summaryrefslogtreecommitdiffstats
path: root/merge.cc
blob: 08a3d0e71980e3f3653cc7df20ba8b2a084bca69 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
/*  Lziprecover - Data recovery tool for lzip files
    Copyright (C) 2009, 2010, 2011, 2012, 2013, 2014 Antonio Diaz Diaz.

    This program is free software: you can redistribute it and/or modify
    it under the terms of the GNU General Public License as published by
    the Free Software Foundation, either version 3 of the License, or
    (at your option) any later version.

    This program is distributed in the hope that it will be useful,
    but WITHOUT ANY WARRANTY; without even the implied warranty of
    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
    GNU General Public License for more details.

    You should have received a copy of the GNU General Public License
    along with this program.  If not, see <http://www.gnu.org/licenses/>.
*/

#define _FILE_OFFSET_BITS 64

#include <cerrno>
#include <climits>
#include <cstdio>
#include <cstdlib>
#include <cstring>
#include <string>
#include <vector>
#include <stdint.h>
#include <unistd.h>
#include <sys/stat.h>

#include "lzip.h"
#include "decoder.h"
#include "file_index.h"


namespace {

// Add 'bv' to 'block_vector' splitting blocks as needed to keep all the
// edges (pos and end of every block).
// 'block_vector' contains the result. 'bv' is destroyed.
void combine( std::vector< Block > & block_vector, std::vector< Block > & bv )
  {
  if( block_vector.empty() ) { block_vector.swap( bv ); return; }
  unsigned i1 = 0, i2 = 0;
  while( i1 < block_vector.size() && i2 < bv.size() )
    {
    Block & b1 = block_vector[i1];
    Block & b2 = bv[i2];
    if( b1.overlaps( b2 ) )
      {
      if( b1 < b2 )
        {
        Block b = b1.split( b2.pos() );
        block_vector.insert( block_vector.begin() + i1, b ); ++i1;
        }
      else if( b2 < b1 )
        {
        Block b( b2.pos(), b1.pos() - b2.pos() );
        b2.split( b1.pos() );
        block_vector.insert( block_vector.begin() + i1, b ); ++i1;
        }
      else if( b1.end() < b2.end() ) { b2.split( b1.end() ); ++i1; }
      else if( b2.end() < b1.end() )
        {
        Block b = b1.split( b2.end() );
        block_vector.insert( block_vector.begin() + i1, b ); ++i1; ++i2;
        }
      else { ++i1; ++i2; }		// blocks are identical
      }
    else if( b1 < b2 ) ++i1;
    else { block_vector.insert( block_vector.begin() + i1, b2 ); ++i1; ++i2; }
    }
  if( i2 < bv.size() )				// tail copy
    block_vector.insert( block_vector.end(), bv.begin() + i2, bv.end() );
  }


bool diff_member( const long long mpos, const long long msize,
                  const std::vector< int > & infd_vector,
                  std::vector< Block > & block_vector )
  {
  const int files = infd_vector.size();
  const int buffer_size = 65536;
  uint8_t * const buffer1 = new uint8_t[buffer_size];
  uint8_t * const buffer2 = new uint8_t[buffer_size];

  bool error = false;
  for( int i1 = 0; i1 + 1 < files && !error; ++i1 )
    {
    for( int i2 = i1 + 1; i2 < files && !error; ++i2 )
      {
      std::vector< Block > bv;
      long long partial_pos = 0;
      const int fd1 = infd_vector[i1], fd2 = infd_vector[i2];
      int begin = -1;			// begin of block. -1 means no block
      bool prev_equal = true;
      if( !safe_seek( fd1, mpos ) || !safe_seek( fd2, mpos ) )
        { error = true; break; }

      while( msize > partial_pos )
        {
        const int size = std::min( (long long)buffer_size, msize - partial_pos );
        const int rd = readblock( fd1, buffer1, size );
        if( rd != size && errno )
          { show_error( "Error reading input file", errno ); error = true; break; }
        if( rd > 0 )
          {
          if( readblock( fd2, buffer2, rd ) != rd )
            { show_error( "Error reading input file", errno );
              error = true; break; }
          for( int i = 0; i < rd; ++i )
            {
            if( buffer1[i] != buffer2[i] )
              {
              prev_equal = false;
              if( begin < 0 ) begin = partial_pos + i;	// begin block
              }
            else if( !prev_equal ) prev_equal = true;
            else if( begin >= 0 )			// end block
              {
              Block b( mpos + begin, partial_pos + i - 1 - begin );
              begin = -1;
              bv.push_back( b );
              }
            }
          partial_pos += rd;
          }
        if( rd < buffer_size ) break;			// EOF
        }
      if( begin >= 0 )					// finish last block
        {
        Block b( mpos + begin, partial_pos - prev_equal - begin );
        bv.push_back( b );
        }
      combine( block_vector, bv );
      }
    }
  delete[] buffer2; delete[] buffer1;
  return !error;
  }


int ipow( const unsigned base, const unsigned exponent )
  {
  unsigned result = 1;
  for( unsigned i = 0; i < exponent; ++i )
    {
    if( INT_MAX / base >= result ) result *= base;
    else { result = INT_MAX; break; }
    }
  return result;
  }


int open_input_files( const std::vector< std::string > & filenames,
                      std::vector< int > & infd_vector,
                      File_index & file_index, const int verbosity )
  {
  const int files = filenames.size();
  bool identical = false;
  for( int i = 1; i < files; ++i )
    if( filenames[0] == filenames[i] )
      { identical = true; break; }
  if( !identical )
    for( int i = 0; i < files; ++i )
      {
      struct stat in_stats;
      ino_t st_ino0 = 0;
      dev_t st_dev0 = 0;
      infd_vector[i] = open_instream( filenames[i].c_str(), &in_stats, true, true );
      if( infd_vector[i] < 0 ) return 1;
      if( i == 0 ) { st_ino0 = in_stats.st_ino; st_dev0 = in_stats.st_dev; }
      else if( st_ino0 == in_stats.st_ino && st_dev0 == in_stats.st_dev )
        { identical = true; break; }
      }
  if( identical ) { show_error( "Two input files are the same." ); return 2; }

  long long isize = 0;
  for( int i = 0; i < files; ++i )
    {
    long long tmp;
    const File_index fi( infd_vector[i] );
    if( fi.retval() == 0 )		// file format is intact
      {
      if( file_index.retval() != 0 ) file_index = fi;
      else if( file_index != fi )
        { show_error( "Input files are different." ); return 2; }
      tmp = file_index.file_size();
      }
    else				// file format is damaged
      {
      tmp = lseek( infd_vector[i], 0, SEEK_END );
      if( tmp < 0 )
        {
        show_error2( "Input file", filenames[i].c_str(), "is not seekable." );
        return 1;
        }
      }
    if( i == 0 )
      {
      isize = tmp;
      if( isize < min_member_size )
        { show_error2( "Input file", filenames[i].c_str(), "is too short." );
          return 2; }
      }
    else if( isize != tmp )
      { show_error( "Sizes of input files are different." ); return 2; }
    }

  if( file_index.retval() != 0 )
    {
    const File_index fi( infd_vector, isize );
    if( fi.retval() == 0 )		// file format could be recovered
      file_index = fi;
    else
      { show_error( "Format damaged in all input files." ); return 2; }
    }

  for( int i = 0; i < files; ++i )
    {
    const int infd = infd_vector[i];
    bool error = false;
    for( int j = 0; j < file_index.members(); ++j )
      {
      const long long mpos = file_index.mblock( j ).pos();
      const long long msize = file_index.mblock( j ).size();
      if( !safe_seek( infd, mpos ) ) return 1;
      if( !try_decompress_member( infd, msize ) ) { error = true; break; }
      }
    if( !error )
      {
      if( verbosity >= 1 )
        std::printf( "File '%s' has no errors. Recovery is not needed.\n",
                     filenames[i].c_str() );
      return 0;
      }
    }
  return -1;
  }

} // end namespace


void cleanup_and_fail( const std::string & output_filename,
                       const int outfd, const int retval )
  {
  if( outfd >= 0 ) close( outfd );
  if( std::remove( output_filename.c_str() ) != 0 && errno != ENOENT )
    show_error( "WARNING: deletion of output file (apparently) failed." );
  std::exit( retval );
  }


// max_size < 0 means no size limit.
bool copy_file( const int infd, const int outfd, const long long max_size )
  {
  const int buffer_size = 65536;
  // remaining number of bytes to copy
  long long rest = ( ( max_size >= 0 ) ? max_size : buffer_size );
  uint8_t * const buffer = new uint8_t[buffer_size];
  bool error = false;

  while( rest > 0 )
    {
    const int size = std::min( (long long)buffer_size, rest );
    if( max_size >= 0 ) rest -= size;
    const int rd = readblock( infd, buffer, size );
    if( rd != size && errno )
      { show_error( "Error reading input file", errno ); error = true; break; }
    if( rd > 0 )
      {
      const int wr = writeblock( outfd, buffer, rd );
      if( wr != rd )
        { show_error( "Error writing output file", errno );
          error = true; break; }
      }
    if( rd < size ) break;				// EOF
    }
  delete[] buffer;
  return !error;
  }


bool try_decompress_member( const int fd, const unsigned long long msize,
                            long long * failure_posp )
  {
  try {
    Range_decoder rdec( fd );
    File_header header;
    rdec.read_data( header.data, File_header::size );
    if( !rdec.finished() &&			// End Of File
        header.verify_magic() && header.verify_version() &&
        header.dictionary_size() >= min_dictionary_size &&
        header.dictionary_size() <= max_dictionary_size )
      {
      LZ_decoder decoder( header, rdec, -1 );
      Pretty_print dummy( "", -1 );

      if( decoder.decode_member( dummy ) == 0 &&
          rdec.member_position() == msize ) return true;
      if( failure_posp ) *failure_posp = rdec.member_position();
      }
    }
  catch( std::bad_alloc )
    { show_error( "Not enough memory." ); std::exit( 1 ); }
  catch( Error e ) {}
  return false;
  }


int merge_files( const std::vector< std::string > & filenames,
                 const std::string & output_filename, const int verbosity,
                 const bool force )
  {
  const int files = filenames.size();
  std::vector< int > infd_vector( files );
  File_index file_index;
  const int retval =
    open_input_files( filenames, infd_vector, file_index, verbosity );
  if( retval >= 0 ) return retval;
  if( !safe_seek( infd_vector[0], 0 ) ) return 1;

  const int outfd = open_outstream_rw( output_filename, force );
  if( outfd < 0 ) return 1;
  if( !copy_file( infd_vector[0], outfd ) )		// copy whole file
    cleanup_and_fail( output_filename, outfd, 1 );

  for( int j = 0; j < file_index.members(); ++j )
    {
    const long long mpos = file_index.mblock( j ).pos();
    const long long msize = file_index.mblock( j ).size();
    // vector of data blocks differing among the copies of the current member
    std::vector< Block > block_vector;
    if( !diff_member( mpos, msize, infd_vector, block_vector ) ||
        !safe_seek( outfd, mpos ) )
      cleanup_and_fail( output_filename, outfd, 1 );

    if( block_vector.empty() )
      {
      if( file_index.members() > 1 && try_decompress_member( outfd, msize ) )
        continue;
      show_error( "Input files are (partially) identical. Recovery is not possible." );
      cleanup_and_fail( output_filename, outfd, 2 );
      }

    const int size0 = block_vector[0].size();
    const bool single_block = ( block_vector.size() == 1 );
    if( ipow( files, block_vector.size() ) >= INT_MAX ||
        ( single_block && ipow( files, 2 ) >= INT_MAX / size0 ) )
      { show_error( "Input files are too damaged. Recovery is not possible." );
        cleanup_and_fail( output_filename, outfd, 2 ); }

    const int shifts = ( single_block && size0 > 1 ) ? size0 - 1 : 1;
    if( single_block && size0 > 1 )
      {
      Block b( block_vector[0].pos() + 1, size0 - 1 );
      block_vector[0].size( 1 );
      block_vector.push_back( b );
      }

    if( verbosity >= 1 && file_index.members() > 1 )
      {
      std::printf( "Merging member %d\n", j + 1 );
      std::fflush( stdout );
      }
    const int base_variations = ipow( files, block_vector.size() );
    const int variations = base_variations * shifts;
    bool done = false;
    for( int var = 0; var < variations; ++var )
      {
      if( verbosity >= 1 )
        {
        std::printf( "Trying variation %d of %d \r", var + 1, variations );
        std::fflush( stdout );
        }
      int tmp = var;
      for( unsigned i = 0; i < block_vector.size(); ++i )
        {
        const int infd = infd_vector[tmp % files];
        tmp /= files;
        if( lseek( infd, block_vector[i].pos(), SEEK_SET ) < 0 ||
            lseek( outfd, block_vector[i].pos(), SEEK_SET ) < 0 ||
            !copy_file( infd, outfd, block_vector[i].size() ) )
          { show_error( "Error reading output file", errno );
            cleanup_and_fail( output_filename, outfd, 1 ); }
        }
      if( !safe_seek( outfd, mpos ) )
        cleanup_and_fail( output_filename, outfd, 1 );
      if( try_decompress_member( outfd, msize ) )
        { done = true; break; }
      if( var > 0 && var % base_variations == 0 )
        block_vector[0].shift( block_vector[1] );
      }
    if( verbosity >= 1 ) std::printf( "\n" );
    if( !done )
      {
      if( verbosity >= 2 )
        for( unsigned i = 0; i < block_vector.size(); ++i )
          std::fprintf( stderr, "area %2d from offset %6lld to %6lld\n", i + 1,
                        block_vector[i].pos(), block_vector[i].end() - 1 );
      show_error( "Some error areas overlap. Can't recover input file." );
      cleanup_and_fail( output_filename, outfd, 2 );
      }
    }

  if( close( outfd ) != 0 )
    {
    show_error( "Error closing output file", errno );
    cleanup_and_fail( output_filename, -1, 1 );
    }
  if( verbosity >= 1 )
    std::printf( "Input files merged successfully.\n" );
  return 0;
  }