1 
2 /*-----------------------------------------------------------*/
3 /*--- Block recoverer program for bzip2                   ---*/
4 /*---                                      bzip2recover.c ---*/
5 /*-----------------------------------------------------------*/
6 
7 /*--
8   This program is bzip2recover, a program to attempt data 
9   salvage from damaged files created by the accompanying
10   bzip2-1.0 program.
11 
12   Copyright (C) 1996-2000 Julian R Seward.  All rights reserved.
13 
14   Redistribution and use in source and binary forms, with or without
15   modification, are permitted provided that the following conditions
16   are met:
17 
18   1. Redistributions of source code must retain the above copyright
19      notice, this list of conditions and the following disclaimer.
20 
21   2. The origin of this software must not be misrepresented; you must 
22      not claim that you wrote the original software.  If you use this 
23      software in a product, an acknowledgment in the product 
24      documentation would be appreciated but is not required.
25 
26   3. Altered source versions must be plainly marked as such, and must
27      not be misrepresented as being the original software.
28 
29   4. The name of the author may not be used to endorse or promote 
30      products derived from this software without specific prior written 
31      permission.
32 
33   THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS
34   OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
35   WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
36   ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY
37   DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
38   DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE
39   GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
40   INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
41   WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
42   NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
43   SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
44 
45   Julian Seward, Cambridge, UK.
46   jseward@acm.org
47   bzip2/libbzip2 version 1.0 of 21 March 2000
48 --*/
49 
50 /*--
51   This program is a complete hack and should be rewritten
52   properly.  It isn't very complicated.
53 --*/
54 
55 #include <stdio.h>
56 #include <errno.h>
57 #include <stdlib.h>
58 #include <string.h>
59 
60 typedef  unsigned int   UInt32;
61 typedef  int            Int32;
62 typedef  unsigned char  UChar;
63 typedef  char           Char;
64 typedef  unsigned char  Bool;
65 #define True    ((Bool)1)
66 #define False   ((Bool)0)
67 
68 
69 Char inFileName[2000];
70 Char outFileName[2000];
71 Char progName[2000];
72 
73 UInt32 bytesOut = 0;
74 UInt32 bytesIn  = 0;
75 
76 
77 /*---------------------------------------------------*/
78 /*--- I/O errors                                  ---*/
79 /*---------------------------------------------------*/
80 
81 /*---------------------------------------------*/
82 void readError ( void )
83 {
84    fprintf ( stderr,
85              "%s: I/O error reading `%s', possible reason follows.\n",
86             progName, inFileName );
87    perror ( progName );
88    fprintf ( stderr, "%s: warning: output file(s) may be incomplete.\n",
89              progName );
90    exit ( 1 );
91 }
92 
93 
94 /*---------------------------------------------*/
95 void writeError ( void )
96 {
97    fprintf ( stderr,
98              "%s: I/O error reading `%s', possible reason follows.\n",
99             progName, inFileName );
100    perror ( progName );
101    fprintf ( stderr, "%s: warning: output file(s) may be incomplete.\n",
102              progName );
103    exit ( 1 );
104 }
105 
106 
107 /*---------------------------------------------*/
108 void mallocFail ( Int32 n )
109 {
110    fprintf ( stderr,
111              "%s: malloc failed on request for %d bytes.\n",
112             progName, n );
113    fprintf ( stderr, "%s: warning: output file(s) may be incomplete.\n",
114              progName );
115    exit ( 1 );
116 }
117 
118 
119 /*---------------------------------------------------*/
120 /*--- Bit stream I/O                              ---*/
121 /*---------------------------------------------------*/
122 
123 typedef
124    struct {
125       FILE*  handle;
126       Int32  buffer;
127       Int32  buffLive;
128       Char   mode;
129    }
130    BitStream;
131 
132 
133 /*---------------------------------------------*/
134 BitStream* bsOpenReadStream ( FILE* stream )
135 {
136    BitStream *bs = malloc ( sizeof(BitStream) );
137    if (bs == NULL) mallocFail ( sizeof(BitStream) );
138    bs->handle = stream;
139    bs->buffer = 0;
140    bs->buffLive = 0;
141    bs->mode = 'r';
142    return bs;
143 }
144 
145 
146 /*---------------------------------------------*/
147 BitStream* bsOpenWriteStream ( FILE* stream )
148 {
149    BitStream *bs = malloc ( sizeof(BitStream) );
150    if (bs == NULL) mallocFail ( sizeof(BitStream) );
151    bs->handle = stream;
152    bs->buffer = 0;
153    bs->buffLive = 0;
154    bs->mode = 'w';
155    return bs;
156 }
157 
158 
159 /*---------------------------------------------*/
160 void bsPutBit ( BitStream* bs, Int32 bit )
161 {
162    if (bs->buffLive == 8) {
163       Int32 retVal = putc ( (UChar) bs->buffer, bs->handle );
164       if (retVal == EOF) writeError();
165       bytesOut++;
166       bs->buffLive = 1;
167       bs->buffer = bit & 0x1;
168    } else {
169       bs->buffer = ( (bs->buffer << 1) | (bit & 0x1) );
170       bs->buffLive++;
171    };
172 }
173 
174 
175 /*---------------------------------------------*/
176 /*--
177    Returns 0 or 1, or 2 to indicate EOF.
178 --*/
179 Int32 bsGetBit ( BitStream* bs )
180 {
181    if (bs->buffLive > 0) {
182       bs->buffLive --;
183       return ( ((bs->buffer) >> (bs->buffLive)) & 0x1 );
184    } else {
185       Int32 retVal = getc ( bs->handle );
186       if ( retVal == EOF ) {
187          if (errno != 0) readError();
188          return 2;
189       }
190       bs->buffLive = 7;
191       bs->buffer = retVal;
192       return ( ((bs->buffer) >> 7) & 0x1 );
193    }
194 }
195 
196 
197 /*---------------------------------------------*/
198 void bsClose ( BitStream* bs )
199 {
200    Int32 retVal;
201 
202    if ( bs->mode == 'w' ) {
203       while ( bs->buffLive < 8 ) {
204          bs->buffLive++;
205          bs->buffer <<= 1;
206       };
207       retVal = putc ( (UChar) (bs->buffer), bs->handle );
208       if (retVal == EOF) writeError();
209       bytesOut++;
210       retVal = fflush ( bs->handle );
211       if (retVal == EOF) writeError();
212    }
213    retVal = fclose ( bs->handle );
214    if (retVal == EOF) {
215       if (bs->mode == 'w') writeError(); else readError();
216    }
217    free ( bs );
218 }
219 
220 
221 /*---------------------------------------------*/
222 void bsPutUChar ( BitStream* bs, UChar c )
223 {
224    Int32 i;
225    for (i = 7; i >= 0; i--)
226       bsPutBit ( bs, (((UInt32) c) >> i) & 0x1 );
227 }
228 
229 
230 /*---------------------------------------------*/
231 void bsPutUInt32 ( BitStream* bs, UInt32 c )
232 {
233    Int32 i;
234 
235    for (i = 31; i >= 0; i--)
236       bsPutBit ( bs, (c >> i) & 0x1 );
237 }
238 
239 
240 /*---------------------------------------------*/
241 Bool endsInBz2 ( Char* name )
242 {
243    Int32 n = strlen ( name );
244    if (n <= 4) return False;
245    return
246       (name[n-4] == '.' &&
247        name[n-3] == 'b' &&
248        name[n-2] == 'z' &&
249        name[n-1] == '2');
250 }
251 
252 
253 /*---------------------------------------------------*/
254 /*---                                             ---*/
255 /*---------------------------------------------------*/
256 
257 #define BLOCK_HEADER_HI  0x00003141UL
258 #define BLOCK_HEADER_LO  0x59265359UL
259 
260 #define BLOCK_ENDMARK_HI 0x00001772UL
261 #define BLOCK_ENDMARK_LO 0x45385090UL
262 
263 
264 UInt32 bStart[20000];
265 UInt32 bEnd[20000];
266 UInt32 rbStart[20000];
267 UInt32 rbEnd[20000];
268 
269 Int32 main ( Int32 argc, Char** argv )
270 {
271    FILE*       inFile;
272    FILE*       outFile;
273    BitStream*  bsIn, *bsWr;
274    Int32       currBlock, b, wrBlock;
275    UInt32      bitsRead;
276    Int32       rbCtr;
277 
278 
279    UInt32      buffHi, buffLo, blockCRC;
280    Char*       p;
281 
282    strcpy ( progName, argv[0] );
283    inFileName[0] = outFileName[0] = 0;
284 
285    fprintf ( stderr, "bzip2recover 1.0: extracts blocks from damaged .bz2 files.\n" );
286 
287    if (argc != 2) {
288       fprintf ( stderr, "%s: usage is `%s damaged_file_name'.\n",
289                         progName, progName );
290       exit(1);
291    }
292 
293    strcpy ( inFileName, argv[1] );
294 
295    inFile = fopen ( inFileName, "rb" );
296    if (inFile == NULL) {
297       fprintf ( stderr, "%s: can't read `%s'\n", progName, inFileName );
298       exit(1);
299    }
300 
301    bsIn = bsOpenReadStream ( inFile );
302    fprintf ( stderr, "%s: searching for block boundaries ...\n", progName );
303 
304    bitsRead = 0;
305    buffHi = buffLo = 0;
306    currBlock = 0;
307    bStart[currBlock] = 0;
308 
309    rbCtr = 0;
310 
311    while (True) {
312       b = bsGetBit ( bsIn );
313       bitsRead++;
314       if (b == 2) {
315          if (bitsRead >= bStart[currBlock] &&
316             (bitsRead - bStart[currBlock]) >= 40) {
317             bEnd[currBlock] = bitsRead-1;
318             if (currBlock > 0)
319                fprintf ( stderr, "   block %d runs from %d to %d (incomplete)\n",
320                          currBlock,  bStart[currBlock], bEnd[currBlock] );
321          } else
322             currBlock--;
323          break;
324       }
325       buffHi = (buffHi << 1) | (buffLo >> 31);
326       buffLo = (buffLo << 1) | (b & 1);
327       if ( ( (buffHi & 0x0000ffff) == BLOCK_HEADER_HI 
328              && buffLo == BLOCK_HEADER_LO)
329            || 
330            ( (buffHi & 0x0000ffff) == BLOCK_ENDMARK_HI 
331              && buffLo == BLOCK_ENDMARK_LO)
332          ) {
333          if (bitsRead > 49)
334             bEnd[currBlock] = bitsRead-49; else
335             bEnd[currBlock] = 0;
336          if (currBlock > 0 &&
337 	     (bEnd[currBlock] - bStart[currBlock]) >= 130) {
338             fprintf ( stderr, "   block %d runs from %d to %d\n",
339                       rbCtr+1,  bStart[currBlock], bEnd[currBlock] );
340             rbStart[rbCtr] = bStart[currBlock];
341             rbEnd[rbCtr] = bEnd[currBlock];
342             rbCtr++;
343          }
344          currBlock++;
345 
346          bStart[currBlock] = bitsRead;
347       }
348    }
349 
350    bsClose ( bsIn );
351 
352    /*-- identified blocks run from 1 to rbCtr inclusive. --*/
353 
354    if (rbCtr < 1) {
355       fprintf ( stderr,
356                 "%s: sorry, I couldn't find any block boundaries.\n",
357                 progName );
358       exit(1);
359    };
360 
361    fprintf ( stderr, "%s: splitting into blocks\n", progName );
362 
363    inFile = fopen ( inFileName, "rb" );
364    if (inFile == NULL) {
365       fprintf ( stderr, "%s: can't open `%s'\n", progName, inFileName );
366       exit(1);
367    }
368    bsIn = bsOpenReadStream ( inFile );
369 
370    /*-- placate gcc's dataflow analyser --*/
371    blockCRC = 0; bsWr = 0;
372 
373    bitsRead = 0;
374    outFile = NULL;
375    wrBlock = 0;
376    while (True) {
377       b = bsGetBit(bsIn);
378       if (b == 2) break;
379       buffHi = (buffHi << 1) | (buffLo >> 31);
380       buffLo = (buffLo << 1) | (b & 1);
381       if (bitsRead == 47+rbStart[wrBlock]) 
382          blockCRC = (buffHi << 16) | (buffLo >> 16);
383 
384       if (outFile != NULL && bitsRead >= rbStart[wrBlock]
385                           && bitsRead <= rbEnd[wrBlock]) {
386          bsPutBit ( bsWr, b );
387       }
388 
389       bitsRead++;
390 
391       if (bitsRead == rbEnd[wrBlock]+1) {
392          if (outFile != NULL) {
393             bsPutUChar ( bsWr, 0x17 ); bsPutUChar ( bsWr, 0x72 );
394             bsPutUChar ( bsWr, 0x45 ); bsPutUChar ( bsWr, 0x38 );
395             bsPutUChar ( bsWr, 0x50 ); bsPutUChar ( bsWr, 0x90 );
396             bsPutUInt32 ( bsWr, blockCRC );
397             bsClose ( bsWr );
398          }
399          if (wrBlock >= rbCtr) break;
400          wrBlock++;
401       } else
402       if (bitsRead == rbStart[wrBlock]) {
403          outFileName[0] = 0;
404          sprintf ( outFileName, "rec%4d", wrBlock+1 );
405          for (p = outFileName; *p != 0; p++) if (*p == ' ') *p = '0';
406          strcat ( outFileName, inFileName );
407          if ( !endsInBz2(outFileName)) strcat ( outFileName, ".bz2" );
408 
409          fprintf ( stderr, "   writing block %d to `%s' ...\n",
410                            wrBlock+1, outFileName );
411 
412          outFile = fopen ( outFileName, "wb" );
413          if (outFile == NULL) {
414             fprintf ( stderr, "%s: can't write `%s'\n",
415                       progName, outFileName );
416             exit(1);
417          }
418          bsWr = bsOpenWriteStream ( outFile );
419          bsPutUChar ( bsWr, 'B' ); bsPutUChar ( bsWr, 'Z' );
420          bsPutUChar ( bsWr, 'h' ); bsPutUChar ( bsWr, '9' );
421          bsPutUChar ( bsWr, 0x31 ); bsPutUChar ( bsWr, 0x41 );
422          bsPutUChar ( bsWr, 0x59 ); bsPutUChar ( bsWr, 0x26 );
423          bsPutUChar ( bsWr, 0x53 ); bsPutUChar ( bsWr, 0x59 );
424       }
425    }
426 
427    fprintf ( stderr, "%s: finished\n", progName );
428    return 0;
429 }
430 
431 
432 
433 /*-----------------------------------------------------------*/
434 /*--- end                                  bzip2recover.c ---*/
435 /*-----------------------------------------------------------*/