.
1123 #define Q_INIT_BITSTREAM do { bitsleft = 0; bitbuf = 0; } while (0)
1125 #define Q_FILL_BUFFER do { \
1126 if (bitsleft <= (int)(ULONG_BITS - 16)) { \
1127 bitbuf |= ((inpos[0]<<8)|inpos[1]) << (ULONG_BITS-16 - bitsleft); \
1128 bitsleft += 16; inpos += 2; \
1132 #define Q_PEEK_BITS(n) (bitbuf >> (ULONG_BITS - (n)))
1133 #define Q_REMOVE_BITS(n) ((bitbuf <<= (n)), (bitsleft -= (n)))
1135 #define Q_READ_BITS(v,n) do { \
1137 for (bitsneed = (n); bitsneed; bitsneed -= bitrun) { \
1139 bitrun = (bitsneed > bitsleft) ? bitsleft : bitsneed; \
1140 (v) = ((v) << bitrun) | Q_PEEK_BITS(bitrun); \
1141 Q_REMOVE_BITS(bitrun); \
1145 #define Q_MENTRIES(model) (QTM(model).entries)
1146 #define Q_MSYM(model,symidx) (QTM(model).syms[(symidx)].sym)
1147 #define Q_MSYMFREQ(model,symidx) (QTM(model).syms[(symidx)].cumfreq)
1149 /* GET_SYMBOL(model, var) fetches the next symbol from the stated model
1150 * and puts it in var. it may need to read the bitstream to do this.
1152 #define GET_SYMBOL(m, var) do { \
1153 range = ((H - L) & 0xFFFF) + 1; \
1154 symf = ((((C - L + 1) * Q_MSYMFREQ(m,0)) - 1) / range) & 0xFFFF; \
1156 for (i=1; i < Q_MENTRIES(m); i++) { \
1157 if (Q_MSYMFREQ(m,i) <= symf) break; \
1159 (var) = Q_MSYM(m,i-1); \
1161 range = (H - L) + 1; \
1162 H = L + ((Q_MSYMFREQ(m,i-1) * range) / Q_MSYMFREQ(m,0)) - 1; \
1163 L = L + ((Q_MSYMFREQ(m,i) * range) / Q_MSYMFREQ(m,0)); \
1165 if ((L & 0x8000) != (H & 0x8000)) { \
1166 if ((L & 0x4000) && !(H & 0x4000)) { \
1167 /* underflow case */ \
1168 C ^= 0x4000; L &= 0x3FFF; H |= 0x4000; \
1172 L <<= 1; H = (H << 1) | 1; \
1174 C = (C << 1) | Q_PEEK_BITS(1); \
1178 QTMupdatemodel(&(QTM(m)), i); \
1182 static int QTMdecompress(int inlen, int outlen) {
1183 UBYTE *inpos = CAB(inbuf);
1184 UBYTE *window = QTM(window);
1185 UBYTE *runsrc, *rundest;
1187 ULONG window_posn = QTM(window_posn);
1188 ULONG window_size = QTM(window_size);
1190 /* used by bitstream macros */
1191 register int bitsleft, bitrun, bitsneed;
1192 register ULONG bitbuf;
1194 /* used by GET_SYMBOL */
1199 int extra, togo = outlen, match_length = 0; /* Prevent: ... might be used uninitialized in this function */
1201 UBYTE selector, sym;
1202 ULONG match_offset = 0; /* Prevent: ... might be used uninitialized in this function */
1204 UWORD H = 0xFFFF, L = 0, C;
1206 /* read initial value of C */
1210 /* apply 2^x-1 mask */
1211 window_posn &= window_size - 1;
1212 /* runs can't straddle the window wraparound */
1213 if ((window_posn + togo) > window_size) {
1214 D(("straddled run\n"))
1215 return DECR_DATAFORMAT;
1219 GET_SYMBOL(model7, selector);
1222 GET_SYMBOL(model00, sym); window[window_posn++] = sym; togo--;
1225 GET_SYMBOL(model40, sym); window[window_posn++] = sym; togo--;
1228 GET_SYMBOL(model80, sym); window[window_posn++] = sym; togo--;
1231 GET_SYMBOL(modelC0, sym); window[window_posn++] = sym; togo--;
1235 /* selector 4 = fixed length of 3 */
1236 GET_SYMBOL(model4, sym);
1237 Q_READ_BITS(extra, q_extra_bits[sym]);
1238 match_offset = q_position_base[sym] + extra + 1;
1243 /* selector 5 = fixed length of 4 */
1244 GET_SYMBOL(model5, sym);
1245 Q_READ_BITS(extra, q_extra_bits[sym]);
1246 match_offset = q_position_base[sym] + extra + 1;
1251 /* selector 6 = variable length */
1252 GET_SYMBOL(model6len, sym);
1253 Q_READ_BITS(extra, q_length_extra[sym]);
1254 match_length = q_length_base[sym] + extra + 5;
1255 GET_SYMBOL(model6pos, sym);
1256 Q_READ_BITS(extra, q_extra_bits[sym]);
1257 match_offset = q_position_base[sym] + extra + 1;
1261 D(("Selector is bogus\n"))
1262 return DECR_ILLEGALDATA;
1265 /* if this is a match */
1266 if (selector >= 4) {
1267 rundest = window + window_posn;
1268 togo -= match_length;
1270 /* copy any wrapped around source data */
1271 if (window_posn >= match_offset) {
1273 runsrc = rundest - match_offset;
1275 runsrc = rundest + (window_size - match_offset);
1276 copy_length = match_offset - window_posn;
1277 if (copy_length < match_length) {
1278 match_length -= copy_length;
1279 window_posn += copy_length;
1280 while (copy_length-- > 0) *rundest++ = *runsrc++;
1284 window_posn += match_length;
1286 /* copy match data - no worries about destination wraps */
1287 while (match_length-- > 0) *rundest++ = *runsrc++;
1289 } /* while (togo > 0) */
1292 D(("Frame overflow, this_run = %d\n", togo))
1293 return DECR_ILLEGALDATA;
1296 memcpy(CAB(outbuf), window + ((!window_posn) ? window_size : window_posn) -
1299 QTM(window_posn) = window_posn;
1305 /* LZX decruncher */
1307 /* Microsoft's LZX document and their implementation of the
1308 * com.ms.util.cab Java package do not concur.
1310 * In the LZX document, there is a table showing the correlation between
1311 * window size and the number of position slots. It states that the 1MB
1312 * window = 40 slots and the 2MB window = 42 slots. In the implementation,
1313 * 1MB = 42 slots, 2MB = 50 slots. The actual calculation is 'find the
1314 * first slot whose position base is equal to or more than the required
1315 * window size'. This would explain why other tables in the document refer
1316 * to 50 slots rather than 42.
1318 * The constant NUM_PRIMARY_LENGTHS used in the decompression pseudocode
1319 * is not defined in the specification.
1321 * The LZX document does not state the uncompressed block has an
1322 * uncompressed length field. Where does this length field come from, so
1323 * we can know how large the block is? The implementation has it as the 24
1324 * bits following after the 3 blocktype bits, before the alignment
1327 * The LZX document states that aligned offset blocks have their aligned
1328 * offset huffman tree AFTER the main and length trees. The implementation
1329 * suggests that the aligned offset tree is BEFORE the main and length
1332 * The LZX document decoding algorithm states that, in an aligned offset
1333 * block, if an extra_bits value is 1, 2 or 3, then that number of bits
1334 * should be read and the result added to the match offset. This is
1335 * correct for 1 and 2, but not 3, where just a huffman symbol (using the
1336 * aligned tree) should be read.
1338 * Regarding the E8 preprocessing, the LZX document states 'No translation
1339 * may be performed on the last 6 bytes of the input block'. This is
1340 * correct. However, the pseudocode provided checks for the *E8 leader*
1341 * up to the last 6 bytes. If the leader appears between -10 and -7 bytes
1342 * from the end, this would cause the next four bytes to be modified, at
1343 * least one of which would be in the last 6 bytes, which is not allowed
1344 * according to the spec.
1346 * The specification states that the huffman trees must always contain at
1347 * least one element. However, many CAB files contain blocks where the
1348 * length tree is completely empty (because there are no matches), and
1349 * this is expected to succeed.
1353 /* LZX uses what it calls 'position slots' to represent match offsets.
1354 * What this means is that a small 'position slot' number and a small
1355 * offset from that slot are encoded instead of one large offset for
1357 * - lzx_position_base is an index to the position slot bases
1358 * - lzx_extra_bits states how many bits of offset-from-base data is needed.
1360 static ULONG lzx_position_base[51];
1361 static UBYTE extra_bits[51];
1363 static int LZXinit(int window) {
1364 ULONG wndsize = 1 << window;
1365 int i, j, posn_slots;
1367 /* LZX supports window sizes of 2^15 (32Kb) through 2^21 (2Mb) */
1368 /* if a previously allocated window is big enough, keep it */
1369 if (window < 15 || window > 21) return DECR_DATAFORMAT;
1370 if (LZX(actual_size) < wndsize) {
1371 if (LZX(window)) free(LZX(window));
1375 if (!(LZX(window) = malloc(wndsize))) return DECR_NOMEMORY;
1376 LZX(actual_size) = wndsize;
1378 LZX(window_size) = wndsize;
1380 /* initialise static tables */
1381 for (i=0, j=0; i <= 50; i += 2) {
1382 extra_bits[i] = extra_bits[i+1] = j; /* 0,0,0,0,1,1,2,2,3,3... */
1383 if ((i != 0) && (j < 17)) j++; /* 0,0,1,2,3,4...15,16,17,17,17,17... */
1385 for (i=0, j=0; i <= 50; i++) {
1386 lzx_position_base[i] = j; /* 0,1,2,3,4,6,8,12,16,24,32,... */
1387 j += 1 << extra_bits[i]; /* 1,1,1,1,2,2,4,4,8,8,16,16,32,32,... */
1390 /* calculate required position slots */
1391 if (window == 20) posn_slots = 42;
1392 else if (window == 21) posn_slots = 50;
1393 else posn_slots = window << 1;
1395 /*posn_slots=i=0; while (i < wndsize) i += 1 << extra_bits[posn_slots++]; */
1398 LZX(R0) = LZX(R1) = LZX(R2) = 1;
1399 LZX(main_elements) = LZX_NUM_CHARS + (posn_slots << 3);
1400 LZX(header_read) = 0;
1401 LZX(frames_read) = 0;
1402 LZX(block_remaining) = 0;
1403 LZX(block_type) = LZX_BLOCKTYPE_INVALID;
1404 LZX(intel_curpos) = 0;
1405 LZX(intel_started) = 0;
1406 LZX(window_posn) = 0;
1408 /* initialise tables to 0 (because deltas will be applied to them) */
1409 for (i = 0; i < LZX_MAINTREE_MAXSYMBOLS; i++) LZX(MAINTREE_len)[i] = 0;
1410 for (i = 0; i < LZX_LENGTH_MAXSYMBOLS; i++) LZX(LENGTH_len)[i] = 0;
1415 /* Bitstream reading macros (LZX / intel little-endian byte order)
1417 * INIT_BITSTREAM should be used first to set up the system
1418 * READ_BITS(var,n) takes N bits from the buffer and puts them in var
1420 * ENSURE_BITS(n) ensures there are at least N bits in the bit buffer.
1421 * it can guarantee up to 17 bits (i.e. it can read in
1422 * 16 new bits when there is down to 1 bit in the buffer,
1423 * and it can read 32 bits when there are 0 bits in the
1425 * PEEK_BITS(n) extracts (without removing) N bits from the bit buffer
1426 * REMOVE_BITS(n) removes N bits from the bit buffer
1428 * These bit access routines work by using the area beyond the MSB and the
1429 * LSB as a free source of zeroes. This avoids having to mask any bits.
1430 * So we have to know the bit width of the bitbuffer variable.
1433 #define INIT_BITSTREAM do { bitsleft = 0; bitbuf = 0; } while (0)
1435 /* Quantum reads bytes in normal order; LZX is little-endian order */
1436 #define ENSURE_BITS(n) \
1437 while (bitsleft < (n)) { \
1438 bitbuf |= ((inpos[1]<<8)|inpos[0]) << (ULONG_BITS-16 - bitsleft); \
1439 bitsleft += 16; inpos+=2; \
1442 #define PEEK_BITS(n) (bitbuf >> (ULONG_BITS - (n)))
1443 #define REMOVE_BITS(n) ((bitbuf <<= (n)), (bitsleft -= (n)))
1445 #define READ_BITS(v,n) do { \
1448 (v) = PEEK_BITS(n); \
1456 /* Huffman macros */
1458 #define TABLEBITS(tbl) (LZX_##tbl##_TABLEBITS)
1459 #define MAXSYMBOLS(tbl) (LZX_##tbl##_MAXSYMBOLS)
1460 #define SYMTABLE(tbl) (LZX(tbl##_table))
1461 #define LENTABLE(tbl) (LZX(tbl##_len))
1463 /* BUILD_TABLE(tablename) builds a huffman lookup table from code lengths.
1464 * In reality, it just calls make_decode_table() with the appropriate
1465 * values - they're all fixed by some #defines anyway, so there's no point
1466 * writing each call out in full by hand.
1468 #define BUILD_TABLE(tbl) \
1469 if (make_decode_table( \
1470 MAXSYMBOLS(tbl), TABLEBITS(tbl), LENTABLE(tbl), SYMTABLE(tbl) \
1471 )) { return DECR_ILLEGALDATA; }
1474 /* READ_HUFFSYM(tablename, var) decodes one huffman symbol from the
1475 * bitstream using the stated table and puts it in var.
1477 #define READ_HUFFSYM(tbl,var) do { \
1479 hufftbl = SYMTABLE(tbl); \
1480 if ((i = hufftbl[PEEK_BITS(TABLEBITS(tbl))]) >= MAXSYMBOLS(tbl)) { \
1481 j = 1 << (ULONG_BITS - TABLEBITS(tbl)); \
1483 j >>= 1; i <<= 1; i |= (bitbuf & j) ? 1 : 0; \
1484 if (!j) { return DECR_ILLEGALDATA; } \
1485 } while ((i = hufftbl[i]) >= MAXSYMBOLS(tbl)); \
1487 j = LENTABLE(tbl)[(var) = i]; \
1492 /* READ_LENGTHS(tablename, first, last) reads in code lengths for symbols
1493 * first to last in the given table. The code lengths are stored in their
1494 * own special LZX way.
1496 #define READ_LENGTHS(tbl,first,last) do { \
1497 lb.bb = bitbuf; lb.bl = bitsleft; lb.ip = inpos; \
1498 if (lzx_read_lens(LENTABLE(tbl),(first),(last),&lb)) { \
1499 return DECR_ILLEGALDATA; \
1501 bitbuf = lb.bb; bitsleft = lb.bl; inpos = lb.ip; \
1505 /* make_decode_table(nsyms, nbits, length[], table[])
1507 * This function was coded by David Tritscher. It builds a fast huffman
1508 * decoding table out of just a canonical huffman code lengths table.
1510 * nsyms = total number of symbols in this huffman tree.
1511 * nbits = any symbols with a code length of nbits or less can be decoded
1512 * in one lookup of the table.
1513 * length = A table to get code lengths from [0 to syms-1]
1514 * table = The table to fill up with decoded symbols and pointers.
1516 * Returns 0 for OK or 1 for error
1519 static int make_decode_table(ULONG nsyms, ULONG nbits, UBYTE *length, UWORD *table) {
1521 register ULONG leaf;
1522 register UBYTE bit_num = 1;
1524 ULONG pos = 0; /* the current position in the decode table */
1525 ULONG table_mask = 1 << nbits;
1526 ULONG bit_mask = table_mask >> 1; /* don't do 0 length codes */
1527 ULONG next_symbol = bit_mask; /* base of allocation for long codes */
1529 /* fill entries for codes short enough for a direct mapping */
1530 while (bit_num <= nbits) {
1531 for (sym = 0; sym < nsyms; sym++) {
1532 if (length[sym] == bit_num) {
1535 if((pos += bit_mask) > table_mask) return 1; /* table overrun */
1537 /* fill all possible lookups of this symbol with the symbol itself */
1539 while (fill-- > 0) table[leaf++] = sym;
1546 /* if there are any codes longer than nbits */
1547 if (pos != table_mask) {
1548 /* clear the remainder of the table */
1549 for (sym = pos; sym < table_mask; sym++) table[sym] = 0;
1551 /* give ourselves room for codes to grow by up to 16 more bits */
1556 while (bit_num <= 16) {
1557 for (sym = 0; sym < nsyms; sym++) {
1558 if (length[sym] == bit_num) {
1560 for (fill = 0; fill < bit_num - nbits; fill++) {
1561 /* if this path hasn't been taken yet, 'allocate' two entries */
1562 if (table[leaf] == 0) {
1563 table[(next_symbol << 1)] = 0;
1564 table[(next_symbol << 1) + 1] = 0;
1565 table[leaf] = next_symbol++;
1567 /* follow the path and select either left or right for next bit */
1568 leaf = table[leaf] << 1;
1569 if ((pos >> (15-fill)) & 1) leaf++;
1573 if ((pos += bit_mask) > table_mask) return 1; /* table overflow */
1582 if (pos == table_mask) return 0;
1584 /* either erroneous table, or all elements are 0 - let's find out. */
1585 for (sym = 0; sym < nsyms; sym++) if (length[sym]) return 1;
1595 static int lzx_read_lens(UBYTE *lens, ULONG first, ULONG last, struct lzx_bits *lb) {
1599 register ULONG bitbuf = lb->bb;
1600 register int bitsleft = lb->bl;
1601 UBYTE *inpos = lb->ip;
1604 for (x = 0; x < 20; x++) {
1606 LENTABLE(PRETREE)[x] = y;
1608 BUILD_TABLE(PRETREE);
1610 for (x = first; x < last; ) {
1611 READ_HUFFSYM(PRETREE, z);
1613 READ_BITS(y, 4); y += 4;
1614 while (y--) lens[x++] = 0;
1617 READ_BITS(y, 5); y += 20;
1618 while (y--) lens[x++] = 0;
1621 READ_BITS(y, 1); y += 4;
1622 READ_HUFFSYM(PRETREE, z);
1623 z = lens[x] - z; if (z < 0) z += 17;
1624 while (y--) lens[x++] = z;
1627 z = lens[x] - z; if (z < 0) z += 17;
1638 static int LZXdecompress(int inlen, int outlen) {
1639 UBYTE *inpos = CAB(inbuf);
1640 UBYTE *endinp = inpos + inlen;
1641 UBYTE *window = LZX(window);
1642 UBYTE *runsrc, *rundest;
1643 UWORD *hufftbl; /* used in READ_HUFFSYM macro as chosen decoding table */
1645 ULONG window_posn = LZX(window_posn);
1646 ULONG window_size = LZX(window_size);
1651 register ULONG bitbuf;
1652 register int bitsleft;
1653 ULONG match_offset, i,j,k; /* ijk used in READ_HUFFSYM macro */
1654 struct lzx_bits lb; /* used in READ_LENGTHS macro */
1656 int togo = outlen, this_run, main_element, aligned_bits;
1657 int match_length, copy_length, length_footer, extra, verbatim_bits;
1661 /* read header if necessary */
1662 if (!LZX(header_read)) {
1664 READ_BITS(k, 1); if (k) { READ_BITS(i,16); READ_BITS(j,16); }
1665 LZX(intel_filesize) = (i << 16) | j; /* or 0 if not encoded */
1666 LZX(header_read) = 1;
1669 /* main decoding loop */
1671 /* last block finished, new block expected */
1672 if (LZX(block_remaining) == 0) {
1673 if (LZX(block_type) == LZX_BLOCKTYPE_UNCOMPRESSED) {
1674 if (LZX(block_length) & 1) inpos++; /* realign bitstream to word */
1678 READ_BITS(LZX(block_type), 3);
1681 LZX(block_remaining) = LZX(block_length) = (i << 8) | j;
1683 switch (LZX(block_type)) {
1684 case LZX_BLOCKTYPE_ALIGNED:
1685 for (i = 0; i < 8; i++) { READ_BITS(j, 3); LENTABLE(ALIGNED)[i] = j; }
1686 BUILD_TABLE(ALIGNED);
1687 /* rest of aligned header is same as verbatim */
1689 case LZX_BLOCKTYPE_VERBATIM:
1690 READ_LENGTHS(MAINTREE, 0, 256);
1691 READ_LENGTHS(MAINTREE, 256, LZX(main_elements));
1692 BUILD_TABLE(MAINTREE);
1693 if (LENTABLE(MAINTREE)[0xE8] != 0) LZX(intel_started) = 1;
1695 READ_LENGTHS(LENGTH, 0, LZX_NUM_SECONDARY_LENGTHS);
1696 BUILD_TABLE(LENGTH);
1699 case LZX_BLOCKTYPE_UNCOMPRESSED:
1700 LZX(intel_started) = 1; /* because we can't assume otherwise */
1701 ENSURE_BITS(16); /* get up to 16 pad bits into the buffer */
1702 if (bitsleft > 16) inpos -= 2; /* and align the bitstream! */
1703 R0 = inpos[0]|(inpos[1]<<8)|(inpos[2]<<16)|(inpos[3]<<24);inpos+=4;
1704 R1 = inpos[0]|(inpos[1]<<8)|(inpos[2]<<16)|(inpos[3]<<24);inpos+=4;
1705 R2 = inpos[0]|(inpos[1]<<8)|(inpos[2]<<16)|(inpos[3]<<24);inpos+=4;
1709 return DECR_ILLEGALDATA;
1713 /* buffer exhaustion check */
1714 if (inpos > endinp) {
1715 /* it's possible to have a file where the next run is less than
1716 * 16 bits in size. In this case, the READ_HUFFSYM() macro used
1717 * in building the tables will exhaust the buffer, so we should
1718 * allow for this, but not allow those accidentally read bits to
1719 * be used (so we check that there are at least 16 bits
1720 * remaining - in this boundary case they aren't really part of
1721 * the compressed data)
1723 if (inpos > (endinp+2) || bitsleft < 16) return DECR_ILLEGALDATA;
1726 while ((this_run = LZX(block_remaining)) > 0 && togo > 0) {
1727 if (this_run > togo) this_run = togo;
1729 LZX(block_remaining) -= this_run;
1731 /* apply 2^x-1 mask */
1732 window_posn &= window_size - 1;
1733 /* runs can't straddle the window wraparound */
1734 if ((window_posn + this_run) > window_size)
1735 return DECR_DATAFORMAT;
1737 switch (LZX(block_type)) {
1739 case LZX_BLOCKTYPE_VERBATIM:
1740 while (this_run > 0) {
1741 READ_HUFFSYM(MAINTREE, main_element);
1743 if (main_element < LZX_NUM_CHARS) {
1744 /* literal: 0 to LZX_NUM_CHARS-1 */
1745 window[window_posn++] = main_element;
1749 /* match: LZX_NUM_CHARS + ((slot<<3) | length_header (3 bits)) */
1750 main_element -= LZX_NUM_CHARS;
1752 match_length = main_element & LZX_NUM_PRIMARY_LENGTHS;
1753 if (match_length == LZX_NUM_PRIMARY_LENGTHS) {
1754 READ_HUFFSYM(LENGTH, length_footer);
1755 match_length += length_footer;
1757 match_length += LZX_MIN_MATCH;
1759 match_offset = main_element >> 3;
1761 if (match_offset > 2) {
1762 /* not repeated offset */
1763 if (match_offset != 3) {
1764 extra = extra_bits[match_offset];
1765 READ_BITS(verbatim_bits, extra);
1766 match_offset = lzx_position_base[match_offset]
1767 - 2 + verbatim_bits;
1773 /* update repeated offset LRU queue */
1774 R2 = R1; R1 = R0; R0 = match_offset;
1776 else if (match_offset == 0) {
1779 else if (match_offset == 1) {
1781 R1 = R0; R0 = match_offset;
1783 else /* match_offset == 2 */ {
1785 R2 = R0; R0 = match_offset;
1788 rundest = window + window_posn;
1789 this_run -= match_length;
1791 /* copy any wrapped around source data */
1792 if (window_posn >= match_offset) {
1794 runsrc = rundest - match_offset;
1796 runsrc = rundest + (window_size - match_offset);
1797 copy_length = match_offset - window_posn;
1798 if (copy_length < match_length) {
1799 match_length -= copy_length;
1800 window_posn += copy_length;
1801 while (copy_length-- > 0) *rundest++ = *runsrc++;
1805 window_posn += match_length;
1807 /* copy match data - no worries about destination wraps */
1808 while (match_length-- > 0) *rundest++ = *runsrc++;
1813 case LZX_BLOCKTYPE_ALIGNED:
1814 while (this_run > 0) {
1815 READ_HUFFSYM(MAINTREE, main_element);
1817 if (main_element < LZX_NUM_CHARS) {
1818 /* literal: 0 to LZX_NUM_CHARS-1 */
1819 window[window_posn++] = main_element;
1823 /* match: LZX_NUM_CHARS + ((slot<<3) | length_header (3 bits)) */
1824 main_element -= LZX_NUM_CHARS;
1826 match_length = main_element & LZX_NUM_PRIMARY_LENGTHS;
1827 if (match_length == LZX_NUM_PRIMARY_LENGTHS) {
1828 READ_HUFFSYM(LENGTH, length_footer);
1829 match_length += length_footer;
1831 match_length += LZX_MIN_MATCH;
1833 match_offset = main_element >> 3;
1835 if (match_offset > 2) {
1836 /* not repeated offset */
1837 extra = extra_bits[match_offset];
1838 match_offset = lzx_position_base[match_offset] - 2;
1840 /* verbatim and aligned bits */
1842 READ_BITS(verbatim_bits, extra);
1843 match_offset += (verbatim_bits << 3);
1844 READ_HUFFSYM(ALIGNED, aligned_bits);
1845 match_offset += aligned_bits;
1847 else if (extra == 3) {
1848 /* aligned bits only */
1849 READ_HUFFSYM(ALIGNED, aligned_bits);
1850 match_offset += aligned_bits;
1852 else if (extra > 0) { /* extra==1, extra==2 */
1853 /* verbatim bits only */
1854 READ_BITS(verbatim_bits, extra);
1855 match_offset += verbatim_bits;
1857 else /* extra == 0 */ {
1862 /* update repeated offset LRU queue */
1863 R2 = R1; R1 = R0; R0 = match_offset;
1865 else if (match_offset == 0) {
1868 else if (match_offset == 1) {
1870 R1 = R0; R0 = match_offset;
1872 else /* match_offset == 2 */ {
1874 R2 = R0; R0 = match_offset;
1877 rundest = window + window_posn;
1878 this_run -= match_length;
1880 /* copy any wrapped around source data */
1881 if (window_posn >= match_offset) {
1883 runsrc = rundest - match_offset;
1885 runsrc = rundest + (window_size - match_offset);
1886 copy_length = match_offset - window_posn;
1887 if (copy_length < match_length) {
1888 match_length -= copy_length;
1889 window_posn += copy_length;
1890 while (copy_length-- > 0) *rundest++ = *runsrc++;
1894 window_posn += match_length;
1896 /* copy match data - no worries about destination wraps */
1897 while (match_length-- > 0) *rundest++ = *runsrc++;
1902 case LZX_BLOCKTYPE_UNCOMPRESSED:
1903 if ((inpos + this_run) > endinp) return DECR_ILLEGALDATA;
1904 memcpy(window + window_posn, inpos, (size_t) this_run);
1905 inpos += this_run; window_posn += this_run;
1909 return DECR_ILLEGALDATA; /* might as well */
1915 if (togo != 0) return DECR_ILLEGALDATA;
1916 memcpy(CAB(outbuf), window + ((!window_posn) ? window_size : window_posn) -
1917 outlen, (size_t) outlen);
1919 LZX(window_posn) = window_posn;
1924 /* intel E8 decoding */
1925 if ((LZX(frames_read)++ < 32768) && LZX(intel_filesize) != 0) {
1926 if (outlen <= 6 || !LZX(intel_started)) {
1927 LZX(intel_curpos) += outlen;
1930 UBYTE *data = CAB(outbuf);
1931 UBYTE *dataend = data + outlen - 10;
1932 LONG curpos = LZX(intel_curpos);
1933 LONG filesize = LZX(intel_filesize);
1934 LONG abs_off, rel_off;
1936 LZX(intel_curpos) = curpos + outlen;
1938 while (data < dataend) {
1939 if (*data++ != 0xE8) { curpos++; continue; }
1940 abs_off = data[0] | (data[1]<<8) | (data[2]<<16) | (data[3]<<24);
1941 if ((abs_off >= -curpos) && (abs_off < filesize)) {
1942 rel_off = (abs_off >= 0) ? abs_off - curpos : abs_off + filesize;
1943 data[0] = (UBYTE) rel_off;
1944 data[1] = (UBYTE) (rel_off >> 8);
1945 data[2] = (UBYTE) (rel_off >> 16);
1946 data[3] = (UBYTE) (rel_off >> 24);
1960 /* all the file IO is abstracted into these routines:
1961 * cabinet_(open|close|read|seek|skip|getoffset)
1962 * file_(open|close|write)
1965 /* ensure_filepath("a/b/c/d.txt") ensures a, a/b and a/b/c exist as dirs */
1966 int ensure_filepath(char *path) {
1972 m = umask(0); umask(m); /* obtain user's umask */
1974 for (p = path; *p; p++) {
1975 if ((p != path) && (*p == '/')) {
1977 ok = (stat(path, &st_buf) == 0) && S_ISDIR(st_buf.st_mode);
1978 if (!ok) ok = (mkdir(path, 0777 & ~m) == 0);
1986 /* opens a file for output, returns success */
1987 int file_open(struct file *fi, int lower, char *dir) {
1988 char c, *s, *d, *name;
1991 if (!(name = malloc(strlen(fi->filename) + (dir ? strlen(dir) : 0) + 2))) {
1992 g_warning(_("out of memory!"));
1996 /* start with blank name */
1999 /* add output directory if needed */
2006 /* remove leading slashes */
2008 while (*s == '\\') s++;
2010 /* copy from fi->filename to new name, converting MS-DOS slashes to UNIX
2011 * slashes as we go. Also lowercases characters if needed.
2013 d = &name[strlen(name)];
2016 *d++ = (c=='/') ? '\\' : ((c=='\\') ? '/' :
2017 (lower ? tolower((unsigned char) c) : c));
2020 /* create directories if needed, attempt to write file */
2021 if (ensure_filepath(name)) {
2022 fi->fh = fopen(name, "wb");
2026 /* as full filename is no longer needed, free it */
2030 perror(fi->filename);
2036 /* closes a completed file, updates protections and timestamp */
2037 void file_close(struct file *fi) {
2047 m = umask(0); umask(m); /* obtain user's umask */
2050 | (fi->attribs & cffile_A_EXEC ? 0111 : 0)
2051 | (fi->attribs & cffile_A_RDONLY ? 0 : 0222)) & ~ m
2055 time.tm_sec = (fi->time << 1) & 0x3e;
2056 time.tm_min = (fi->time >> 5) & 0x3f;
2057 time.tm_hour = (fi->time >> 11);
2058 time.tm_mday = fi->date & 0x1f;
2059 time.tm_mon =((fi->date >> 5) & 0xf) - 1;
2060 time.tm_year = (fi->date >> 9) + 80;
2063 utb.actime = utb.modtime = mktime(&time);
2064 utime(fi->filename, &utb);
2068 int file_write(struct file *fi, UBYTE *buf, size_t length) {
2069 if (fwrite((void *)buf, 1, length, fi->fh) != length) {
2070 perror(fi->filename);
2076 void cabinet_close(struct cabinet *cab) {
2083 #endif /* CAPTIVE */
2085 static void cabinet_seek(struct cabinet *cab, off_t offset) {
2086 acquire_cabinet_seek(cab->acquire_cabinet,offset);
2089 static void cabinet_skip(struct cabinet *cab, off_t distance) {
2090 acquire_cabinet_seek_skip(cab->acquire_cabinet,distance);
2093 static off_t cabinet_getoffset(struct cabinet *cab) {
2094 return acquire_cabinet_tell(cab->acquire_cabinet);
2097 /* read data from a cabinet, returns success */
2098 static int cabinet_read(struct cabinet *cab, UBYTE *buf, size_t length) {
2099 GnomeVFSResult errvfsresult;
2100 GnomeVFSFileSize bytes_read;
2102 errvfsresult=acquire_cabinet_read(cab->acquire_cabinet,buf,length,&bytes_read);
2103 if (errvfsresult!=GNOME_VFS_OK) {
2104 g_warning(_("%s: cabinet read error: %s"), cab->filename, gnome_vfs_result_to_string(errvfsresult));
2107 if (bytes_read!=length)
2108 g_warning(_("%s: WARNING; cabinet is truncated"), cab->filename);
2114 /* try to open a cabinet file, returns success */
2115 int cabinet_open(struct cabinet *cab) {
2116 char *name = cab->filename;
2119 /* note: this is now case sensitive */
2120 if (!(fh = fopen(name, "rb"))) {
2125 /* seek to end of file */
2126 if (fseek(fh, 0, SEEK_END) != 0) {
2132 /* get length of file */
2133 cab->filelen = ftell(fh);
2135 /* return to the start of the file */
2136 if (fseek(fh, 0, SEEK_SET) != 0) {
2146 #endif /* CAPTIVE */
2148 /* allocate and read an aribitrarily long string from the cabinet */
2149 static char *cabinet_read_string(struct cabinet *cab) {
2150 off_t len=256, base = cabinet_getoffset(cab), maxlen = cab->filelen - base;
2154 if (len > maxlen) len = maxlen;
2155 if (!(buf = realloc(buf, (size_t) len))) break;
2156 if (!cabinet_read(cab, buf, (size_t) len)) break;
2158 /* search for a null terminator in what we've just read */
2159 for (i=0; i < len; i++) {
2160 if (!buf[i]) {ok=1; break;}
2164 if (len == maxlen) {
2165 g_warning(_("%s: WARNING; cabinet is truncated"), cab->filename);
2169 cabinet_seek(cab, base);
2174 if (buf) free(buf); else g_warning(_("out of memory!"));
2178 /* otherwise, set the stream to just after the string and return */
2179 cabinet_seek(cab, base + ((off_t) strlen((char *) buf)) + 1);
2180 return (char *) buf;
2183 /* reads the header and all folder and file entries in this cabinet */
2184 static int cabinet_read_entries(struct cabinet *cab) {
2185 int num_folders, num_files, header_resv, folder_resv = 0, i;
2186 struct folder *fol, *linkfol = NULL;
2187 struct file *file, *linkfile = NULL;
2191 /* read in the CFHEADER */
2192 base_offset = cabinet_getoffset(cab);
2193 if (!cabinet_read(cab, buf, cfhead_SIZEOF)) {
2197 /* check basic MSCF signature */
2198 if (EndGetI32(buf+cfhead_Signature) != 0x4643534d) {
2199 g_warning(_("%s: not a Microsoft cabinet file"), cab->filename);
2203 /* get the number of folders */
2204 num_folders = EndGetI16(buf+cfhead_NumFolders);
2205 if (num_folders == 0) {
2206 g_warning(_("%s: no folders in cabinet"), cab->filename);
2210 /* get the number of files */
2211 num_files = EndGetI16(buf+cfhead_NumFiles);
2212 if (num_files == 0) {
2213 g_warning(_("%s: no files in cabinet"), cab->filename);
2217 /* just check the header revision */
2218 if ((buf[cfhead_MajorVersion] > 1) ||
2219 (buf[cfhead_MajorVersion] == 1 && buf[cfhead_MinorVersion] > 3))
2221 g_warning(_("%s: WARNING; cabinet format version > 1.3"),
2225 /* read the reserved-sizes part of header, if present */
2226 cab->flags = EndGetI16(buf+cfhead_Flags);
2227 if (cab->flags & cfheadRESERVE_PRESENT) {
2228 if (!cabinet_read(cab, buf, cfheadext_SIZEOF)) return 0;
2229 header_resv = EndGetI16(buf+cfheadext_HeaderReserved);
2230 folder_resv = buf[cfheadext_FolderReserved];
2231 cab->block_resv = buf[cfheadext_DataReserved];
2233 if (header_resv > 60000) {
2234 g_warning(_("%s: WARNING; header reserved space > 60000"),
2238 /* skip the reserved header */
2239 if (header_resv) cabinet_skip(cab, (off_t) header_resv);
2242 if (cab->flags & cfheadPREV_CABINET) {
2243 cab->prevname = cabinet_read_string(cab);
2244 if (!cab->prevname) return 0;
2245 cab->previnfo = cabinet_read_string(cab);
2246 if (!cab->previnfo) return 0;
2249 if (cab->flags & cfheadNEXT_CABINET) {
2250 cab->nextname = cabinet_read_string(cab);
2251 if (!cab->nextname) return 0;
2252 cab->nextinfo = cabinet_read_string(cab);
2253 if (!cab->nextinfo) return 0;
2257 for (i = 0; i < num_folders; i++) {
2258 if (!cabinet_read(cab, buf, cffold_SIZEOF)) return 0;
2259 if (folder_resv) cabinet_skip(cab, folder_resv);
2261 fol = (struct folder *) calloc(1, sizeof(struct folder));
2262 if (!fol) { g_warning(_("out of memory!")); return 0; }
2265 fol->offset[0] = base_offset + (off_t) EndGetI32(buf+cffold_DataOffset);
2266 fol->num_blocks = EndGetI16(buf+cffold_NumBlocks);
2267 fol->comp_type = EndGetI16(buf+cffold_CompType);
2269 if (!linkfol) cab->folders = fol; else linkfol->next = fol;
2274 for (i = 0; i < num_files; i++) {
2275 if (!cabinet_read(cab, buf, cffile_SIZEOF)) return 0;
2276 file = (struct file *) calloc(1, sizeof(struct file));
2277 if (!file) { g_warning(_("out of memory!")); return 0; }
2279 file->length = EndGetI32(buf+cffile_UncompressedSize);
2280 file->offset = EndGetI32(buf+cffile_FolderOffset);
2281 file->index = EndGetI16(buf+cffile_FolderIndex);
2282 file->time = EndGetI16(buf+cffile_Time);
2283 file->date = EndGetI16(buf+cffile_Date);
2284 file->attribs = EndGetI16(buf+cffile_Attribs);
2285 file->filename = cabinet_read_string(cab);
2286 if (!file->filename) return 0;
2287 if (!linkfile) cab->files = file; else linkfile->next = file;
2294 /* this does the tricky job of running through every file in the cabinet,
2295 * including spanning cabinets, and working out which file is in which
2296 * folder in which cabinet. It also throws out the duplicate file entries
2297 * that appear in spanning cabinets. There is memory leakage here because
2298 * those entries are not freed. See the XAD CAB client for an
2299 * implementation of this that correctly frees the discarded file entries.
2301 struct file *process_files(struct cabinet *basecab) {
2302 struct cabinet *cab;
2303 struct file *outfi = NULL, *linkfi = NULL, *nextfi, *fi, *cfi;
2304 struct folder *fol, *firstfol, *lastfol = NULL, *predfol;
2307 for (cab = basecab; cab; cab = cab->nextcab) {
2308 /* firstfol = first folder in this cabinet */
2309 /* lastfol = last folder in this cabinet */
2310 /* predfol = last folder in previous cabinet (or NULL if first cabinet) */
2312 firstfol = cab->folders;
2313 for (lastfol = firstfol; lastfol->next;) lastfol = lastfol->next;
2316 for (fi = cab->files; fi; fi = nextfi) {
2320 if (i < cffileCONTINUED_FROM_PREV) {
2321 for (fol = firstfol; fol && i--; ) fol = fol->next;
2322 fi->folder = fol; /* NULL if an invalid folder index */
2325 /* folder merging */
2326 if (i == cffileCONTINUED_TO_NEXT
2327 || i == cffileCONTINUED_PREV_AND_NEXT) {
2328 if (cab->nextcab && !lastfol->contfile) lastfol->contfile = fi;
2331 if (i == cffileCONTINUED_FROM_PREV
2332 || i == cffileCONTINUED_PREV_AND_NEXT) {
2333 /* these files are to be continued in yet another
2334 * cabinet, don't merge them in just yet */
2335 if (i == cffileCONTINUED_PREV_AND_NEXT) mergeok = 0;
2337 /* only merge once per cabinet */
2339 if ((cfi = predfol->contfile)
2340 && (cfi->offset == fi->offset)
2341 && (cfi->length == fi->length)
2342 && (strcmp(cfi->filename, fi->filename) == 0)
2343 && (predfol->comp_type == firstfol->comp_type)) {
2344 /* increase the number of splits */
2345 if ((i = ++(predfol->num_splits)) > CAB_SPLITMAX) {
2347 g_warning(_("%s: internal error, increase CAB_SPLITMAX"),
2351 /* copy information across from the merged folder */
2352 predfol->offset[i] = firstfol->offset[0];
2353 predfol->cab[i] = firstfol->cab[0];
2354 predfol->next = firstfol->next;
2355 predfol->contfile = firstfol->contfile;
2357 if (firstfol == lastfol) lastfol = predfol;
2359 predfol = NULL; /* don't merge again within this cabinet */
2363 /* if the folders won't merge, don't add their files */
2368 if (mergeok) fi->folder = firstfol;
2373 if (linkfi) linkfi->next = fi; else outfi = fi;
2377 } /* for (cab= ...*/
2382 /* validates and reads file entries from a cabinet at offset [offset] in
2383 * file [name]. Returns a cabinet structure if successful, or NULL
2386 static struct cabinet *load_cab_offset(struct acquire_cabinet *acquire_cabinet, off_t offset) {
2387 struct cabinet *cab = (struct cabinet *) calloc(1, sizeof(struct cabinet));
2389 if (!cab) return NULL;
2391 cab->acquire_cabinet = acquire_cabinet;
2392 cab->filename = cab->acquire_cabinet->filename;
2393 /* if ((ok = cabinet_open(cab))) * CAPTIVE */
2394 cab->filelen = acquire_cabinet->size;
2395 cabinet_seek(cab, offset);
2396 ok = cabinet_read_entries(cab);
2397 /* cabinet_close(cab); * CAPTIVE */
2404 /* Searches a file for embedded cabinets (also succeeds on just normal
2405 * cabinet files). The first result of this search will be returned, and
2406 * the remaining results will be chained to it via the cab->next structure
2409 #define SEARCH_SIZE (32*1024)
2410 static UBYTE search_buf[SEARCH_SIZE];
2412 struct cabinet *find_cabs_in_file(struct acquire_cabinet *acquire_cabinet) {
2413 struct cabinet *cab, *cab2, *firstcab = NULL, *linkcab = NULL;
2414 UBYTE *pstart = &search_buf[0], *pend, *p;
2415 ULONG offset, caboff, cablen = 0; /* Prevent: ... might be used uninitialized in this function */
2416 ULONG foffset = 0; /* Prevent: ... might be used uninitialized in this function */
2419 int state = 0, found = 0, ok = 0;
2421 /* open the file and search for cabinet headers */
2422 if ((cab = (struct cabinet *) calloc(1, sizeof(struct cabinet)))) {
2423 cab->acquire_cabinet = acquire_cabinet;
2424 cab->filename = acquire_cabinet->filename;
2425 cab->filelen = acquire_cabinet->size;
2426 if (1 /* cabinet_open(cab) * CAPTIVE */) {
2427 filelen = (ULONG) cab->filelen;
2428 for (offset = 0; offset < filelen; offset += length) {
2429 /* search length is either the full length of the search buffer,
2430 * or the amount of data remaining to the end of the file,
2431 * whichever is less.
2433 length = filelen - offset;
2434 if (length > SEARCH_SIZE) length = SEARCH_SIZE;
2436 /* fill the search buffer with data from disk */
2437 if (!cabinet_read(cab, search_buf, length)) break;
2439 /* read through the entire buffer. */
2441 pend = &search_buf[length];
2444 /* starting state */
2446 /* we spend most of our time in this while loop, looking for
2447 * a leading 'M' of the 'MSCF' signature
2449 while (*p++ != 0x4D && p < pend);
2450 if (p < pend) state = 1; /* if we found tht 'M', advance state */
2453 /* verify that the next 3 bytes are 'S', 'C' and 'F' */
2454 case 1: state = (*p++ == 0x53) ? 2 : 0; break;
2455 case 2: state = (*p++ == 0x43) ? 3 : 0; break;
2456 case 3: state = (*p++ == 0x46) ? 4 : 0; break;
2458 /* we don't care about bytes 4-7 */
2459 /* bytes 8-11 are the overall length of the cabinet */
2460 case 8: cablen = *p++; state++; break;
2461 case 9: cablen |= *p++ << 8; state++; break;
2462 case 10: cablen |= *p++ << 16; state++; break;
2463 case 11: cablen |= *p++ << 24; state++; break;
2465 /* we don't care about bytes 12-15 */
2466 /* bytes 16-19 are the offset within the cabinet of the filedata */
2467 case 16: foffset = *p++; state++; break;
2468 case 17: foffset |= *p++ << 8; state++; break;
2469 case 18: foffset |= *p++ << 16; state++; break;
2470 case 19: foffset |= *p++ << 24;
2471 /* now we have recieved 20 bytes of potential cab header. */
2472 /* work out the offset in the file of this potential cabinet */
2473 caboff = offset + (p-pstart) - 20;
2475 /* check that the files offset is less than the alleged length
2476 * of the cabinet, and that the offset + the alleged length are
2477 * 'roughly' within the end of overall file length
2479 if ((foffset < cablen) &&
2480 ((caboff + foffset) < (filelen + 32)) &&
2481 ((caboff + cablen) < (filelen + 32)) )
2483 /* found a potential result - try loading it */
2485 cab2 = load_cab_offset(acquire_cabinet, (off_t) caboff);
2490 /* cause the search to restart after this cab's data. */
2491 offset = caboff + cablen;
2492 if (offset < cab->filelen) cabinet_seek(cab, offset);
2496 /* link the cab into the list */
2497 if (linkcab == NULL) firstcab = cab2;
2498 else linkcab->next = cab2;
2505 p++, state++; break;
2509 /* cabinet_close(cab); * CAPTIVE */
2514 /* if there were cabinets that were found but are not ok, point this out */
2516 g_warning(_("%s: WARNING; found %d bad cabinets"), acquire_cabinet->filename, found-ok);
2519 /* if no cabinets were found, let the user know */
2521 g_warning(_("%s: not a Microsoft cabinet file."), acquire_cabinet->filename);
2528 /* UTF translates two-byte unicode characters into 1, 2 or 3 bytes.
2529 * %000000000xxxxxxx -> %0xxxxxxx
2530 * %00000xxxxxyyyyyy -> %110xxxxx %10yyyyyy
2531 * %xxxxyyyyyyzzzzzz -> %1110xxxx %10yyyyyy %10zzzzzz
2533 * Therefore, the inverse is as follows:
2535 * 0x00 - 0x7F = one byte char
2536 * 0x80 - 0xBF = invalid
2537 * 0xC0 - 0xDF = 2 byte char (next char only 0x80-0xBF is valid)
2538 * 0xE0 - 0xEF = 3 byte char (next 2 chars only 0x80-0xBF is valid)
2539 * 0xF0 - 0xFF = invalid
2542 /* translate UTF -> ASCII */
2543 static int convertUTF(UBYTE *in) {
2544 UBYTE c, *out = in, *end = in + strlen((char *) in) + 1;
2548 /* read unicode character */
2549 if ((c = *in++) < 0x80) x = c;
2551 if (c < 0xC0) return 0;
2552 else if (c < 0xE0) {
2553 x = (c & 0x1F) << 6;
2554 if ((c = *in++) < 0x80 || c > 0xBF) return 0; else x |= (c & 0x3F);
2556 else if (c < 0xF0) {
2557 x = (c & 0xF) << 12;
2558 if ((c = *in++) < 0x80 || c > 0xBF) return 0; else x |= (c & 0x3F)<<6;
2559 if ((c = *in++) < 0x80 || c > 0xBF) return 0; else x |= (c & 0x3F);
2564 /* terrible unicode -> ASCII conversion */
2565 if (x > 127) x = '_';
2567 if (in > end) return 0; /* just in case */
2568 } while ((*out++ = (UBYTE) x));
2572 void print_fileinfo(struct file *fi) {
2573 int d = fi->date, t = fi->time;
2576 if (fi->attribs & cffile_A_NAME_IS_UTF) {
2577 fname = malloc(strlen(fi->filename) + 1);
2579 strcpy(fname, fi->filename);
2580 convertUTF((UBYTE *) fname);
2584 printf("%9u | %02d.%02d.%04d %02d:%02d:%02d | %s\n",
2586 d & 0x1f, (d>>5) & 0xf, (d>>9) + 1980,
2587 t >> 11, (t>>5) & 0x3f, (t << 1) & 0x3e,
2588 fname ? fname : fi->filename
2591 if (fname) free(fname);
2594 #endif /* CAPTIVE */
2596 static int NONEdecompress(int inlen, int outlen) {
2597 if (inlen != outlen) return DECR_ILLEGALDATA;
2598 memcpy(CAB(outbuf), CAB(inbuf), (size_t) inlen);
2602 static ULONG checksum(UBYTE *data, UWORD bytes, ULONG csum) {
2606 for (len = bytes >> 2; len--; data += 4) {
2607 csum ^= ((data[0]) | (data[1]<<8) | (data[2]<<16) | (data[3]<<24));
2610 switch (bytes & 3) {
2611 case 3: ul |= *data++ << 16;
2612 case 2: ul |= *data++ << 8;
2613 case 1: ul |= *data;
2620 int file_write(struct file *fi, UBYTE *buf, size_t length);
2622 static int decompress(struct file *fi, int savemode, int fix) {
2623 ULONG bytes = savemode ? fi->length : fi->offset - CAB(offset);
2624 struct cabinet *cab = CAB(current)->cab[CAB(split)];
2625 UBYTE buf[cfdata_SIZEOF], *data;
2626 UWORD inlen, len, outlen, cando;
2631 /* cando = the max number of bytes we can do */
2632 cando = CAB(outlen);
2633 if (cando > bytes) cando = bytes;
2636 if (cando && savemode) file_write(fi, CAB(outpos), cando);
2638 CAB(outpos) += cando;
2639 CAB(outlen) -= cando;
2640 bytes -= cando; if (!bytes) break;
2642 /* we only get here if we emptied the output buffer */
2644 /* read data header + data */
2646 while (outlen == 0) {
2647 /* read the block header, skip the reserved part */
2648 if ((NONEdecompress==CAB(decompress) && !savemode && bytes>32768)) {
2649 cabinet_skip(cab, cfdata_SIZEOF);
2650 memset(buf + cfdata_CheckSum, 0, 4); /* no CheckSum */
2651 /* FIXME: Is it safe to assume 'NONEdecompress' block size 32768?
2652 * Probably not but we need to prevent scattering block headers through the file.
2654 buf[cfdata_CompressedSize + 0]=(32768>>0)&0xFF;
2655 buf[cfdata_CompressedSize + 1]=(32768>>8)&0xFF;
2656 buf[cfdata_UncompressedSize + 0]=(32768>>0)&0xFF;
2657 buf[cfdata_UncompressedSize + 1]=(32768>>8)&0xFF;
2659 if (!cabinet_read(cab, buf, cfdata_SIZEOF)) return DECR_INPUT;
2661 cabinet_skip(cab, cab->block_resv);
2663 /* we shouldn't get blocks over CAB_INPUTMAX in size */
2664 data = CAB(inbuf) + inlen;
2665 len = EndGetI16(buf+cfdata_CompressedSize);
2667 if (inlen > CAB_INPUTMAX) return DECR_INPUT;
2668 if ((NONEdecompress==CAB(decompress) && !savemode && bytes>32768)) {
2669 cabinet_skip(cab, len);
2671 if (!cabinet_read(cab, data, len)) return DECR_INPUT;
2674 /* clear two bytes after read-in data */
2675 data[len+1] = data[len+2] = 0;
2677 /* perform checksum test on the block (if one is stored) */
2678 cksum = EndGetI32(buf+cfdata_CheckSum);
2679 if (!(NONEdecompress==CAB(decompress) && !savemode && bytes>32768)) {
2680 if (cksum && cksum != checksum(buf+4, 4, checksum(data, len, 0))) {
2681 /* checksum is wrong */
2682 if (fix && ((fi->folder->comp_type & cffoldCOMPTYPE_MASK)
2683 == cffoldCOMPTYPE_MSZIP))
2685 g_warning(_("%s: WARNING; checksum failed"), fi->filename);
2688 return DECR_CHECKSUM;
2693 /* outlen=0 means this block was part of a split block */
2694 outlen = EndGetI16(buf+cfdata_UncompressedSize);
2698 cab = CAB(current)->cab[++CAB(split)];
2699 if (!cabinet_open(cab)) return DECR_INPUT;
2700 cabinet_seek(cab, CAB(current)->offset[CAB(split)]);
2707 if (!(NONEdecompress==CAB(decompress) && !savemode && bytes>32768)) {
2708 /* decompress block */
2709 if ((err = CAB(decompress)(inlen, outlen))) {
2710 if (fix && ((fi->folder->comp_type & cffoldCOMPTYPE_MASK)
2711 == cffoldCOMPTYPE_MSZIP))
2713 g_warning(_("%s: WARNING; failed decrunching block"),
2721 CAB(outlen) = outlen;
2722 CAB(outpos) = CAB(outbuf);
2729 int extract_file(struct file *fi, int lower, int fix, char *dir) {
2730 struct folder *fol = fi->folder, *oldfol = CAB(current);
2733 /* is a change of folder needed? do we need to reset the current folder? */
2734 if (fol != oldfol || fi->offset < CAB(offset)) {
2735 UWORD comptype = fol->comp_type;
2736 int ct1 = comptype & cffoldCOMPTYPE_MASK;
2737 int ct2 = oldfol ? (oldfol->comp_type & cffoldCOMPTYPE_MASK) : 0;
2739 /* if the archiver has changed, call the old archiver's free() function */
2742 case cffoldCOMPTYPE_LZX:
2748 case cffoldCOMPTYPE_QUANTUM:
2758 case cffoldCOMPTYPE_NONE:
2759 CAB(decompress) = NONEdecompress;
2762 case cffoldCOMPTYPE_MSZIP:
2763 CAB(decompress) = ZIPdecompress;
2766 case cffoldCOMPTYPE_QUANTUM:
2767 CAB(decompress) = QTMdecompress;
2768 err = QTMinit((comptype >> 8) & 0x1f, (comptype >> 4) & 0xF);
2771 case cffoldCOMPTYPE_LZX:
2772 CAB(decompress) = LZXdecompress;
2773 err = LZXinit((comptype >> 8) & 0x1f);
2777 err = DECR_DATAFORMAT;
2779 if (err) goto exit_handler;
2781 /* initialisation OK, set current folder and reset offset */
2783 if (oldfol) cabinet_close(oldfol->cab[CAB(split)]);
2784 if (!cabinet_open(fol->cab[0])) {
2785 err = DECR_ILLEGALDATA;
2788 #endif /* CAPTIVE */
2789 cabinet_seek(fol->cab[0], fol->offset[0]);
2792 CAB(outlen) = 0; /* discard existing block */
2796 if (fi->offset > CAB(offset)) {
2797 /* decode bytes and send them to /dev/null */
2798 if ((err = decompress(fi, 0, fix))) goto exit_handler;
2799 CAB(offset) = fi->offset;
2802 if (!file_open(fi, lower, dir)) return 0;
2803 #endif /* CAPTIVE */
2804 err = decompress(fi, 1, fix);
2805 if (err) CAB(current) = NULL; else CAB(offset) += fi->length;
2808 #endif /* CAPTIVE */
2812 const char *errmsg, *cabname;
2815 errmsg = _("out of memory!"); break;
2816 case DECR_ILLEGALDATA:
2817 errmsg = _("%s: illegal or corrupt data"); break;
2818 case DECR_DATAFORMAT:
2819 errmsg = _("%s: unsupported data format"); break;
2821 errmsg = _("%s: checksum error"); break;
2823 errmsg = _("%s: input error"); break;
2825 errmsg = _("%s: output error"); break;
2827 errmsg = _("%s: unknown error (BUG)");
2831 cabname = CAB(current)->cab[CAB(split)]->filename;
2834 cabname = fi->folder->cab[0]->filename;
2837 g_warning(errmsg, cabname);
2845 /* tries to find *cabname, from the directory path of origcab, correcting the
2846 * case of *cabname if necessary, If found, writes back to *cabname.
2848 void find_cabinet_file(char **cabname, char *origcab) {
2849 char *tail, *cab, *name, *nextpart;
2850 struct dirent *entry;
2855 /* ensure we have a cabinet name at all */
2856 if (!(name = *cabname)) return;
2858 /* find if there's a directory path in the origcab */
2859 tail = origcab ? strrchr(origcab, '/') : NULL;
2861 if ((cab = (char *) malloc((tail ? tail-origcab : 1) + strlen(name) + 2))) {
2862 /* add the directory path from the original cabinet name */
2864 memcpy(cab, origcab, tail-origcab);
2865 cab[tail-origcab] = '\0';
2868 /* default directory path of '.' */
2874 /* we don't want null cabinet filenames */
2875 if (name[0] == '\0') break;
2877 /* if there is a directory component in the cabinet name,
2878 * look for that alone first
2880 nextpart = strchr(name, '\\');
2881 if (nextpart) *nextpart = '\0';
2883 /* try accessing the component with its current name (case-sensitive) */
2884 len = strlen(cab); strcat(cab, "/"); strcat(cab, name);
2885 found = (stat(cab, &st_buf) == 0) &&
2886 nextpart ? S_ISDIR(st_buf.st_mode) : S_ISREG(st_buf.st_mode);
2888 /* if the component was not found, look for it in the current dir */
2891 if ((dir = opendir(cab))) {
2892 while ((entry = readdir(dir))) {
2893 if (strcasecmp(name, entry->d_name) == 0) {
2894 strcat(cab, "/"); strcat(cab, entry->d_name); found = 1;
2901 /* restore the real name and skip to the next directory component
2902 * or actual cabinet name
2904 if (nextpart) *nextpart = '\\', name = &nextpart[1];
2906 /* while there is another directory component, and while we
2907 * successfully found the current component
2909 } while (nextpart && found);
2912 /* if we found the cabinet, change the next cabinet's name.
2913 * otherwise, pretend nothing happened
2916 free((void *) *cabname);
2926 /* process_cabinet() is called by main() for every file listed on the
2927 * command line. It will find every cabinet file in that file, and will
2928 * search for every chained cabinet attached to those cabinets, then it
2929 * will either extract or list the cabinet(s). Returns 0 for success or 1
2930 * for failure (unlike most cabextract functions).
2932 int process_cabinet(char *cabname, char *dir,
2933 int fix, int view, int lower, int quiet) {
2935 struct cabinet *basecab, *cab, *cab1, *cab2;
2936 struct file *filelist, *fi;
2938 /* has the list-mode header been seen before? */
2941 if (view || !quiet) {
2942 printf("%s cabinet: %s\n", view ? "Viewing" : "Extracting", cabname);
2945 /* load the file requested */
2946 basecab = find_cabs_in_file(cabname);
2947 if (!basecab) return 1;
2949 /* iterate over all cabinets found in that file */
2950 for (cab = basecab; cab; cab=cab->next) {
2952 /* bi-directionally load any spanning cabinets -- backwards */
2953 for (cab1 = cab; cab1->flags & cfheadPREV_CABINET; cab1 = cab1->prevcab) {
2954 if (!quiet) printf("%s: extends backwards to %s (%s)\n", cabname,
2955 cab1->prevname, cab1->previnfo);
2956 find_cabinet_file(&(cab1->prevname), cabname);
2957 if (!(cab1->prevcab = load_cab_offset(cab1->prevname, 0))) {
2958 g_warning(_("%s: can't read previous cabinet %s"),
2959 cabname, cab1->prevname);
2962 cab1->prevcab->nextcab = cab1;
2965 /* bi-directionally load any spanning cabinets -- forwards */
2966 for (cab2 = cab; cab2->flags & cfheadNEXT_CABINET; cab2 = cab2->nextcab) {
2967 if (!quiet) printf("%s: extends to %s (%s)\n", cabname,
2968 cab2->nextname, cab2->nextinfo);
2969 find_cabinet_file(&(cab2->nextname), cabname);
2970 if (!(cab2->nextcab = load_cab_offset(cab2->nextname, 0))) {
2971 g_warning(_("%s: can't read next cabinet %s"),
2972 cabname, cab2->nextname);
2975 cab2->nextcab->prevcab = cab2;
2978 filelist = process_files(cab1);
2979 CAB(current) = NULL;
2981 if (view && !viewhdr) {
2982 printf("File size | Date Time | Name\n");
2983 printf("----------+---------------------+-------------\n");
2986 for (fi = filelist; fi; fi = fi->next) {
2991 if (!quiet) printf(" extracting: %s\n", fi->filename);
2992 extract_file(fi, lower, fix, dir);
2997 if (view) printf("\n");
2998 else if (!quiet) printf("Finished processing cabinet.\n\n");
3003 struct option opts[] = {
3004 { "version", 0, NULL, 'v' },
3005 { "help", 0, NULL, 'h' },
3006 { "list", 0, NULL, 'l' },
3007 { "quiet", 0, NULL, 'q' },
3008 { "lowercase", 0, NULL, 'L' },
3009 { "fix", 0, NULL, 'f' },
3010 { "directory", 1, NULL, 'd' },
3011 { NULL, 0, NULL, 0 }
3014 int main(int argc, char *argv[]) {
3015 int help=0, list=0, lower=0, view=0, quiet=0, fix=0, x, err=0;
3017 while ((x = getopt_long(argc, argv, "vhlqLfd:", opts, NULL)) != -1) {
3019 case 'v': view = 1; break;
3020 case 'h': help = 1; break;
3021 case 'l': list = 1; break;
3022 case 'q': quiet = 1; break;
3023 case 'L': lower = 1; break;
3024 case 'f': fix = 1; break;
3025 case 'd': dir = optarg; break;
3031 "Usage: %s [options] [-d dir] <cabinet file(s)>\n\n"
3032 "This will extract all files from a cabinet or executable cabinet.\n"
3033 "For multi-part cabinets, only specify the first file in the set.\n\n"
3035 " -v --version print version / list cabinet\n"
3036 " -h --help show this help page\n"
3037 " -l --list list contents of cabinet\n"
3038 " -q --quiet only print errors and warnings\n"
3039 " -L --lowercase make filenames lowercase\n"
3040 " -f --fix fix (some) corrupted cabinets\n"
3041 " -d --directory extract all files to the given directory\n\n"
3042 "cabextract %s (C) 2000-2002 Stuart Caie <kyzer@4u.net>\n"
3043 "This is free software with ABSOLUTELY NO WARRANTY.\n",
3049 if (optind == argc) {
3050 /* no arguments other than the options */
3052 printf("cabextract version %s\n", VERSION);
3056 fprintf(stderr, "cabextract: No cabinet files specified.\n"
3057 "Try '%s --help' for more information.\n", argv[0]);
3062 while (optind != argc) {
3063 err |= process_cabinet(argv[optind++], dir, fix, view||list, lower, quiet);
3069 #endif /* CAPTIVE */
git://git.jankratochvil.net / captive.git / blob