Merging upstream version 1.21.

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

1 files changed, 167 insertions, 108 deletions

diff --git a/main.cc b/main.cc
index 3307194..142216f 100644
--- a/main.cc
+++ b/main.cc
@@ -1,5 +1,5 @@
 /*  Lziprecover - Data recovery tool for the lzip format
-    Copyright (C) 2009-2018 Antonio Diaz Diaz.
+    Copyright (C) 2009-2019 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
@@ -38,25 +38,29 @@
 #include <unistd.h>
 #include <utime.h>
 #include <sys/stat.h>
-#if defined(__MSVCRT__)
+#if defined(__MSVCRT__) || defined(__OS2__) || defined(__DJGPP__)
 #include <io.h>
+#if defined(__MSVCRT__)
 #define fchmod(x,y) 0
 #define fchown(x,y,z) 0
 #define SIGHUP SIGTERM
 #define S_ISSOCK(x) 0
+#ifndef S_IRGRP
 #define S_IRGRP 0
 #define S_IWGRP 0
 #define S_IROTH 0
 #define S_IWOTH 0
 #endif
-#if defined(__OS2__)
-#include <io.h>
+#endif
+#if defined(__DJGPP__)
+#define S_ISSOCK(x) 0
+#define S_ISVTX 0
+#endif
 #endif
 
 #include "arg_parser.h"
 #include "lzip.h"
 #include "decoder.h"
-#include "block.h"
 
 #ifndef O_BINARY
 #define O_BINARY 0
@@ -67,12 +71,11 @@
 #endif
 
 int verbosity = 0;
-std::string output_filename;		// global vars for output file
-int outfd = -1;
+std::string output_filename;	// global vars for output file
+int outfd = -1;			// see 'delete_output_on_interrupt' below
 
 namespace {
 
-const char * const Program_name = "Lziprecover";
 const char * const program_name = "lziprecover";
 const char * invocation_name = 0;
 
@@ -82,53 +85,58 @@ const struct { const char * from; const char * to; } known_extensions[] = {
   { 0,      0      } };
 
 enum Mode { m_none, m_alone_to_lz, m_debug_decompress, m_debug_delay,
-            m_debug_repair, m_decompress, m_dump_tdata, m_list, m_merge,
-            m_range_dec, m_remove_tdata, m_repair, m_show_packets, m_split,
-            m_strip_tdata, m_test };
+            m_debug_repair, m_decompress, m_dump, m_list, m_merge,
+            m_range_dec, m_remove, m_repair, m_show_packets, m_split,
+            m_strip, m_test };
 
+/* Variable used in signal handler context.
+   It is not declared volatile because the handler never returns. */
 bool delete_output_on_interrupt = false;
 
 
 void show_help()
   {
-  std::printf( "%s - Data recovery tool and decompressor for the lzip format.\n", Program_name );
-  std::printf( "\nLziprecover can repair perfectly most files with small errors (up to one\n"
+  std::printf( "Lziprecover is a data recovery tool and decompressor for files in the lzip\n"
+               "compressed data format (.lz). Lziprecover is able to repair slightly damaged\n"
+               "files, produce a correct file by merging the good parts of two or more\n"
+               "damaged copies, extract data from damaged files, decompress files and test\n"
+               "integrity of files.\n"
+               "\nLziprecover can repair perfectly most files with small errors (up to one\n"
                "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"
-               "\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 provides random access to the data in multimember files; it\n"
-               "only decompresses the members containing the desired data.\n"
-               "\nLziprecover facilitates the management of metadata stored as trailing\n"
-               "data in lzip files.\n"
+               "\nLziprecover can remove the damaged members from multimember files, for\n"
+               "example multimember tar.lz archives.\n"
+               "\nLziprecover provides random access to the data in multimember files; it only\n"
+               "decompresses the members containing the desired data.\n"
+               "\nLziprecover facilitates the management of metadata stored as trailing data\n"
+               "in lzip 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"
-               "  -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"
-               "  -i, --ignore-errors             make '--range-decompress' ignore data errors\n"
-               "  -k, --keep                      keep (don't delete) input files\n"
-               "  -l, --list                      print (un)compressed file sizes\n"
-               "  -m, --merge                     correct errors in file using several copies\n"
-               "  -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 multimember file in single-member files\n"
-               "  -t, --test                      test compressed file integrity\n"
-               "  -v, --verbose                   be verbose (a 2nd -v gives more)\n"
-               "      --loose-trailing            allow trailing data seeming corrupt header\n"
-               "      --dump-tdata                dump trailing data to standard output\n"
-               "      --remove-tdata              remove trailing data from files in place\n"
-               "      --strip-tdata               copy files to stdout without trailing data\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"
+               "  -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=<n-m>  decompress a range of bytes to stdout\n"
+               "  -f, --force                   overwrite existing output files\n"
+               "  -i, --ignore-errors           all errors in -D, format errors in -l, --dump\n"
+               "  -k, --keep                    keep (don't delete) input files\n"
+               "  -l, --list                    print (un)compressed file sizes\n"
+               "  -m, --merge                   correct errors in file using several copies\n"
+               "  -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 multimember file in single-member files\n"
+               "  -t, --test                    test compressed file integrity\n"
+               "  -v, --verbose                 be verbose (a 2nd -v gives more)\n"
+               "      --loose-trailing          allow trailing data seeming corrupt header\n"
+               "      --dump=<list>:d:t         dump members listed/damaged, tdata to stdout\n"
+               "      --remove=<list>:d:t       remove members, tdata from files in place\n"
+               "      --strip=<list>:d:t        copy files to stdout stripping members given\n" );
   if( verbosity >= 1 )
     {
     std::printf( "  -W, --debug-decompress=<pos>,<val>  set pos to val and decompress to stdout\n"
@@ -202,6 +210,46 @@ void show_header( const unsigned dictionary_size )
 #include "main_common.cc"
 
 
+// Colon-separated list of "damaged", "tdata", [r][^]<list> (1 1,3-5,8)
+void Member_list::parse( const char * p )
+  {
+  while( true )
+    {
+    const char * tp = p;	// points to terminator; ':' or null
+    while( *tp && *tp != ':' ) ++tp;
+    const unsigned len = tp - p;
+    if( std::isalpha( (const unsigned char)*p ) )
+      {
+      if( len <= 7 && std::strncmp( "damaged", p, len ) == 0 )
+        { damaged = true; goto next; }
+      if( len <= 5 && std::strncmp( "tdata", p, len ) == 0 )
+        { tdata = true; goto next; }
+      }
+    {
+    const bool reverse = ( *p == 'r' );
+    if( reverse ) ++p;
+    if( *p == '^' ) { ++p; if( reverse ) rin = false; else in = false; }
+    std::vector< Block > * rvp = reverse ? &rrange_vector : &range_vector;
+    while( std::isdigit( (const unsigned char)*p ) )
+      {
+      const char * tail;
+      const int pos = getnum( p, 0, 1, INT_MAX, &tail ) - 1;
+      if( rvp->size() && pos < rvp->back().end() ) break;
+      const int size = (*tail == '-') ?
+        getnum( tail + 1, 0, pos + 1, INT_MAX, &tail ) - pos : 1;
+      rvp->push_back( Block( pos, size ) );
+      if( tail == tp ) goto next;
+      if( *tail == ',' ) p = tail + 1; else break;
+      }
+    }
+    show_error( "Invalid list of members." );
+    std::exit( 1 );
+next:
+    if( *(p = tp) != 0 ) ++p; else return;
+    }
+  }
+
+
 namespace {
 
 // Recognized formats: <begin> <begin>-<end> <begin>,<size> ,<size>
@@ -215,11 +263,11 @@ void parse_range( const char * const ptr, Block & range )
     {
     range.pos( value );
     if( tail[0] == 0 ) { range.size( INT64_MAX - value ); return; }
-    const bool issize = ( tail[0] == ',' );
+    const bool is_size = ( tail[0] == ',' );
     value = getnum( tail + 1, 0, 1, INT64_MAX );		// size
-    if( issize || value > range.pos() )
+    if( is_size || value > range.pos() )
       {
-      if( !issize ) value -= range.pos();
+      if( !is_size ) value -= range.pos();
       if( INT64_MAX - range.pos() >= value ) { range.size( value ); return; }
       }
     }
@@ -343,23 +391,23 @@ int open_instream( const char * const name, struct stat * const in_statsp,
 int open_truncable_stream( const char * const name,
                            struct stat * const in_statsp )
   {
-  int infd = open( name, O_RDWR | O_BINARY );
-  if( infd < 0 )
+  int fd = open( name, O_RDWR | O_BINARY );
+  if( fd < 0 )
     show_file_error( name, "Can't open input file", errno );
   else
     {
-    const int i = fstat( infd, in_statsp );
+    const int i = fstat( fd, in_statsp );
     const mode_t mode = in_statsp->st_mode;
     if( i != 0 || !S_ISREG( mode ) )
       {
       if( verbosity >= 0 )
         std::fprintf( stderr, "%s: File '%s' is not a regular file.\n",
                       program_name, name );
-      close( infd );
-      infd = -1;
+      close( fd );
+      fd = -1;
       }
     }
-  return infd;
+  return fd;
   }
 
 
@@ -421,8 +469,17 @@ bool check_tty( const char * const input_filename, const int infd,
   }
 
 
+void set_signals( void (*action)(int) )
+  {
+  std::signal( SIGHUP, action );
+  std::signal( SIGINT, action );
+  std::signal( SIGTERM, action );
+  }
+
+
 void cleanup_and_fail( const int retval )
   {
+  set_signals( SIG_IGN );			// ignore signals
   if( delete_output_on_interrupt )
     {
     delete_output_on_interrupt = false;
@@ -438,6 +495,13 @@ void cleanup_and_fail( const int retval )
 
 namespace {
 
+extern "C" void signal_handler( int )
+  {
+  show_error( "Control-C or similar caught, quitting." );
+  cleanup_and_fail( 1 );
+  }
+
+
      // Set permissions, owner and times.
 void close_and_set_permissions( const struct stat * const in_statsp )
   {
@@ -517,9 +581,9 @@ int decompress( const unsigned long long cfile_size, const int infd,
     Range_decoder rdec( infd );
     for( bool first_member = true; ; first_member = false )
       {
-      File_header header;
+      Lzip_header header;
       rdec.reset_member_position();
-      const int size = rdec.read_data( header.data, File_header::size );
+      const int size = rdec.read_data( header.data, Lzip_header::size );
       if( rdec.finished() )			// End Of File
         {
         if( first_member )
@@ -573,30 +637,16 @@ int decompress( const unsigned long long cfile_size, const int infd,
         { std::fputs( testing ? "ok\n" : "done\n", stderr ); pp.reset(); }
       }
     }
-  catch( std::bad_alloc ) { pp( "Not enough memory." ); retval = 1; }
-  catch( Error e ) { pp(); show_error( e.msg, errno ); retval = 1; }
+  catch( std::bad_alloc & ) { pp( "Not enough memory." ); retval = 1; }
+  catch( Error & e ) { pp(); show_error( e.msg, errno ); retval = 1; }
   if( verbosity == 1 && retval == 0 )
     std::fputs( testing ? "ok\n" : "done\n", stderr );
   return retval;
   }
 
-
-extern "C" void signal_handler( int )
-  {
-  show_error( "Control-C or similar caught, quitting." );
-  cleanup_and_fail( 1 );
-  }
-
-
-void set_signals()
-  {
-  std::signal( SIGHUP, signal_handler );
-  std::signal( SIGINT, signal_handler );
-  std::signal( SIGTERM, signal_handler );
-  }
-
 } // end namespace
 
+void set_signal_handler() { set_signals( signal_handler ); }
 
 int close_outstream( const struct stat * const in_statsp )
   {
@@ -625,23 +675,15 @@ std::string insert_fixed( std::string name )
 void show_file_error( const char * const filename, const char * const msg,
                       const int errcode )
   {
-  if( verbosity < 0 ) return;
-  std::fprintf( stderr, "%s: %s: %s", program_name, filename, msg );
-  if( errcode > 0 ) std::fprintf( stderr, ": %s", std::strerror( errcode ) );
-  std::fputc( '\n', stderr );
-  }
-
-
-void show_error2( const char * const msg1, const char * const name,
-                  const char * const msg2 )
-  {
   if( verbosity >= 0 )
-    std::fprintf( stderr, "%s: %s '%s' %s\n", program_name, msg1, name, msg2 );
+    std::fprintf( stderr, "%s: %s: %s%s%s\n", program_name, filename, msg,
+                  ( errcode > 0 ) ? ": " : "",
+                  ( errcode > 0 ) ? std::strerror( errcode ) : "" );
   }
 
 
-void show_error4( const char * const msg1, const char * const name1,
-                  const char * const name2, const char * const msg2 )
+void show_2file_error( 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",
@@ -684,6 +726,7 @@ int main( const int argc, const char * const argv[] )
   {
   Block range( 0, 0 );
   Bad_byte bad_byte;
+  Member_list member_list;
   std::string default_output_filename;
   std::vector< std::string > filenames;
   Mode program_mode = m_none;
@@ -695,7 +738,7 @@ int main( const int argc, const char * const argv[] )
   bool to_stdout = false;
   invocation_name = argv[0];
 
-  enum { opt_dtd = 256, opt_lt, opt_rtd, opt_std };
+  enum { opt_du = 256, opt_dtd, opt_lt, opt_re, opt_rtd, opt_st, opt_std };
   const Arg_parser::Option options[] =
     {
     { 'a', "trailing-error",     Arg_parser::no  },
@@ -721,9 +764,12 @@ int main( const int argc, const char * const argv[] )
     { 'X', "show-packets",       Arg_parser::maybe },
     { 'Y', "debug-delay",        Arg_parser::yes },
     { 'Z', "debug-repair",       Arg_parser::yes },
+    { opt_du,  "dump",           Arg_parser::yes },
     { opt_dtd, "dump-tdata",     Arg_parser::no  },
     { opt_lt,  "loose-trailing", Arg_parser::no  },
+    { opt_re,  "remove",         Arg_parser::yes },
     { opt_rtd, "remove-tdata",   Arg_parser::no  },
+    { opt_st,  "strip",          Arg_parser::yes },
     { opt_std, "strip-tdata",    Arg_parser::no  },
     {  0 , 0,                    Arg_parser::no  } };
 
@@ -768,15 +814,24 @@ int main( const int argc, const char * const argv[] )
                 parse_range( arg, range ); break;
       case 'Z': set_mode( program_mode, m_debug_repair );
                 parse_pos_value( arg, bad_byte ); break;
-      case opt_dtd: set_mode( program_mode, m_dump_tdata ); break;
+      case opt_du: set_mode( program_mode, m_dump );
+                   member_list.parse( arg ); break;
+      case opt_dtd: set_mode( program_mode, m_dump );
+                    member_list.parse( "tdata" ); break;
       case opt_lt: loose_trailing = true; break;
-      case opt_rtd: set_mode( program_mode, m_remove_tdata ); break;
-      case opt_std: set_mode( program_mode, m_strip_tdata ); break;
+      case opt_re: set_mode( program_mode, m_remove );
+                   member_list.parse( arg ); break;
+      case opt_rtd: set_mode( program_mode, m_remove );
+                    member_list.parse( "tdata" ); break;
+      case opt_st: set_mode( program_mode, m_strip );
+                   member_list.parse( arg ); break;
+      case opt_std: set_mode( program_mode, m_strip );
+                    member_list.parse( "tdata" ); break;
       default : internal_error( "uncaught option." );
       }
     } // end process options
 
-#if defined(__MSVCRT__) || defined(__OS2__)
+#if defined(__MSVCRT__) || defined(__OS2__) || defined(__DJGPP__)
   setmode( STDIN_FILENO, O_BINARY );
   setmode( STDOUT_FILENO, O_BINARY );
 #endif
@@ -794,6 +849,7 @@ int main( const int argc, const char * const argv[] )
     if( filenames.back() != "-" ) filenames_given = true;
     }
 
+  const char terminator = isatty( STDOUT_FILENO ) ? '\r' : '\n';
   try {
   switch( program_mode )
     {
@@ -804,56 +860,54 @@ int main( const int argc, const char * const argv[] )
       return debug_decompress( filenames[0], bad_byte, false );
     case m_debug_delay:
       one_file( filenames.size() );
-      return debug_delay( filenames[0], range );
+      return debug_delay( filenames[0], range, terminator );
     case m_debug_repair:
       one_file( filenames.size() );
-      return debug_repair( filenames[0], bad_byte );
+      return debug_repair( filenames[0], bad_byte, terminator );
     case m_decompress: break;
-    case m_dump_tdata:
-    case m_strip_tdata:
+    case m_dump:
+    case m_strip:
       if( filenames.size() < 1 )
         { show_error( "You must specify at least 1 file.", 0, true ); return 1; }
-      if( default_output_filename.size() ) set_signals();
-      return dump_tdata( filenames, default_output_filename, force,
-                         program_mode == m_strip_tdata, loose_trailing );
+      return dump_members( filenames, default_output_filename, member_list,
+                           force, ignore_errors, ignore_trailing,
+                           loose_trailing, program_mode == m_strip );
     case m_list: break;
     case m_merge:
       if( filenames.size() < 2 )
         { show_error( "You must specify at least 2 files.", 0, true ); return 1; }
-      set_signals();
-      return merge_files( filenames, default_output_filename, force );
+      return merge_files( filenames, default_output_filename, force, terminator );
     case m_range_dec:
       one_file( filenames.size() );
-      set_signals();
       return range_decompress( filenames[0], default_output_filename, range,
                                force, ignore_errors, ignore_trailing,
                                loose_trailing, to_stdout );
-    case m_remove_tdata:
+    case m_remove:
       if( filenames.size() < 1 )
         { show_error( "You must specify at least 1 file.", 0, true ); return 1; }
-      return remove_tdata( filenames, loose_trailing );
+      return remove_members( filenames, member_list, ignore_errors,
+                             ignore_trailing, loose_trailing );
     case m_repair:
       one_file( filenames.size() );
-      set_signals();
-      return repair_file( filenames[0], default_output_filename, force );
+      return repair_file( filenames[0], default_output_filename, force, terminator );
     case m_show_packets:
       one_file( filenames.size() );
       return debug_decompress( filenames[0], bad_byte, true );
     case m_split:
       one_file( filenames.size() );
-      set_signals();
       return split_file( filenames[0], default_output_filename, force );
     case m_test: break;
     }
     }
-  catch( std::bad_alloc )
+  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 ); }
+  catch( Error & e ) { show_error( e.msg, errno ); cleanup_and_fail( 1 ); }
 
   if( filenames.empty() ) filenames.push_back("-");
 
   if( program_mode == m_list )
-    return list_files( filenames, ignore_trailing, loose_trailing );
+    return list_files( filenames, ignore_errors, ignore_trailing,
+                       loose_trailing );
 
   if( program_mode == m_test )
     outfd = -1;
@@ -862,7 +916,7 @@ int main( const int argc, const char * const argv[] )
 
   if( !to_stdout && program_mode != m_test &&
       ( filenames_given || default_output_filename.size() ) )
-    set_signals();
+    set_signals( signal_handler );
 
   Pretty_print pp( filenames );
 
@@ -941,6 +995,12 @@ int main( const int argc, const char * const argv[] )
     else
       tmp = decompress( cfile_size, infd, pp, ignore_trailing,
                         loose_trailing, program_mode == m_test );
+    if( close( infd ) != 0 )
+      {
+      show_error( input_filename.size() ? "Error closing input file" :
+                                          "Error closing stdin", errno );
+      if( tmp < 1 ) tmp = 1;
+      }
     if( tmp > retval ) retval = tmp;
     if( tmp )
       { if( program_mode != m_test ) cleanup_and_fail( retval );
@@ -950,7 +1010,6 @@ int main( const int argc, const char * const argv[] )
       close_and_set_permissions( in_statsp );
     if( input_filename.size() )
       {
-      close( infd );
       if( !keep_input_files && !to_stdout && program_mode != m_test )
         std::remove( input_filename.c_str() );
       }