summaryrefslogtreecommitdiffstats
path: root/lzip_index.cc
blob: 66eb30d333571097d7bd062cae86ee2bc8a37db9 (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
/* Lziprecover - Data recovery tool for the lzip format
   Copyright (C) 2009-2021 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 2 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 <algorithm>
#include <cerrno>
#include <cstdio>
#include <cstring>
#include <string>
#include <vector>
#include <stdint.h>
#include <unistd.h>

#include "lzip.h"
#include "lzip_index.h"


int seek_read( const int fd, uint8_t * const buf, const int size,
               const long long pos )
  {
  if( lseek( fd, pos, SEEK_SET ) == pos )
    return readblock( fd, buf, size );
  return 0;
  }


bool Lzip_index::check_header_error( const Lzip_header & header,
                                     const bool ignore_bad_ds )
  {
  if( !header.verify_magic() )
    { error_ = bad_magic_msg; retval_ = 2; return true; }
  if( !header.verify_version() )
    { error_ = bad_version( header.version() ); retval_ = 2; return true; }
  if( !ignore_bad_ds && !isvalid_ds( header.dictionary_size() ) )
    { error_ = bad_dict_msg; retval_ = 2; return true; }
  return false;
  }

void Lzip_index::set_errno_error( const char * const msg )
  {
  error_ = msg; error_ += std::strerror( errno );
  retval_ = 1;
  }

void Lzip_index::set_num_error( const char * const msg, unsigned long long num )
  {
  char buf[80];
  snprintf( buf, sizeof buf, "%s%llu", msg, num );
  error_ = buf;
  retval_ = 2;
  }


bool Lzip_index::read_header( const int fd, Lzip_header & header,
                              const long long pos )
  {
  if( seek_read( fd, header.data, Lzip_header::size, pos ) != Lzip_header::size )
    { set_errno_error( "Error reading member header: " ); return false; }
  return true;
  }

bool Lzip_index::read_trailer( const int fd, Lzip_trailer & trailer,
                               const long long pos )
  {
  if( seek_read( fd, trailer.data, Lzip_trailer::size,
                 pos - Lzip_trailer::size ) != Lzip_trailer::size )
    { set_errno_error( "Error reading member trailer: " ); return false; }
  return true;
  }


/* Skip backwards the gap or trailing data ending at pos.
   'ignore_gaps' also ignores format errors and a truncated last member.
   If successful, push member preceding gap and set pos to member header. */
bool Lzip_index::skip_gap( const int fd, unsigned long long & pos,
                   const bool ignore_trailing, const bool loose_trailing,
                   const bool ignore_bad_ds, const bool ignore_gaps )
  {
  if( pos < min_member_size )
    {
    if( ignore_gaps && !member_vector.empty() ) { pos = 0; return true; }
    return false;
    }
  enum { block_size = 16384,
         buffer_size = block_size + Lzip_trailer::size - 1 + Lzip_header::size };
  uint8_t buffer[buffer_size];
  int bsize = pos % block_size;			// total bytes in buffer
  if( bsize <= buffer_size - block_size ) bsize += block_size;
  int search_size = bsize;			// bytes to search for trailer
  int rd_size = bsize;				// bytes to read from file
  unsigned long long ipos = pos - rd_size;	// aligned to block_size

  while( true )
    {
    if( seek_read( fd, buffer, rd_size, ipos ) != rd_size )
      { set_errno_error( "Error seeking member trailer: " ); return false; }
    const uint8_t max_msb = ( ipos + search_size ) >> 56;
    for( int i = search_size; i >= Lzip_trailer::size; --i )
      if( buffer[i-1] <= max_msb )	// most significant byte of member_size
        {
        const Lzip_trailer & trailer =
          *(const Lzip_trailer *)( buffer + i - Lzip_trailer::size );
        const unsigned long long member_size = trailer.member_size();
        if( member_size == 0 )			// skip trailing zeros
          { while( i > Lzip_trailer::size && buffer[i-9] == 0 ) --i; continue; }
        if( member_size > ipos + i || !trailer.verify_consistency() )
          continue;
        Lzip_header header;
        if( !read_header( fd, header, ipos + i - member_size ) ) return false;
        if( !header.verify( ignore_bad_ds ) ) continue;
        const Lzip_header & header2 = *(const Lzip_header *)( buffer + i );
        const bool full_h2 = bsize - i >= Lzip_header::size;
        if( header2.verify_prefix( bsize - i ) )	// next header
          {
          if( !ignore_gaps && member_vector.empty() )	// last member
            {
            if( !full_h2 ) error_ = "Last member in input file is truncated.";
            else if( !check_header_error( header2, ignore_bad_ds ) )
              error_ = "Last member in input file is truncated or corrupt.";
            retval_ = 2; return false;
            }
          const unsigned dictionary_size =
                         full_h2 ? header2.dictionary_size() : 0;
          const unsigned long long member_size = pos - ( ipos + i );
          pos = ipos + i;
          // approximate data and member sizes for '-i -D'
          member_vector.push_back( Member( 0, member_size, pos,
                                           member_size, dictionary_size ) );
          }
        if( !ignore_gaps && member_vector.empty() )
          {
          if( !loose_trailing && full_h2 && header2.verify_corrupt() )
            { error_ = corrupt_mm_msg; retval_ = 2; return false; }
          if( !ignore_trailing )
            { error_ = trailing_msg; retval_ = 2; return false; }
          }
        pos = ipos + i - member_size;
        const unsigned dictionary_size = header.dictionary_size();
        member_vector.push_back( Member( 0, trailer.data_size(), pos,
                                         member_size, dictionary_size ) );
        if( dictionary_size_ < dictionary_size )
          dictionary_size_ = dictionary_size;
        return true;
        }
    if( ipos == 0 )
      {
      if( ignore_gaps && !member_vector.empty() )
        {
        const Lzip_header * header = (const Lzip_header *)buffer;
        const unsigned dictionary_size = header->dictionary_size();
        // approximate data and member sizes for '-i -D'
        member_vector.push_back( Member( 0, pos, 0, pos, dictionary_size ) );
        pos = 0; return true;
        }
      set_num_error( "Bad trailer at pos ", pos - Lzip_trailer::size );
      return false;
      }
    bsize = buffer_size;
    search_size = bsize - Lzip_header::size;
    rd_size = block_size;
    ipos -= rd_size;
    std::memcpy( buffer + rd_size, buffer, buffer_size - rd_size );
    }
  }


Lzip_index::Lzip_index( const int infd, const bool ignore_trailing,
                        const bool loose_trailing, const bool ignore_bad_ds,
                        const bool ignore_gaps, const long long max_pos )
  : insize( lseek( infd, 0, SEEK_END ) ), retval_( 0 ), dictionary_size_( 0 )
  {
  if( insize < 0 )
    { set_errno_error( "Input file is not seekable: " ); return; }
  if( insize < min_member_size )
    { error_ = "Input file is too short."; retval_ = 2; return; }
  if( insize > INT64_MAX )
    { error_ = "Input file is too long (2^63 bytes or more).";
      retval_ = 2; return; }

  Lzip_header header;
  if( !read_header( infd, header, 0 ) ) return;
  if( check_header_error( header, ignore_bad_ds ) ) return;

  // pos always points to a header or to ( EOF || max_pos )
  unsigned long long pos = ( max_pos > 0 ) ? max_pos : insize;
  while( pos >= min_member_size )
    {
    Lzip_trailer trailer;
    if( !read_trailer( infd, trailer, pos ) ) break;
    const unsigned long long member_size = trailer.member_size();
    // if gaps are being ignored, verify consistency of last trailer only.
    if( member_size > pos || member_size < min_member_size ||
        ( ( !ignore_gaps || member_vector.empty() ) &&
        !trailer.verify_consistency() ) )		// bad trailer
      {
      if( ignore_gaps || member_vector.empty() )
        { if( skip_gap( infd, pos, ignore_trailing, loose_trailing,
              ignore_bad_ds, ignore_gaps ) ) continue; else return; }
      set_num_error( "Bad trailer at pos ", pos - Lzip_trailer::size );
      break;
      }
    if( !read_header( infd, header, pos - member_size ) ) break;
    if( !header.verify( ignore_bad_ds ) )		// bad header
      {
      if( ignore_gaps || member_vector.empty() )
        { if( skip_gap( infd, pos, ignore_trailing, loose_trailing,
              ignore_bad_ds, ignore_gaps ) ) continue; else return; }
      set_num_error( "Bad header at pos ", pos - member_size );
      break;
      }
    pos -= member_size;
    const unsigned dictionary_size = header.dictionary_size();
    member_vector.push_back( Member( 0, trailer.data_size(), pos,
                                     member_size, dictionary_size ) );
    if( dictionary_size_ < dictionary_size )
      dictionary_size_ = dictionary_size;
    }
  // block at pos == 0 must be a member unless shorter than min_member_size
  if( pos >= min_member_size || ( pos != 0 && !ignore_gaps ) ||
      member_vector.empty() )
    {
    member_vector.clear();
    if( retval_ == 0 ) { error_ = "Can't create file index."; retval_ = 2; }
    return;
    }
  std::reverse( member_vector.begin(), member_vector.end() );
  for( unsigned long i = 0; ; ++i )
    {
    const long long end = member_vector[i].dblock.end();
    if( end < 0 || end > INT64_MAX )
      {
      member_vector.clear();
      error_ = "Data in input file is too long (2^63 bytes or more).";
      retval_ = 2; return;
      }
    if( i + 1 >= member_vector.size() ) break;
    member_vector[i+1].dblock.pos( end );
    if( member_vector[i].mblock.end() > member_vector[i+1].mblock.pos() )
      internal_error( "two mblocks overlap after constructing a Lzip_index." );
    }
  }


// All files in 'infd_vector' must be at least 'fsize' bytes long.
Lzip_index::Lzip_index( const std::vector< int > & infd_vector,
                        const long long fsize )
  : insize( fsize ), retval_( 0 ), dictionary_size_( 0 )	// DS not used
  {
  if( insize < 0 )
    { set_errno_error( "Input file is not seekable: " ); return; }
  if( insize < min_member_size )
    { error_ = "Input file is too short."; retval_ = 2; return; }
  if( insize > INT64_MAX )
    { error_ = "Input file is too long (2^63 bytes or more).";
      retval_ = 2; return; }

  const int files = infd_vector.size();
  Lzip_header header;
  bool done = false;
  for( int i = 0; i < files && !done; ++i )
    {
    const int infd = infd_vector[i];
    if( !read_header( infd, header, 0 ) ) return;
    if( header.verify_magic() && header.verify_version() ) done = true;
    }
  if( !done )
    { error_ = bad_magic_msg; retval_ = 2; return; }

  long long pos = insize;		// always points to a header or to EOF
  while( pos >= min_member_size )
    {
    unsigned long long member_size;
    Lzip_trailer trailer;
    done = false;
    for( int it = 0; it < files && !done; ++it )
      {
      const int tfd = infd_vector[it];
      if( !read_trailer( tfd, trailer, pos ) ) goto error;
      member_size = trailer.member_size();
      if( member_size <= (unsigned long long)pos && trailer.verify_consistency() )
        for( int ih = 0; ih < files && !done; ++ih )
          {
          const int hfd = infd_vector[ih];
          if( !read_header( hfd, header, pos - member_size ) ) goto error;
          if( header.verify_magic() && header.verify_version() ) done = true;
          }
      }
    if( !done )
      {
      if( member_vector.empty() ) { --pos; continue; }	// maybe trailing data
      set_num_error( "Member size in trailer may be corrupt at pos ", pos - 8 );
      break;
      }
    if( member_vector.empty() && insize > pos )
      {
      const int size = std::min( (long long)Lzip_header::size, insize - pos );
      for( int i = 0; i < files; ++i )
        {
        const int infd = infd_vector[i];
        if( seek_read( infd, header.data, size, pos ) == size &&
            header.verify_prefix( size ) )
          {
          error_ = "Last member in input file is truncated or corrupt.";
          retval_ = 2; goto error;
          }
        }
      }
    pos -= member_size;
    member_vector.push_back( Member( 0, trailer.data_size(), pos,
                                     member_size, 0 ) );
    }
error:
  if( pos != 0 || member_vector.empty() )
    {
    member_vector.clear();
    if( retval_ == 0 ) { error_ = "Can't create file index."; retval_ = 2; }
    return;
    }
  std::reverse( member_vector.begin(), member_vector.end() );
  for( unsigned long i = 0; ; ++i )
    {
    const long long end = member_vector[i].dblock.end();
    if( end < 0 || end > INT64_MAX )
      {
      member_vector.clear();
      error_ = "Data in input file is too long (2^63 bytes or more).";
      retval_ = 2; return;
      }
    if( i + 1 >= member_vector.size() ) break;
    member_vector[i+1].dblock.pos( end );
    }
  }


// Returns members + gaps [+ trailing data].
long Lzip_index::blocks( const bool count_tdata ) const
  {
  long n = member_vector.size() + ( count_tdata && cdata_size() < file_size() );
  if( member_vector.size() && member_vector[0].mblock.pos() > 0 ) ++n;
  for( unsigned long i = 1; i < member_vector.size(); ++i )
    if( member_vector[i-1].mblock.end() < member_vector[i].mblock.pos() ) ++n;
  return n;
  }