Merging upstream version 1.9.

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

1 files changed, 268 insertions, 0 deletions

diff --git a/file_index.c b/file_index.c
new file mode 100644
index 0000000..e737608
--- /dev/null
+++ b/file_index.c
@@ -0,0 +1,268 @@
+/*  Lunzip - Decompressor for the lzip format
+    Copyright (C) 2010-2017 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 <errno.h>
+#include <stdbool.h>
+#include <stdio.h>
+#include <string.h>
+#include <stdint.h>
+#include <stdlib.h>
+#include <unistd.h>
+
+#include "lzip.h"
+#include "file_index.h"
+
+
+static 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;
+  }
+
+
+static bool add_error( struct File_index * const fi, const char * const msg )
+  {
+  const int len = strlen( msg );
+  void * tmp = resize_buffer( fi->error, fi->error_size + len + 1 );
+  if( !tmp ) return false;
+  fi->error = (char *)tmp;
+  strncpy( fi->error + fi->error_size, msg, len + 1 );
+  fi->error_size += len;
+  return true;
+  }
+
+
+static bool push_back_member( struct File_index * const fi,
+                              const long long dp, const long long ds,
+                              const long long mp, const long long ms,
+                              const unsigned dict_size )
+  {
+  struct Member * p;
+  void * tmp = resize_buffer( fi->member_vector,
+               ( fi->members + 1 ) * sizeof fi->member_vector[0] );
+  if( !tmp )
+    { add_error( fi, "Not enough memory." ); fi->retval = 1; return false; }
+  fi->member_vector = (struct Member *)tmp;
+  p = &(fi->member_vector[fi->members]);
+  init_member( p, dp, ds, mp, ms, dict_size );
+  ++fi->members;
+  return true;
+  }
+
+
+static void Fi_free_member_vector( struct File_index * const fi )
+  {
+  if( fi->member_vector )
+    { free( fi->member_vector ); fi->member_vector = 0; }
+  fi->members = 0;
+  }
+
+
+static void Fi_reverse_member_vector( struct File_index * const fi )
+  {
+  struct Member tmp;
+  long i;
+  for( i = 0; i < fi->members / 2; ++i )
+    {
+    tmp = fi->member_vector[i];
+    fi->member_vector[i] = fi->member_vector[fi->members-i-1];
+    fi->member_vector[fi->members-i-1] = tmp;
+    }
+  }
+
+
+static void Fi_set_errno_error( struct File_index * const fi,
+                                const char * const msg )
+  {
+  add_error( fi, msg ); add_error( fi, strerror( errno ) );
+  fi->retval = 1;
+  }
+
+static void Fi_set_num_error( struct File_index * const fi,
+                              const char * const msg, unsigned long long num )
+  {
+  char buf[80];
+  snprintf( buf, sizeof buf, "%s%llu", msg, num );
+  add_error( fi, buf );
+  fi->retval = 2;
+  }
+
+
+/* If successful, push last member and set pos to member header. */
+static bool Fi_skip_trailing_data( struct File_index * const fi,
+                                   const int fd, long long * const pos )
+  {
+  enum { block_size = 16384,
+         buffer_size = block_size + Ft_size - 1 + Fh_size };
+  uint8_t buffer[buffer_size];
+  int bsize = *pos % block_size;		/* total bytes in buffer */
+  int search_size, rd_size;
+  unsigned long long ipos;
+  int i;
+  if( bsize <= buffer_size - block_size ) bsize += block_size;
+  search_size = bsize;				/* bytes to search for trailer */
+  rd_size = bsize;				/* bytes to read from file */
+  ipos = *pos - rd_size;			/* aligned to block_size */
+  if( *pos < min_member_size ) return false;
+
+  while( true )
+    {
+    const uint8_t max_msb = ( ipos + search_size ) >> 56;
+    if( seek_read( fd, buffer, rd_size, ipos ) != rd_size )
+      { Fi_set_errno_error( fi, "Error seeking member trailer: " );
+        return false; }
+    for( i = search_size; i >= Ft_size; --i )
+      if( buffer[i-1] <= max_msb )	/* most significant byte of member_size */
+        {
+        File_header header;
+        File_trailer * trailer = (File_trailer *)( buffer + i - Ft_size );
+        const unsigned long long member_size = Ft_get_member_size( *trailer );
+        unsigned dictionary_size;
+        if( member_size == 0 )
+          { while( i > Ft_size && buffer[i-9] == 0 ) --i; continue; }
+        if( member_size < min_member_size || member_size > ipos + i )
+          continue;
+        if( seek_read( fd, header, Fh_size,
+                       ipos + i - member_size ) != Fh_size )
+          { Fi_set_errno_error( fi, "Error reading member header: " );
+            return false; }
+        dictionary_size = Fh_get_dictionary_size( header );
+        if( !Fh_verify_magic( header ) || !Fh_verify_version( header ) ||
+            !isvalid_ds( dictionary_size ) ) continue;
+        if( Fh_verify_prefix( buffer + i, bsize - i ) )
+          {
+          add_error( fi, "Last member in input file is truncated or corrupt." );
+          fi->retval = 2; return false;
+          }
+        *pos = ipos + i - member_size;
+        return push_back_member( fi, 0, Ft_get_data_size( *trailer ), *pos,
+                                 member_size, dictionary_size );
+        }
+    if( ipos <= 0 )
+      { Fi_set_num_error( fi, "Member size in trailer is corrupt at pos ",
+                          *pos - 8 );
+        return false; }
+    bsize = buffer_size;
+    search_size = bsize - Fh_size;
+    rd_size = block_size;
+    ipos -= rd_size;
+    memcpy( buffer + rd_size, buffer, buffer_size - rd_size );
+    }
+  }
+
+
+bool Fi_init( struct File_index * const fi, const int infd,
+              const bool ignore_trailing )
+  {
+  File_header header;
+  long long pos;
+  long i;
+  fi->member_vector = 0;
+  fi->error = 0;
+  fi->isize = lseek( infd, 0, SEEK_END );
+  fi->members = 0;
+  fi->error_size = 0;
+  fi->retval = 0;
+  if( fi->isize < 0 )
+    { Fi_set_errno_error( fi, "Input file is not seekable: " ); return false; }
+  if( fi->isize < min_member_size )
+    { add_error( fi, "Input file is too short." ); fi->retval = 2;
+      return false; }
+  if( fi->isize > INT64_MAX )
+    { add_error( fi, "Input file is too long (2^63 bytes or more)." );
+      fi->retval = 2; return false; }
+
+  if( seek_read( infd, header, Fh_size, 0 ) != Fh_size )
+    { Fi_set_errno_error( fi, "Error reading member header: " ); return false; }
+  if( !Fh_verify_magic( header ) )
+    { add_error( fi, bad_magic_msg ); fi->retval = 2; return false; }
+  if( !Fh_verify_version( header ) )
+    { add_error( fi, bad_version( Fh_version( header ) ) ); fi->retval = 2;
+      return false; }
+  if( !isvalid_ds( Fh_get_dictionary_size( header ) ) )
+    { add_error( fi, bad_dict_msg ); fi->retval = 2; return false; }
+
+  pos = fi->isize;		/* always points to a header or to EOF */
+  while( pos >= min_member_size )
+    {
+    File_trailer trailer;
+    unsigned long long member_size;
+    unsigned dictionary_size;
+    if( seek_read( infd, trailer, Ft_size, pos - Ft_size ) != Ft_size )
+      { Fi_set_errno_error( fi, "Error reading member trailer: " ); break; }
+    member_size = Ft_get_member_size( trailer );
+    if( member_size < min_member_size || member_size > (unsigned long long)pos )
+      {
+      if( fi->members > 0 )
+        Fi_set_num_error( fi, "Member size in trailer is corrupt at pos ",
+                          pos - 8 );
+      else if( Fi_skip_trailing_data( fi, infd, &pos ) )
+        { if( ignore_trailing ) continue;
+          add_error( fi, trailing_msg ); fi->retval = 2; return false; }
+      break;
+      }
+    if( seek_read( infd, header, Fh_size, pos - member_size ) != Fh_size )
+      { Fi_set_errno_error( fi, "Error reading member header: " ); break; }
+    dictionary_size = Fh_get_dictionary_size( header );
+    if( !Fh_verify_magic( header ) || !Fh_verify_version( header ) ||
+        !isvalid_ds( dictionary_size ) )
+      {
+      if( fi->members > 0 )
+        Fi_set_num_error( fi, "Bad header at pos ", pos - member_size );
+      else if( Fi_skip_trailing_data( fi, infd, &pos ) )
+        { if( ignore_trailing ) continue;
+          add_error( fi, trailing_msg ); fi->retval = 2; return false; }
+      break;
+      }
+    pos -= member_size;
+    if( !push_back_member( fi, 0, Ft_get_data_size( trailer ), pos,
+                           member_size, dictionary_size ) )
+      return false;
+    }
+  if( pos != 0 || fi->members <= 0 )
+    {
+    Fi_free_member_vector( fi );
+    if( fi->retval == 0 )
+      { add_error( fi, "Can't create file index." ); fi->retval = 2; }
+    return false;
+    }
+  Fi_reverse_member_vector( fi );
+  for( i = 0; i < fi->members - 1; ++i )
+    {
+    const long long end = block_end( fi->member_vector[i].dblock );
+    if( end < 0 || end > INT64_MAX )
+      {
+      Fi_free_member_vector( fi );
+      add_error( fi, "Data in input file is too long (2^63 bytes or more)." );
+      fi->retval = 2; return false;
+      }
+    fi->member_vector[i+1].dblock.pos = end;
+    }
+  return true;
+  }
+
+
+void Fi_free( struct File_index * const fi )
+  {
+  Fi_free_member_vector( fi );
+  if( fi->error ) { free( fi->error ); fi->error = 0; }
+  fi->error_size = 0;
+  }