Merging upstream version 1.18.

Signed-off-by: Daniel Baumann <mail@daniel-baumann.ch>

1 files changed, 170 insertions, 112 deletions

diff --git a/main.cc b/main.cc
index 7845d90..d3ef70c 100644
--- a/main.cc
+++ b/main.cc
@@ -1,5 +1,5 @@
 /*  Lziprecover - Data recovery tool for the lzip format
-    Copyright (C) 2009-2015 Antonio Diaz Diaz.
+    Copyright (C) 2009-2016 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
@@ -66,12 +66,14 @@
 #error "Environments where CHAR_BIT != 8 are not supported."
 #endif
 
+std::string output_filename;		// global vars for output file
+int outfd = -1;
 
 namespace {
 
 const char * const Program_name = "Lziprecover";
 const char * const program_name = "lziprecover";
-const char * const program_year = "2015";
+const char * const program_year = "2016";
 const char * invocation_name = 0;
 
 struct { const char * from; const char * to; } const known_extensions[] = {
@@ -79,15 +81,11 @@ struct { const char * from; const char * to; } const known_extensions[] = {
   { ".tlz", ".tar" },
   { 0,      0      } };
 
-enum Mode { m_none, m_debug_delay, m_debug_repair, m_decompress, m_list,
-            m_merge, m_range_dec, m_repair, m_show_packets, m_split, m_test };
+enum Mode { m_none, m_alone_to_lz, m_debug_decompress, m_debug_delay,
+            m_debug_repair, m_decompress, m_list, m_merge, m_range_dec,
+            m_repair, m_show_packets, m_split, m_test };
 
-std::string output_filename;
-int outfd = -1;
 int verbosity = 0;
-const mode_t usr_rw = S_IRUSR | S_IWUSR;
-const mode_t all_rw = usr_rw | S_IRGRP | S_IWGRP | S_IROTH | S_IWOTH;
-mode_t outfd_mode = usr_rw;
 bool delete_output_on_interrupt = false;
 
 
@@ -98,15 +96,18 @@ void show_help()
                "single-byte error per member), without the need of any extra redundance\n"
                "at all. Losing an entire archive just because of a corrupt byte near the\n"
                "beginning is a thing of the past.\n"
-               "Lziprecover can also produce a correct file by merging the good parts of\n"
+               "\nLziprecover can also produce a correct file by merging the good parts of\n"
                "two or more damaged copies, extract data from damaged files, decompress\n"
                "files and test integrity of files.\n"
+               "\nLziprecover is not a replacement for regular backups, but a last line of\n"
+               "defense for the case where the backups are also damaged.\n"
                "\nUsage: %s [options] [files]\n", invocation_name );
   std::printf( "\nOptions:\n"
                "  -h, --help                      display this help and exit\n"
                "  -V, --version                   output version information and exit\n"
                "  -a, --trailing-error            exit with error status if trailing data\n"
-               "  -c, --stdout                    send decompressed output to standard output\n"
+               "  -A, --alone-to-lz               convert lzma-alone files to lzip format\n"
+               "  -c, --stdout                    write to standard output, keep input files\n"
                "  -d, --decompress                decompress\n"
                "  -D, --range-decompress=<range>  decompress a range of bytes (N-M) to stdout\n"
                "  -f, --force                     overwrite existing output files\n"
@@ -117,16 +118,19 @@ void show_help()
                "  -o, --output=<file>             place the output into <file>\n"
                "  -q, --quiet                     suppress all messages\n"
                "  -R, --repair                    try to repair a small error in file\n"
-               "  -s, --split                     split multi-member file in single-member files\n"
+               "  -s, --split                     split multimember file in single-member files\n"
                "  -t, --test                      test compressed file integrity\n"
                "  -v, --verbose                   be verbose (a 2nd -v gives more)\n" );
   if( verbosity >= 1 )
     {
-    std::printf( "  -x, --show-packets[=<pos>,<val>]  show in stdout the decoded LZMA packets\n"
-                 "  -y, --debug-delay=<range>         find max error detection delay in <range>\n"
-                 "  -z, --debug-repair=<pos>,<val>    test repair one-byte error at <pos>\n" );
+    std::printf( "  -W, --debug-decompress=<pos>,<val>  set pos to val and decompress to stdout\n"
+                 "  -X, --show-packets[=<pos>,<val>]  show in stdout the decoded LZMA packets\n"
+                 "  -Y, --debug-delay=<range>         find max error detection delay in <range>\n"
+                 "  -Z, --debug-repair=<pos>,<val>    test repair one-byte error at <pos>\n" );
     }
-  std::printf( "Numbers may be followed by a multiplier: k = kB = 10^3 = 1000,\n"
+  std::printf( "If no file names are given, or if a file is '-', lziprecover decompresses\n"
+               "from standard input to standard output.\n"
+               "Numbers may be followed by a multiplier: k = kB = 10^3 = 1000,\n"
                "Ki = KiB = 2^10 = 1024, M = 10^6, Mi = 2^20, G = 10^9, Gi = 2^30, etc...\n"
                "\nExit status: 0 for a normal exit, 1 for environmental problems (file\n"
                "not found, invalid flags, I/O errors, etc), 2 to indicate a corrupt or\n"
@@ -211,14 +215,15 @@ int parse_long_long( const char * const ptr, long long & value )
   }
 
 
-// Recognized formats: <begin> <begin>-<end> <begin>,<size>
+// Recognized formats: <begin> <begin>-<end> <begin>,<size> ,<size>
 //
 void parse_range( const char * const ptr, Block & range )
   {
   long long value = 0;
-  int c = parse_long_long( ptr, value );		// pos
-  if( c && value >= 0 && value < INT64_MAX &&
-      ( ptr[c] == 0 || ptr[c] == ',' || ptr[c] == '-' ) )
+  const bool size_only = ( ptr[0] == ',' );
+  int c = size_only ? 0 : parse_long_long( ptr, value );	// pos
+  if( size_only || ( c && value >= 0 && value < INT64_MAX &&
+                     ( ptr[c] == 0 || ptr[c] == ',' || ptr[c] == '-' ) ) )
     {
     range.pos( value );
     if( ptr[c] == 0 ) { range.size( INT64_MAX - value ); return; }
@@ -321,6 +326,18 @@ int open_instream( const char * const name, struct stat * const in_statsp,
 
 namespace {
 
+void set_a_outname( const std::string & name )
+  {
+  output_filename = name;
+  if( name.size() > 5 && name.compare( name.size() - 5, 5, ".lzma" ) == 0 )
+    output_filename.erase( name.size() - 2 );
+  else if( name.size() > 4 && name.compare( name.size() - 4, 4, ".tlz" ) == 0 )
+    output_filename.insert( name.size() - 2, "ar." );
+  else if( name.size() <= 3 || name.compare( name.size() - 3, 3, ".lz" ) != 0 )
+    output_filename += known_extensions[0].from;
+  }
+
+
 void set_d_outname( const std::string & name, const int i )
   {
   if( i >= 0 )
@@ -339,18 +356,25 @@ void set_d_outname( const std::string & name, const int i )
                   program_name, name.c_str(), output_filename.c_str() );
   }
 
+} // end namespace
 
-bool open_outstream( const bool force )
+bool open_outstream( const bool force, const bool from_stdin,
+                     const bool rw, const bool skipping )
   {
-  int flags = O_CREAT | O_WRONLY | O_BINARY;
+  const mode_t usr_rw = S_IRUSR | S_IWUSR;
+  const mode_t all_rw = usr_rw | S_IRGRP | S_IWGRP | S_IROTH | S_IWOTH;
+  const mode_t outfd_mode = from_stdin ? all_rw : usr_rw;
+  int flags = O_CREAT | ( rw ? O_RDWR : O_WRONLY ) | O_BINARY;
   if( force ) flags |= O_TRUNC; else flags |= O_EXCL;
 
   outfd = open( output_filename.c_str(), flags, outfd_mode );
-  if( outfd < 0 && verbosity >= 0 )
+  if( outfd >= 0 ) delete_output_on_interrupt = true;
+  else if( verbosity >= 0 )
     {
     if( errno == EEXIST )
-      std::fprintf( stderr, "%s: Output file '%s' already exists, skipping.\n",
-                    program_name, output_filename.c_str() );
+      std::fprintf( stderr, "%s: Output file '%s' already exists%s.\n",
+                    program_name, output_filename.c_str(), skipping ?
+                    ", skipping" : ". Use '--force' to overwrite it" );
     else
       std::fprintf( stderr, "%s: Can't create output file '%s': %s\n",
                     program_name, output_filename.c_str(), std::strerror( errno ) );
@@ -359,6 +383,37 @@ bool open_outstream( const bool force )
   }
 
 
+bool file_exists( const std::string & filename )
+  {
+  struct stat st;
+  if( stat( filename.c_str(), &st ) == 0 )
+    {
+    if( verbosity >= 0 )
+      std::fprintf( stderr, "%s: Output file '%s' already exists."
+                            " Use '--force' to overwrite it.\n",
+                    program_name, filename.c_str() );
+    return true;
+    }
+  return false;
+  }
+
+
+bool check_tty( const int infd, const Mode program_mode )
+  {
+  if( program_mode == m_alone_to_lz && isatty( outfd ) )
+    {
+    show_error( "I won't write compressed data to a terminal.", 0, true );
+    return false;
+    }
+  if( isatty( infd ) )			// all modes read compressed data
+    {
+    show_error( "I won't read compressed data from a terminal.", 0, true );
+    return false;
+    }
+  return true;
+  }
+
+
 void cleanup_and_fail( const int retval )
   {
   if( delete_output_on_interrupt )
@@ -374,6 +429,7 @@ void cleanup_and_fail( const int retval )
   std::exit( retval );
   }
 
+namespace {
 
      // Set permissions, owner and times.
 void close_and_set_permissions( const struct stat * const in_statsp )
@@ -390,7 +446,11 @@ void close_and_set_permissions( const struct stat * const in_statsp )
           fchmod( outfd, mode & ~( S_ISUID | S_ISGID | S_ISVTX ) ) != 0 )
         warning = true;
     }
-  if( close( outfd ) != 0 ) cleanup_and_fail( 1 );
+  if( close( outfd ) != 0 )
+    {
+    show_error( "Error closing output file", errno );
+    cleanup_and_fail( 1 );
+    }
   outfd = -1;
   delete_output_on_interrupt = false;
   if( in_statsp )
@@ -405,19 +465,6 @@ void close_and_set_permissions( const struct stat * const in_statsp )
   }
 
 
-std::string insert_fixed( std::string name )
-  {
-  if( name.size() > 7 && name.compare( name.size() - 7, 7, ".tar.lz" ) == 0 )
-    name.insert( name.size() - 7, "_fixed" );
-  else if( name.size() > 3 && name.compare( name.size() - 3, 3, ".lz" ) == 0 )
-    name.insert( name.size() - 3, "_fixed" );
-  else if( name.size() > 4 && name.compare( name.size() - 4, 4, ".tlz" ) == 0 )
-    name.insert( name.size() - 4, "_fixed" );
-  else name += "_fixed.lz";
-  return name;
-  }
-
-
 unsigned char xdigit( const int value )
   {
   if( value >= 0 && value <= 9 ) return '0' + value;
@@ -475,7 +522,7 @@ int decompress( const int infd, const Pretty_print & pp,
       const int size = rdec.read_data( header.data, File_header::size );
       if( rdec.finished() )			// End Of File
         {
-        if( first_member )
+        if( first_member || header.verify_prefix( size ) )
           { pp( "File ends unexpectedly at member header." ); retval = 2; }
         else if( size > 0 && !show_trailing_data( header.data, size, pp,
                                                   true, ignore_trailing ) )
@@ -499,14 +546,13 @@ int decompress( const int infd, const Pretty_print & pp,
         retval = 2; break;
         }
       const unsigned dictionary_size = header.dictionary_size();
-      if( dictionary_size < min_dictionary_size ||
-          dictionary_size > max_dictionary_size )
+      if( !isvalid_ds( dictionary_size ) )
         { pp( "Invalid dictionary size in member header." ); retval = 2; break; }
 
       if( verbosity >= 2 || ( verbosity == 1 && first_member ) )
         { pp(); show_header( dictionary_size ); }
 
-      LZ_decoder decoder( header, rdec, outfd );
+      LZ_decoder decoder( rdec, dictionary_size, outfd );
       const int result = decoder.decode_member( pp );
       partial_file_pos += rdec.member_position();
       if( result != 0 )
@@ -549,56 +595,42 @@ void set_signals()
 } // end namespace
 
 
-bool file_exists( const std::string & filename )
+int close_outstream( const struct stat * const in_statsp )
   {
-  struct stat st;
-  if( stat( filename.c_str(), &st ) == 0 )
-    {
-    if( verbosity >= 0 )
-      std::fprintf( stderr, "%s: Output file '%s' already exists."
-                            " Use '--force' to overwrite it.\n",
-                    program_name, filename.c_str() );
-    return true;
-    }
-  return false;
+  if( delete_output_on_interrupt )
+    close_and_set_permissions( in_statsp );
+  if( outfd >= 0 && close( outfd ) != 0 )
+    { show_error( "Can't close stdout", errno ); return 1; }
+  outfd = -1;
+  return 0;
   }
 
 
-int open_outstream_rw( const std::string & output_filename, const bool force )
+std::string insert_fixed( std::string name )
   {
-  int flags = O_CREAT | O_RDWR | O_BINARY;
-  if( force ) flags |= O_TRUNC; else flags |= O_EXCL;
-
-  int outfd = open( output_filename.c_str(), flags, all_rw );
-  if( outfd < 0 && verbosity >= 0 )
-    {
-    if( errno == EEXIST )
-      std::fprintf( stderr, "%s: Output file '%s' already exists."
-                            " Use '--force' to overwrite it.\n",
-                    program_name, output_filename.c_str() );
-    else
-      std::fprintf( stderr, "%s: Can't create output file '%s': %s\n",
-                    program_name, output_filename.c_str(), std::strerror( errno ) );
-    }
-  return outfd;
+  if( name.size() > 7 && name.compare( name.size() - 7, 7, ".tar.lz" ) == 0 )
+    name.insert( name.size() - 7, "_fixed" );
+  else if( name.size() > 3 && name.compare( name.size() - 3, 3, ".lz" ) == 0 )
+    name.insert( name.size() - 3, "_fixed" );
+  else if( name.size() > 4 && name.compare( name.size() - 4, 4, ".tlz" ) == 0 )
+    name.insert( name.size() - 4, "_fixed" );
+  else name += "_fixed.lz";
+  return name;
   }
 
 
 void show_error( const char * const msg, const int errcode, const bool help )
   {
-  if( verbosity >= 0 )
+  if( verbosity < 0 ) return;
+  if( msg && msg[0] )
     {
-    if( msg && msg[0] )
-      {
-      std::fprintf( stderr, "%s: %s", program_name, msg );
-      if( errcode > 0 )
-        std::fprintf( stderr, ": %s", std::strerror( errcode ) );
-      std::fputc( '\n', stderr );
-      }
-    if( help )
-      std::fprintf( stderr, "Try '%s --help' for more information.\n",
-                    invocation_name );
+    std::fprintf( stderr, "%s: %s", program_name, msg );
+    if( errcode > 0 ) std::fprintf( stderr, ": %s", std::strerror( errcode ) );
+    std::fputc( '\n', stderr );
     }
+  if( help )
+    std::fprintf( stderr, "Try '%s --help' for more information.\n",
+                  invocation_name );
   }
 
 
@@ -618,6 +650,15 @@ void show_error2( const char * const msg1, const char * const name,
   }
 
 
+void show_error4( const char * const msg1, const char * const name1,
+                  const char * const name2, const char * const msg2 )
+  {
+  if( verbosity >= 0 )
+    std::fprintf( stderr, "%s: %s '%s' and '%s' %s\n",
+                  program_name, msg1, name1, name2, msg2 );
+  }
+
+
 int main( const int argc, const char * const argv[] )
   {
   Block range( 0, 0 );
@@ -638,6 +679,7 @@ int main( const int argc, const char * const argv[] )
   const Arg_parser::Option options[] =
     {
     { 'a', "trailing-error",   Arg_parser::no  },
+    { 'A', "alone-to-lz",      Arg_parser::no  },
     { 'c', "stdout",           Arg_parser::no  },
     { 'd', "decompress",       Arg_parser::no  },
     { 'D', "range-decompress", Arg_parser::yes },
@@ -655,10 +697,11 @@ int main( const int argc, const char * const argv[] )
     { 't', "test",             Arg_parser::no  },
     { 'v', "verbose",          Arg_parser::no  },
     { 'V', "version",          Arg_parser::no  },
-    { 'x', "show-packets",     Arg_parser::maybe },
-    { 'y', "debug-delay",      Arg_parser::yes },
-    { 'z', "debug-repair",     Arg_parser::yes },
-    {  0 ,  0,                 Arg_parser::no  } };
+    { 'W', "debug-decompress", Arg_parser::yes },
+    { 'X', "show-packets",     Arg_parser::maybe },
+    { 'Y', "debug-delay",      Arg_parser::yes },
+    { 'Z', "debug-repair",     Arg_parser::yes },
+    {  0 , 0,                  Arg_parser::no  } };
 
   const Arg_parser parser( argc, argv, options );
   if( parser.error().size() )				// bad option
@@ -670,13 +713,15 @@ int main( const int argc, const char * const argv[] )
     const int code = parser.code( argind );
     if( !code ) break;					// no more options
     const std::string & arg = parser.argument( argind );
+    const char * const ptr = arg.c_str();
     switch( code )
       {
       case 'a': ignore_trailing = false; break;
+      case 'A': set_mode( program_mode, m_alone_to_lz ); break;
       case 'c': to_stdout = true; break;
       case 'd': set_mode( program_mode, m_decompress ); break;
       case 'D': set_mode( program_mode, m_range_dec );
-                parse_range( arg.c_str(), range ); break;
+                parse_range( ptr, range ); break;
       case 'f': force = true; break;
       case 'h': show_help(); return 0;
       case 'i': ignore_errors = true; break;
@@ -691,13 +736,14 @@ int main( const int argc, const char * const argv[] )
       case 't': set_mode( program_mode, m_test ); break;
       case 'v': if( verbosity < 4 ) ++verbosity; break;
       case 'V': show_version(); return 0;
-      case 'x': set_mode( program_mode, m_show_packets );
-                if( arg.size() )
-                  parse_pos_value( arg.c_str(), bad_pos, bad_value ); break;
-      case 'y': set_mode( program_mode, m_debug_delay );
-                parse_range( arg.c_str(), range ); break;
-      case 'z': set_mode( program_mode, m_debug_repair );
-                parse_pos_value( arg.c_str(), bad_pos, bad_value ); break;
+      case 'W': set_mode( program_mode, m_debug_decompress );
+                parse_pos_value( ptr, bad_pos, bad_value ); break;
+      case 'X': set_mode( program_mode, m_show_packets );
+                if( ptr[0] ) parse_pos_value( ptr, bad_pos, bad_value ); break;
+      case 'Y': set_mode( program_mode, m_debug_delay );
+                parse_range( ptr, range ); break;
+      case 'Z': set_mode( program_mode, m_debug_repair );
+                parse_pos_value( ptr, bad_pos, bad_value ); break;
       default : internal_error( "uncaught option." );
       }
     } // end process options
@@ -724,6 +770,10 @@ int main( const int argc, const char * const argv[] )
   switch( program_mode )
     {
     case m_none: internal_error( "invalid operation." ); break;
+    case m_alone_to_lz: break;
+    case m_debug_decompress:
+      one_file( filenames.size() );
+      return debug_decompress( filenames[0], bad_pos, verbosity, bad_value, false );
     case m_debug_delay:
       one_file( filenames.size() );
       return debug_delay( filenames[0], range, verbosity );
@@ -738,34 +788,35 @@ int main( const int argc, const char * const argv[] )
     case m_merge:
       if( filenames.size() < 2 )
         { show_error( "You must specify at least 2 files.", 0, true ); return 1; }
-      if( default_output_filename.empty() )
-        default_output_filename = insert_fixed( filenames[0] );
+      set_signals();
       return merge_files( filenames, default_output_filename, verbosity, force );
     case m_range_dec:
       one_file( filenames.size() );
+      set_signals();
       return range_decompress( filenames[0], default_output_filename, range,
                                verbosity, force, ignore_errors, to_stdout );
     case m_repair:
       one_file( filenames.size() );
-      if( default_output_filename.empty() )
-        default_output_filename = insert_fixed( filenames[0] );
+      set_signals();
       return repair_file( filenames[0], default_output_filename, verbosity,
                           force );
     case m_show_packets:
       one_file( filenames.size() );
-      return debug_show_packets( filenames[0], bad_pos, verbosity, bad_value );
+      return debug_decompress( filenames[0], bad_pos, verbosity, bad_value, true );
     case m_split:
       one_file( filenames.size() );
+      set_signals();
       return split_file( filenames[0], default_output_filename, verbosity, force );
     case m_test: break;
     }
     }
-  catch( std::bad_alloc ) { show_error( "Not enough memory." ); return 1; }
-  catch( Error e ) { show_error( e.msg, errno ); return 1; }
+  catch( std::bad_alloc )
+    { show_error( "Not enough memory." ); cleanup_and_fail( 1 ); }
+  catch( Error e ) { show_error( e.msg, errno ); cleanup_and_fail( 1 ); }
 
   if( program_mode == m_test )
     outfd = -1;
-  else if( program_mode != m_decompress )
+  else if( program_mode != m_alone_to_lz && program_mode != m_decompress )
     internal_error( "invalid decompressor operation." );
 
   if( filenames.empty() ) filenames.push_back("-");
@@ -776,6 +827,7 @@ int main( const int argc, const char * const argv[] )
   Pretty_print pp( filenames, verbosity );
 
   int retval = 0;
+  bool stdin_used = false;
   for( unsigned i = 0; i < filenames.size(); ++i )
     {
     struct stat in_stats;
@@ -783,6 +835,7 @@ int main( const int argc, const char * const argv[] )
 
     if( filenames[i].empty() || filenames[i] == "-" )
       {
+      if( stdin_used ) continue; else stdin_used = true;
       input_filename.clear();
       infd = STDIN_FILENO;
       if( program_mode != m_test )
@@ -792,8 +845,10 @@ int main( const int argc, const char * const argv[] )
         else
           {
           output_filename = default_output_filename;
-          outfd_mode = all_rw;
-          if( !open_outstream( force ) )
+          if( program_mode == m_alone_to_lz &&
+              extension_index( default_output_filename ) < 0 )
+            output_filename += known_extensions[0].from;
+          if( !open_outstream( force, true ) )
             {
             if( retval < 1 ) retval = 1;
             close( infd ); infd = -1;
@@ -813,9 +868,10 @@ int main( const int argc, const char * const argv[] )
         if( to_stdout ) outfd = STDOUT_FILENO;
         else
           {
-          set_d_outname( input_filename, extension_index( input_filename ) );
-          outfd_mode = usr_rw;
-          if( !open_outstream( force ) )
+          if( program_mode == m_alone_to_lz )
+            set_a_outname( input_filename );
+          else set_d_outname( input_filename, extension_index( input_filename ) );
+          if( !open_outstream( force, false ) )
             {
             if( retval < 1 ) retval = 1;
             close( infd ); infd = -1;
@@ -825,17 +881,19 @@ int main( const int argc, const char * const argv[] )
         }
       }
 
-    if( isatty( infd ) )
+    if( !check_tty( infd, program_mode ) )
       {
-      show_error( "I won't read compressed data from a terminal.", 0, true );
-      return 1;
+      if( retval < 1 ) retval = 1;
+      cleanup_and_fail( retval );
       }
 
-    if( output_filename.size() && !to_stdout && program_mode != m_test )
-      delete_output_on_interrupt = true;
     const struct stat * const in_statsp = input_filename.size() ? &in_stats : 0;
     pp.set_name( input_filename );
-    const int tmp = decompress( infd, pp, ignore_trailing, program_mode == m_test );
+    int tmp;
+    if( program_mode == m_alone_to_lz )
+      tmp = alone_to_lz( infd, pp );
+    else
+      tmp = decompress( infd, pp, ignore_trailing, program_mode == m_test );
     if( tmp > retval ) retval = tmp;
     if( tmp && program_mode != m_test ) cleanup_and_fail( retval );