path: root/zcmp.cc

/* Zcmp - decompress and compare two files byte by byte
   Copyright (C) 2010-2023 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 <cctype>
#include <cerrno>
#include <climits>
#include <csignal>
#include <cstdio>
#include <cstdlib>
#include <cstring>
#include <string>
#include <vector>
#include <fcntl.h>
#include <stdint.h>
#include <unistd.h>
#include <sys/stat.h>
#if defined __MSVCRT__ || defined __OS2__
#include <io.h>
#endif

#include "arg_parser.h"
#include "rc.h"
#include "zutils.h"

#ifndef LLONG_MAX
#define LLONG_MAX 0x7FFFFFFFFFFFFFFFLL
#endif


namespace {

#include "zcmpdiff.cc"

void show_help()
  {
  std::printf( "zcmp compares two files and, if they differ, writes to standard output the\n"
               "first byte and line number where they differ. Bytes and lines are numbered\n"
               "starting with 1. A hyphen '-' used as a file argument means standard input.\n"
               "If any file given is compressed, its decompressed content is used. Compressed\n"
               "files are decompressed on the fly; no temporary files are created.\n"
               "\nThe formats supported are bzip2, gzip, lzip, xz, and zstd.\n"
               "\nUsage: zcmp [options] file1 [file2]\n"
               "\nzcmp compares file1 to file2. The standard input is used only if file1 or\n"
               "file2 refers to standard input. If file2 is omitted zcmp tries the\n"
               "following:\n"
               "\n  - If file1 is compressed, compares its decompressed contents with\n"
               "  the corresponding uncompressed file (the name of file1 with the\n"
               "  extension removed).\n"
               "\n  - If file1 is uncompressed, compares it with the decompressed\n"
               "  contents of file1.[lz|bz2|gz|zst|xz] (the first one that is found).\n"
               "\nExit status is 0 if inputs are identical, 1 if different, 2 if trouble.\n"
               "\nOptions:\n"
               "  -h, --help                        display this help and exit\n"
               "  -V, --version                     output version information and exit\n"
               "  -b, --print-bytes                 print differing bytes\n"
               "  -H, --hexadecimal                 print hexadecimal values instead of octal\n"
               "  -i, --ignore-initial=<n>[:<n2>]   ignore differences in the first <n> bytes\n"
               "  -l, --list                        list position, value of all differing bytes\n"
               "  -M, --format=<list>               process only the formats in <list>\n"
               "  -n, --bytes=<n>                   compare at most <n> bytes\n"
               "  -N, --no-rcfile                   don't read runtime configuration file\n"
               "  -O, --force-format=[<f1>][,<f2>]  force one or both input formats\n"
               "  -q, --quiet, --silent             suppress diagnostics written to stderr\n"
               "  -s, --script                      suppress messages about file differences\n"
               "  -v, --verbose                     verbose mode (opposite of --quiet)\n"
               "      --bz2=<command>               set compressor and options for bzip2 format\n"
               "      --gz=<command>                set compressor and options for gzip format\n"
               "      --lz=<command>                set compressor and options for lzip format\n"
               "      --xz=<command>                set compressor and options for xz format\n"
               "      --zst=<command>               set compressor and options for zstd format\n"
               "\nValid formats for options '-M' and '-O' are 'bz2', 'gz', 'lz', 'xz', 'zst',\n"
               "and 'un' for uncompressed.\n"
               "\nByte counts given as arguments to options may be expressed in decimal,\n"
               "hexadecimal, or octal (using the same syntax as integer constants in C++),\n"
               "and 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" );
  show_help_addr();
  }


// separate numbers of 5 or more digits in groups of 3 digits using '_'
const char * format_num3( long long num )
  {
  const char * const si_prefix = "kMGTPEZY";
  const char * const binary_prefix = "KMGTPEZY";
  enum { buffers = 8, bufsize = 4 * sizeof num };
  static char buffer[buffers][bufsize];	// circle of static buffers for printf
  static int current = 0;

  char * const buf = buffer[current++]; current %= buffers;
  char * p = buf + bufsize - 1;		// fill the buffer backwards
  *p = 0;	// terminator
  const bool negative = num < 0;
  char prefix = 0;			// try binary first, then si
  for( int i = 0; i < 8 && num != 0 && ( num / 1024 ) * 1024 == num; ++i )
    { num /= 1024; prefix = binary_prefix[i]; }
  if( prefix ) *(--p) = 'i';
  else
    for( int i = 0; i < 8 && num != 0 && ( num / 1000 ) * 1000 == num; ++i )
      { num /= 1000; prefix = si_prefix[i]; }
  if( prefix ) *(--p) = prefix;
  const bool split = num >= 10000 || num <= -10000;

  for( int i = 0; ; )
    {
    long long onum = num; num /= 10;
    *(--p) = llabs( onum - ( 10 * num ) ) + '0'; if( num == 0 ) break;
    if( split && ++i >= 3 ) { i = 0; *(--p) = '_'; }
    }
  if( negative ) *(--p) = '-';
  return p;
  }


long long getnum( const char * const arg, const char * const option_name,
                  const char ** const tailp = 0,
                  const long long llimit = 0,
                  const long long ulimit = LLONG_MAX )
  {
  char * tail;
  errno = 0;
  long long result = strtoll( arg, &tail, 0 );
  if( tail == arg )
    { show_option_error( arg, "Bad or missing numerical argument in",
                         option_name ); std::exit( 2 ); }
  if( result < 0 ) errno = ERANGE;

  if( !errno && tail[0] && std::isalpha( tail[0] ) )
    {
    const unsigned char ch = *tail++;
    int factor;
    bool bsuf;					// 'B' suffix is present
    if( tail[0] == 'i' ) { ++tail; factor = 1024; } else factor = 1000;
    if( tail[0] == 'B' ) { ++tail; bsuf = true; } else bsuf = false;
    int exponent = -1;				// -1 = bad multiplier
    switch( ch )
      {
      case 'Y': exponent = 8; break;
      case 'Z': exponent = 7; break;
      case 'E': exponent = 6; break;
      case 'P': exponent = 5; break;
      case 'T': exponent = 4; break;
      case 'G': exponent = 3; break;
      case 'M': exponent = 2; break;
      case 'K': if( factor == 1024 ) exponent = 1; break;
      case 'k': if( factor == 1000 ) exponent = 1; break;
      case 'B': if( factor == 1000 && !bsuf ) exponent = 0; break;
      }
    if( exponent < 0 )
      { show_option_error( arg, "Bad multiplier in numerical argument of",
                           option_name ); std::exit( 2 ); }
    for( int i = 0; i < exponent; ++i )
      {
      if( ulimit / factor >= result ) result *= factor;
      else { errno = ERANGE; break; }
      }
    }
  if( !errno && ( result < llimit || result > ulimit ) ) errno = ERANGE;
  if( errno )
    {
    if( verbosity >= 0 )
      std::fprintf( stderr, "%s: '%s': Value out of limits [%s,%s] in "
                    "option '%s'.\n", program_name, arg, format_num3( llimit ),
                    format_num3( ulimit ), option_name );
    std::exit( 2 );
    }
  if( tailp ) *tailp = tail;
  return result;
  }


void parse_ignore_initial( const char * const arg, const char * const pn,
                           long long ignore_initial[2] )
  {
  const char * tail;
  ignore_initial[0] = getnum( arg, pn, &tail );
  if( *tail == ':' || *tail == ',' )
    ignore_initial[1] = getnum( ++tail, pn );
  else if( *tail == 0 ) ignore_initial[1] = ignore_initial[0];
  else { show_option_error( arg, "Missing colon in", pn ); std::exit( 2 ); }
  }


bool skip_ignore_initial( const long long ignore_initial, const int infd )
  {
  if( ignore_initial > 0 )
    {
    const int buffer_size = 4096;
    long long rest = ignore_initial;
    uint8_t buffer[buffer_size];
    while( rest > 0 )
      {
      const int size = std::min( rest, (long long)buffer_size );
      const int rd = readblock( infd, buffer, size );
      if( rd != size && errno ) return false;
      if( rd < size ) break;
      rest -= rd;
      }
    }
  return true;
  }


/* Put into buf the unsigned char c, making unprintable bytes visible by
   quoting like cat -t does. */
void sprintc( char * const buf, unsigned char c )
  {
  int i = 0;

  if( c < 32 || c >= 127 )
    {
    if( c >= 128 ) { c -= 128; buf[i++] = 'M'; buf[i++] = '-'; }
    if( c < 32 ) { c += 64; buf[i++] = '^'; }
    else if( c == 127 ) { c = '?'; buf[i++] = '^'; }
    }
  buf[i++] = c;
  buf[i++] = 0;
  }


int block_compare( const uint8_t * const buffer0,
                   const uint8_t * const buffer1,
                   unsigned long long * const line_numberp )
  {
  const uint8_t * p0 = buffer0;
  const uint8_t * p1 = buffer1;

  if( line_numberp )
    {
    int nl_count = 0;
    while( *p0 == *p1 )
      { if( *p0 == '\n' ) { ++nl_count; } ++p0; ++p1; }
    *line_numberp += nl_count;
    }
  else while( *p0 == *p1 ) { ++p0; ++p1; }
  return p0 - buffer0;
  }


int cmp( const long long max_size, const int infd[2],
         const std::string filenames[2], bool finished[2],
         const bool hexadecimal, const bool list, const bool print_bytes,
         const bool scripted )
  {
  const int buffer_size = 4096;
  unsigned long long byte_number = 1;
  unsigned long long line_number = 1;
  // remaining number of bytes to compare
  long long rest = ( max_size >= 0 ) ? max_size : buffer_size;
  // buffers with space for sentinels at the end
  uint8_t * const buffer0 = new uint8_t[2*(buffer_size+1)];
  uint8_t * const buffer1 = buffer0 + buffer_size + 1;
  uint8_t * buffer[2];
  buffer[0] = buffer0; buffer[1] = buffer1;
  int retval = 0;

  while( rest > 0 )
    {
    const int size = std::min( (long long)buffer_size, rest );
    if( max_size >= 0 ) rest -= size;
    int rd[2];			// number of bytes read from each file
    for( int i = 0; i < 2; ++i )
      {
      rd[i] = readblock( infd[i], buffer[i], size );
      if( rd[i] != size && errno )
        { show_file_error( filenames[i].c_str(), "Read error", errno );
          retval = 2; goto done; }
      }
    for( int i = 0; i < 2; ++i )
      if( rd[i] < size ) finished[i] = true;

    const int min_rd = std::min( rd[0], rd[1] );
    buffer0[min_rd] = 0;		// sentinels for the block compare
    buffer1[min_rd] = 1;

    int first_diff = block_compare( buffer0, buffer1, list ? 0 : &line_number );
    byte_number += first_diff;

    if( first_diff < min_rd )
      {
      retval = 1;				// difference found
      if( scripted ) break;			// status only
      if( !list )				// show first difference
        {
        if( !print_bytes )
          std::printf( "%s %s differ: byte %llu, line %llu\n",
                       filenames[0].c_str(), filenames[1].c_str(),
                       byte_number, line_number );
        else
          {
          const unsigned char c0 = buffer0[first_diff];
          const unsigned char c1 = buffer1[first_diff];
          char buf0[5], buf1[5];
          sprintc( buf0, c0 ); sprintc( buf1, c1 );
          std::printf( hexadecimal ?
                       "%s %s differ: byte %llu, line %llu is %02X %s %02X %s\n" :
                       "%s %s differ: byte %llu, line %llu is %3o %s %3o %s\n",
                       filenames[0].c_str(), filenames[1].c_str(),
                       byte_number, line_number, c0, buf0, c1, buf1 );
          }
        std::fflush( stdout );
        break;
        }
      else			// list ; show all differences
        {
        for( ; first_diff < min_rd; ++byte_number, ++first_diff )
          {
          const unsigned char c0 = buffer0[first_diff];
          const unsigned char c1 = buffer1[first_diff];
          if( c0 != c1 )
            {
            if( !print_bytes )
              std::printf( hexadecimal ? "%llu %02X %02X\n" : "%llu %3o %3o\n",
                           byte_number, c0, c1 );
            else
              {
              char buf0[5], buf1[5];
              sprintc( buf0, c0 ); sprintc( buf1, c1 );
              std::printf( hexadecimal ? "%llu %02X %-4s %02X %s\n" :
                           "%llu %3o %-4s %3o %s\n",
                           byte_number, c0, buf0, c1, buf1 );
              }
            }
          }
        std::fflush( stdout );
        }
      }

    if( rd[0] != rd[1] )
      {
      if( verbosity >= 0 )
        std::fprintf( stderr, list ?
                      "%s: EOF on %s after byte %llu\n" :
                      "%s: EOF on %s after byte %llu, in line %llu\n",
                      program_name, filenames[rd[1]<rd[0]].c_str(),
                      byte_number - 1, line_number );
      retval = 1; break;
      }
    if( min_rd != buffer_size ) break;
    }
done:
  delete[] buffer0;
  return retval;
  }

} // end namespace


int main( const int argc, const char * const argv[] )
  {
  enum { bz2_opt = 256, gz_opt, lz_opt, xz_opt, zst_opt };
  // number of initial bytes ignored for each file
  long long ignore_initial[2] = { 0, 0 };
  long long max_size = -1;			// < 0 means unlimited size
  int format_types[2] = { -1, -1 };		// < 0 means undefined
  bool hexadecimal = false;
  bool list = false;		// list position, value of all differing bytes
  bool print_bytes = false;	// print differing bytes
  bool scripted = false;	// suppress messages about file differences
  program_name = "zcmp";
  invocation_name = ( argc > 0 ) ? argv[0] : program_name;

  const Arg_parser::Option options[] =
    {
    { 'b', "print-bytes",    Arg_parser::no  },
    { 'h', "help",           Arg_parser::no  },
    { 'H', "hexadecimal",    Arg_parser::no  },
    { 'i', "ignore-initial", Arg_parser::yes },
    { 'l', "list",           Arg_parser::no  },
    { 'M', "format",         Arg_parser::yes },
    { 'n', "bytes",          Arg_parser::yes },
    { 'N', "no-rcfile",      Arg_parser::no  },
    { 'O', "force-format",   Arg_parser::yes },
    { 'q', "quiet",          Arg_parser::no  },
    { 'q', "silent",         Arg_parser::no  },
    { 's', "script",         Arg_parser::no  },
    { 'v', "verbose",        Arg_parser::no  },
    { 'V', "version",        Arg_parser::no  },
    { bz2_opt, "bz2",        Arg_parser::yes },
    { gz_opt,  "gz",         Arg_parser::yes },
    { lz_opt,  "lz",         Arg_parser::yes },
    { xz_opt,  "xz",         Arg_parser::yes },
    { zst_opt, "zst",        Arg_parser::yes },
    {  0,   0,               Arg_parser::no  } };

  const Arg_parser parser( argc, argv, options );
  if( parser.error().size() )				// bad option
    { show_error( parser.error().c_str(), 0, true ); return 2; }

  maybe_process_config_file( parser );

  int argind = 0;
  for( ; argind < parser.arguments(); ++argind )
    {
    const int code = parser.code( argind );
    if( !code ) break;					// no more options
    const char * const pn = parser.parsed_name( argind ).c_str();
    const std::string & sarg = parser.argument( argind );
    const char * const arg = sarg.c_str();
    switch( code )
      {
      case 'b': print_bytes = true; break;
      case 'h': show_help(); return 0;
      case 'H': hexadecimal = true; break;
      case 'i': parse_ignore_initial( arg, pn, ignore_initial ); break;
      case 'l': list = true; break;
      case 'M': parse_format_list( sarg, pn ); break;
      case 'n': max_size = getnum( arg, pn ); break;
      case 'N': break;
      case 'O': parse_format_types2( sarg, pn, format_types ); break;
      case 'q': verbosity = -1; break;
      case 's': scripted = true; break;
      case 'v': if( verbosity < 4 ) ++verbosity; break;
      case 'V': show_version(); return 0;
      case bz2_opt: parse_compressor( sarg, pn, fmt_bz2 ); break;
      case gz_opt: parse_compressor( sarg, pn, fmt_gz ); break;
      case lz_opt: parse_compressor( sarg, pn, fmt_lz ); break;
      case xz_opt: parse_compressor( sarg, pn, fmt_xz ); break;
      case zst_opt: parse_compressor( sarg, pn, fmt_zst ); break;
      default : internal_error( "uncaught option." );
      }
    } // end process options

#if defined __MSVCRT__ || defined __OS2__
  setmode( STDIN_FILENO, O_BINARY );
  setmode( STDOUT_FILENO, O_BINARY );
#endif

  const int files = parser.arguments() - argind;
  if( files < 1 ) { show_error( "No files given.", 0, true ); return 2; }
  if( files > 2 ) { show_error( "Too many files.", 0, true ); return 2; }

  std::string filenames[2];		// file names of the two input files
  filenames[0] = parser.argument( argind );
  if( files == 2 ) filenames[1] = parser.argument( argind + 1 );

  int infd[2];				// file descriptors of the two files
  infd[0] = ( filenames[0] == "-" ) ?
    STDIN_FILENO : open_instream( filenames[0] );
  if( infd[0] < 0 ) return 2;

  if( files == 2 )
    {
    if( check_identical( filenames[0].c_str(), filenames[1].c_str() ) )
      {
      if( ignore_initial[0] == ignore_initial[1] ) return 0;
      else { show_error( "Can't compare parts of same file." ); return 2; }
      }
    infd[1] = ( filenames[1] == "-" ) ?
      STDIN_FILENO : open_instream( filenames[1] );
    if( infd[1] < 0 ) return 2;
    }
  else
    {
    if( filenames[0] == "-" )
      { show_error( "Missing operand after '-'.", 0, true ); return 2; }
    if( format_types[0] >= 0 || format_types[1] >= 0 )
      { show_error( "Two files must be given when format is specified.", 0, true );
        return 2; }
    filenames[1] = filenames[0];
    infd[1] = open_other_instream( filenames[1] );
    if( infd[1] < 0 )
      {
      if( verbosity >= 0 )
        std::fprintf( stderr, "%s: Can't find file to compare with '%s'.\n",
                      program_name, filenames[0].c_str() );
      show_error( 0, 0, true ); return 2;
      }
    }

  int old_infd[2];		// copy of file descriptors of the two files
  old_infd[0] = infd[0]; old_infd[1] = infd[1];
  Children children[2];
  if( !set_data_feeder( filenames[0], &infd[0], children[0], format_types[0] ) ||
      !set_data_feeder( filenames[1], &infd[1], children[1], format_types[1] ) )
    return 2;

  for( int i = 0; i < 2; ++i )
    if( !skip_ignore_initial( ignore_initial[i], infd[i] ) )
      {
      show_file_error( filenames[i].c_str(),
                       "Read error skipping initial bytes", errno );
      return 2;
      }

  bool finished[2] = { false, false };
  int retval = cmp( max_size, infd, filenames, finished, hexadecimal, list,
                    print_bytes, scripted );

  for( int i = 0; i < 2; ++i )
    if( !good_status( children[i], finished[i] ) ) retval = 2;

  for( int i = 0; i < 2; ++i )
    {
    if( close( infd[i] ) != 0 )
      { show_close_error(); retval = 2; }
    if( filenames[i] != "-" && close( old_infd[i] ) != 0 )
      {
      show_file_error( filenames[i].c_str(), "Error closing input file", errno );
      retval = 2;
      }
    }
  if( std::fclose( stdout ) != 0 )
    {
    show_error( "Error closing stdout", errno );
    retval = 2;
    }

  return retval;
  }