summaryrefslogtreecommitdiffstats
path: root/main.cc
diff options
context:
space:
mode:
authorDaniel Baumann <daniel.baumann@progress-linux.org>2022-09-24 01:57:46 +0000
committerDaniel Baumann <daniel.baumann@progress-linux.org>2022-09-24 01:57:46 +0000
commit321b4b38cb22eb640619cebcb1230c604b2e9297 (patch)
treebea175e053890b150463151752b7a49dc49c436e /main.cc
parentAdding upstream version 0.22. (diff)
downloadtarlz-321b4b38cb22eb640619cebcb1230c604b2e9297.tar.xz
tarlz-321b4b38cb22eb640619cebcb1230c604b2e9297.zip
Adding upstream version 0.23.upstream/0.23
Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
Diffstat (limited to 'main.cc')
-rw-r--r--main.cc255
1 files changed, 168 insertions, 87 deletions
diff --git a/main.cc b/main.cc
index e04d37d..5a85ebb 100644
--- a/main.cc
+++ b/main.cc
@@ -15,16 +15,17 @@
along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
/*
- Exit status: 0 for a normal exit, 1 for environmental problems (file not
- found, files differ, invalid flags, I/O errors, etc), 2 to indicate a
- corrupt or invalid input file, 3 for an internal consistency error
- (e.g. bug) which caused tarlz to panic.
+ Exit status: 0 for a normal exit, 1 for environmental problems
+ (file not found, files differ, invalid command line options, I/O errors,
+ etc), 2 to indicate a corrupt or invalid input file, 3 for an internal
+ consistency error (e.g., bug) which caused tarlz to panic.
*/
#define _FILE_OFFSET_BITS 64
#include <cctype>
#include <cerrno>
+#include <cstdarg>
#include <cstdio>
#include <cstdlib>
#include <ctime>
@@ -81,30 +82,29 @@ void show_help( const long num_online )
"(uncompressed) tar. Moreover, the option '--keep-damaged' can be used to\n"
"recover as much data as possible from each damaged member, and lziprecover\n"
"can be used to recover some of the damaged members.\n"
- "\nUsage: %s [options] [files]\n", invocation_name );
- std::printf( "\nOptions:\n"
+ "\nUsage: %s operation [options] [files]\n", invocation_name );
+ std::printf( "\nOperations:\n"
" --help display this help and exit\n"
" -V, --version output version information and exit\n"
" -A, --concatenate append archives to the end of an archive\n"
- " -B, --data-size=<bytes> set target size of input data blocks [2x8=16 MiB]\n"
" -c, --create create a new archive\n"
- " -C, --directory=<dir> change to directory <dir>\n"
" -d, --diff find differences between archive and file system\n"
- " --ignore-ids ignore differences in owner and group IDs\n"
" --delete delete files/directories from an archive\n"
- " --exclude=<pattern> exclude files matching a shell pattern\n"
+ " -r, --append append files to the end of an archive\n"
+ " -t, --list list the contents of an archive\n"
+ " -x, --extract extract files/directories from an archive\n"
+ " -z, --compress compress existing POSIX tar archives\n"
+ " --check-lib check version of lzlib and exit\n"
+ "\nOptions:\n"
+ " -B, --data-size=<bytes> set target size of input data blocks [2x8=16 MiB]\n"
+ " -C, --directory=<dir> change to directory <dir>\n"
" -f, --file=<archive> use archive file <archive>\n"
" -h, --dereference follow symlinks; archive the files they point to\n"
- " --mtime=<date> use <date> as mtime for files added to archive\n"
" -n, --threads=<n> set number of (de)compression threads [%ld]\n"
" -o, --output=<file> compress to <file>\n"
" -p, --preserve-permissions don't subtract the umask on extraction\n"
" -q, --quiet suppress all messages\n"
- " -r, --append append files to the end of an archive\n"
- " -t, --list list the contents of an archive\n"
" -v, --verbose verbosely list files processed\n"
- " -x, --extract extract files/directories from an archive\n"
- " -z, --compress compress existing POSIX tar archives\n"
" -0 .. -9 set compression level [default 6]\n"
" --uncompressed don't compress the archive created\n"
" --asolid create solidly compressed appendable archive\n"
@@ -115,10 +115,13 @@ void show_help( const long num_online )
" --anonymous equivalent to '--owner=root --group=root'\n"
" --owner=<owner> use <owner> name/ID for files added to archive\n"
" --group=<group> use <group> name/ID for files added to archive\n"
+ " --exclude=<pattern> exclude files matching a shell pattern\n"
+ " --ignore-ids ignore differences in owner and group IDs\n"
+ " --ignore-overflow ignore mtime overflow differences on 32-bit\n"
" --keep-damaged don't delete partially extracted files\n"
" --missing-crc exit with error status if missing extended CRC\n"
+ " --mtime=<date> use <date> as mtime for files added to archive\n"
" --out-slots=<n> number of 1 MiB output packets buffered [64]\n"
- " --check-lib compare version of lzlib.h with liblz.{a,so}\n"
" --warn-newer warn if any file is newer than the archive\n"
/* " --permissive allow repeated extended headers and records\n"*/,
num_online );
@@ -126,10 +129,12 @@ void show_help( const long num_online )
{
std::printf( " --debug=<level> (0-1) print debug statistics to stderr\n" );
}
- std::printf( "\nExit status: 0 for a normal exit, 1 for environmental problems (file not\n"
- "found, files differ, invalid flags, I/O errors, etc), 2 to indicate a\n"
- "corrupt or invalid input file, 3 for an internal consistency error (e.g. bug)\n"
- "which caused tarlz to panic.\n"
+ std::printf( "\nIf no archive is specified, tarlz tries to read it from standard input or\n"
+ "write it to standard output.\n"
+ "\nExit status: 0 for a normal exit, 1 for environmental problems\n"
+ "(file not found, files differ, invalid command line options, I/O errors,\n"
+ "etc), 2 to indicate a corrupt or invalid input file, 3 for an internal\n"
+ "consistency error (e.g., bug) which caused tarlz to panic.\n"
"\nReport bugs to lzip-bug@nongnu.org\n"
"Tarlz home page: http://www.nongnu.org/lzip/tarlz.html\n" );
}
@@ -204,56 +209,63 @@ int check_lib()
}
-// separate large numbers >= 100_000 in groups of 3 digits using '_'
-const char * format_num3( unsigned long long num )
+// separate numbers of 6 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 (long long) };
+ 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 >= 1024 && num % 1024 == 0; ++i )
+ 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 >= 1000 && num % 1000 == 0; ++i )
+ 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 >= 100000;
+ const bool split = num >= 100000 || num <= -100000;
for( int i = 0; ; )
{
- *(--p) = num % 10 + '0'; num /= 10; if( num == 0 ) break;
+ 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;
}
-unsigned long long getnum( const char * const arg,
- const char * const option_name,
- const unsigned long long llimit,
- const unsigned long long ulimit )
+void show_option_error( const char * const arg, const char * const msg,
+ const char * const option_name )
+ {
+ if( verbosity >= 0 )
+ std::fprintf( stderr, "%s: '%s': %s option '%s'.\n",
+ program_name, arg, msg, option_name );
+ }
+
+
+long long getnum( const char * const arg, const char * const option_name,
+ const long long llimit = LLONG_MIN,
+ const long long ulimit = LLONG_MAX )
{
char * tail;
errno = 0;
- unsigned long long result = strtoull( arg, &tail, 0 );
+ long long result = strtoll( arg, &tail, 0 );
if( tail == arg )
- {
- if( verbosity >= 0 )
- std::fprintf( stderr, "%s: Bad or missing numerical argument in "
- "option '%s'.\n", program_name, option_name );
- std::exit( 1 );
- }
+ { show_option_error( arg, "Bad or missing numerical argument in",
+ option_name ); std::exit( 1 ); }
if( !errno && tail[0] )
{
- const unsigned factor = ( tail[1] == 'i' ) ? 1024 : 1000;
+ const int factor = ( tail[1] == 'i' ) ? 1024 : 1000;
int exponent = 0; // 0 = bad multiplier
switch( tail[0] )
{
@@ -268,15 +280,12 @@ unsigned long long getnum( const char * const arg,
case 'k': if( factor == 1000 ) exponent = 1; break;
}
if( exponent <= 0 )
- {
- if( verbosity >= 0 )
- std::fprintf( stderr, "%s: Bad multiplier in numerical argument of "
- "option '%s'.\n", program_name, option_name );
- std::exit( 1 );
- }
+ { show_option_error( arg, "Bad multiplier in numerical argument of",
+ option_name ); std::exit( 1 ); }
for( int i = 0; i < exponent; ++i )
{
- if( ulimit / factor >= result ) result *= factor;
+ if( ( result >= 0 && LLONG_MAX / factor >= result ) ||
+ ( result < 0 && LLONG_MIN / factor <= result ) ) result *= factor;
else { errno = ERANGE; break; }
}
}
@@ -284,8 +293,8 @@ unsigned long long getnum( const char * const arg,
if( errno )
{
if( verbosity >= 0 )
- std::fprintf( stderr, "%s: Numerical argument out of limits [%s,%s] "
- "in option '%s'.\n", program_name, format_num3( llimit ),
+ 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( 1 );
}
@@ -315,51 +324,61 @@ void set_mode( Program_mode & program_mode, const Program_mode new_mode )
}
-void set_mtime( long long & mtime, const char * arg, const char * const pn )
+// parse time as 'long long' even if time_t is 32-bit
+long long parse_mtime( const char * arg, const char * const pn )
{
- if( *arg == '@' )
- { mtime = getnum( arg + 1, pn, 0, ( 1ULL << 33 ) - 1 ); return; }
+ if( *arg == '@' ) return getnum( arg + 1, pn );
else if( *arg == '.' || *arg == '/' )
{
struct stat st;
- if( stat( arg, &st ) == 0 ) { mtime = st.st_mtime; return; }
- show_file_error( arg, "Can't stat", errno ); std::exit( 1 );
+ if( stat( arg, &st ) == 0 ) return st.st_mtime;
+ show_file_error( arg, "Can't stat mtime reference file", errno );
+ std::exit( 1 );
}
- else // format 'YYYY-MM-DD HH:MM:SS'
+ else // format '[-]YYYY-MM-DD[[[<separator>HH]:MM]:SS]'
{
- unsigned y, mo, d, h, m, s;
- const int n = std::sscanf( arg, "%u-%u-%u %u:%u:%u",
- &y, &mo, &d, &h, &m, &s );
- if( n == 6 && y >= 1970 && mo >= 1 )
+ long long y; // long long because 2147483648-01-01 overflows int
+ unsigned mo, d, h, m, s;
+ char sep;
+ const int n = std::sscanf( arg, "%lld-%u-%u%c%u:%u:%u",
+ &y, &mo, &d, &sep, &h, &m, &s );
+ if( n >= 3 && n <= 7 && n != 4 && ( n == 3 || sep == ' ' || sep == 'T' ) )
{
- struct tm t;
- t.tm_year = y - 1900; t.tm_mon = mo - 1; t.tm_mday = d;
- t.tm_hour = h; t.tm_min = m; t.tm_sec = s; t.tm_isdst = -1;
- mtime = std::mktime( &t ); if( mtime >= 0 ) return;
+ if( y >= INT_MIN + 1900 && y <= INT_MAX && mo >= 1 && mo <= 12 )
+ {
+ struct tm t;
+ t.tm_year = y - 1900; t.tm_mon = mo - 1; t.tm_mday = d;
+ t.tm_hour = ( n >= 5 ) ? h : 0; t.tm_min = ( n >= 6 ) ? m : 0;
+ t.tm_sec = ( n >= 7 ) ? s : 0; t.tm_isdst = -1;
+ errno = 0;
+ const long long mtime = std::mktime( &t );
+ if( mtime != -1 || errno == 0 ) return mtime; // valid datetime
+ }
+ show_option_error( arg, "Date out of limits in", pn ); std::exit( 1 );
}
}
- show_error( "Invalid mtime.", 0, true ); std::exit( 1 );
+ show_option_error( arg, "Unknown date format in", pn ); std::exit( 1 );
}
-void set_owner( int & owner, const char * const arg, const char * const pn )
+long long parse_owner( const char * const arg, const char * const pn )
{
const struct passwd * const pw = getpwnam( arg );
- if( pw ) owner = pw->pw_uid;
- else if( std::isdigit( (unsigned char)arg[0] ) )
- owner = getnum( arg, pn, 0, INT_MAX );
- else if( std::strcmp( arg, "root" ) == 0 ) owner = 0;
- else { show_file_error( arg, "Invalid owner" ); std::exit( 1 ); }
+ if( pw ) return pw->pw_uid;
+ if( std::isdigit( (unsigned char)arg[0] ) )
+ return getnum( arg, pn, 0, LLONG_MAX );
+ if( std::strcmp( arg, "root" ) == 0 ) return 0;
+ show_option_error( arg, "Invalid owner in", pn ); std::exit( 1 );
}
-void set_group( int & group, const char * const arg, const char * const pn )
+long long parse_group( const char * const arg, const char * const pn )
{
const struct group * const gr = getgrnam( arg );
- if( gr ) group = gr->gr_gid;
- else if( std::isdigit( (unsigned char)arg[0] ) )
- group = getnum( arg, pn, 0, INT_MAX );
- else if( std::strcmp( arg, "root" ) == 0 ) group = 0;
- else { show_file_error( arg, "Invalid group" ); std::exit( 1 ); }
+ if( gr ) return gr->gr_gid;
+ if( std::isdigit( (unsigned char)arg[0] ) )
+ return getnum( arg, pn, 0, LLONG_MAX );
+ if( std::strcmp( arg, "root" ) == 0 ) return 0;
+ show_option_error( arg, "Invalid group in", pn ); std::exit( 1 );
}
} // end namespace
@@ -378,7 +397,7 @@ int open_instream( const std::string & name )
return -1; }
struct stat st; // infd must not be a directory
if( fstat( infd, &st ) == 0 && S_ISDIR( st.st_mode ) )
- { show_file_error( name.c_str(), "Is a directory." );
+ { show_file_error( name.c_str(), "Can't read. Is a directory." );
close( infd ); return -1; }
return infd;
}
@@ -397,9 +416,7 @@ int open_outstream( const std::string & name, const bool create,
const char * msg = !create ? "Error opening file" :
( ( errno == EEXIST ) ? "Skipping file" : "Can't create file" );
if( !rbufp ) show_file_error( name.c_str(), msg, errno );
- else
- snprintf( (*rbufp)(), (*rbufp).size(), "%s: %s: %s\n", name.c_str(),
- msg, std::strerror( errno ) );
+ else format_file_error( *rbufp, name.c_str(), msg, errno );
}
return outfd;
}
@@ -432,6 +449,41 @@ void show_error( const char * const msg, const int errcode, const bool help )
}
+bool format_error( Resizable_buffer & rbuf, const int errcode,
+ const char * const format, ... )
+ {
+ if( verbosity < 0 ) { rbuf.resize( 1 ); rbuf()[0] = 0; return false; }
+ va_list args;
+ for( int i = 0; i < 2; ++i ) // resize rbuf if not large enough
+ {
+ int len = snprintf( rbuf(), rbuf.size(), "%s: ", program_name );
+ if( len >= (int)rbuf.size() && !rbuf.resize( len + 1 ) ) break;
+ va_start( args, format );
+ len += vsnprintf( rbuf() + len, rbuf.size() - len, format, args );
+ va_end( args );
+ if( len >= (int)rbuf.size() && !rbuf.resize( len + 1 ) ) break;
+ if( errcode <= 0 ) rbuf()[len++] = '\n';
+ else len += snprintf( rbuf() + len, rbuf.size() - len, ": %s\n",
+ std::strerror( errcode ) );
+ if( len < (int)rbuf.size() || !rbuf.resize( len + 1 ) ) break;
+ }
+ return true;
+ }
+
+
+void print_error( const int errcode, const char * const format, ... )
+ {
+ if( verbosity < 0 ) return;
+ va_list args;
+ std::fprintf( stderr, "%s: ", program_name );
+ va_start( args, format );
+ std::vfprintf( stderr, format, args );
+ va_end( args );
+ if( errcode <= 0 ) std::fputc( '\n', stderr );
+ else std::fprintf( stderr, ": %s\n", std::strerror( errcode ) );
+ }
+
+
void format_file_error( std::string & estr, const char * const filename,
const char * const msg, const int errcode )
{
@@ -442,6 +494,19 @@ void format_file_error( std::string & estr, const char * const filename,
estr += '\n';
}
+bool format_file_error( Resizable_buffer & rbuf, const char * const filename,
+ const char * const msg, const int errcode )
+ {
+ if( verbosity < 0 ) { rbuf.resize( 1 ); rbuf()[0] = 0; return false; }
+ for( int i = 0; i < 2; ++i ) // resize rbuf if not large enough
+ {
+ const int len = snprintf( rbuf(), rbuf.size(), "%s: %s: %s%s%s\n",
+ program_name, filename, msg, ( errcode > 0 ) ? ": " : "",
+ ( errcode > 0 ) ? std::strerror( errcode ) : "" );
+ if( len < (int)rbuf.size() || !rbuf.resize( len + 1 ) ) break;
+ }
+ return true;
+ }
void show_file_error( const char * const filename, const char * const msg,
const int errcode )
@@ -467,7 +532,7 @@ int main( const int argc, const char * const argv[] )
enum { opt_ano = 256, opt_aso, opt_bso, opt_chk, opt_crc, opt_dbg, opt_del,
opt_dso, opt_exc, opt_grp, opt_hlp, opt_id, opt_kd, opt_mti, opt_nso,
- opt_out, opt_own, opt_per, opt_sol, opt_un, opt_wn };
+ opt_ofl, opt_out, opt_own, opt_per, opt_sol, opt_un, opt_wn };
const Arg_parser::Option options[] =
{
{ '0', 0, Arg_parser::no },
@@ -513,6 +578,7 @@ int main( const int argc, const char * const argv[] )
{ opt_crc, "missing-crc", Arg_parser::no },
{ opt_mti, "mtime", Arg_parser::yes },
{ opt_nso, "no-solid", Arg_parser::no },
+ { opt_ofl, "ignore-overflow", Arg_parser::no },
{ opt_out, "out-slots", Arg_parser::yes },
{ opt_own, "owner", Arg_parser::yes },
{ opt_per, "permissive", Arg_parser::no },
@@ -571,8 +637,8 @@ int main( const int argc, const char * const argv[] )
case 'V': show_version(); return 0;
case 'x': set_mode( cl_opts.program_mode, m_extract ); break;
case 'z': set_mode( cl_opts.program_mode, m_compress ); z_pn = pn; break;
- case opt_ano: set_owner( cl_opts.owner, "root", pn );
- set_group( cl_opts.group, "root", pn ); break;
+ case opt_ano: cl_opts.uid = parse_owner( "root", pn );
+ cl_opts.gid = parse_group( "root", pn ); break;
case opt_aso: cl_opts.solidity = asolid; break;
case opt_bso: cl_opts.solidity = bsolid; break;
case opt_crc: cl_opts.missing_crc = true; break;
@@ -581,14 +647,16 @@ int main( const int argc, const char * const argv[] )
case opt_del: set_mode( cl_opts.program_mode, m_delete ); break;
case opt_dso: cl_opts.solidity = dsolid; break;
case opt_exc: Exclude::add_pattern( sarg ); break;
- case opt_grp: set_group( cl_opts.group, arg, pn ); break;
+ case opt_grp: cl_opts.gid = parse_group( arg, pn ); break;
case opt_hlp: show_help( num_online ); return 0;
case opt_id: cl_opts.ignore_ids = true; break;
case opt_kd: cl_opts.keep_damaged = true; break;
- case opt_mti: set_mtime( cl_opts.mtime, arg, pn ); break;
+ case opt_mti: cl_opts.mtime = parse_mtime( arg, pn );
+ cl_opts.mtime_set = true; break;
case opt_nso: cl_opts.solidity = no_solid; break;
+ case opt_ofl: cl_opts.ignore_overflow = true; break;
case opt_out: cl_opts.out_slots = getnum( arg, pn, 1, 1024 ); break;
- case opt_own: set_owner( cl_opts.owner, arg, pn ); break;
+ case opt_own: cl_opts.uid = parse_owner( arg, pn ); break;
case opt_per: cl_opts.permissive = true; break;
case opt_sol: cl_opts.solidity = solid; break;
case opt_un: cl_opts.level = -1; break;
@@ -597,7 +665,7 @@ int main( const int argc, const char * const argv[] )
}
} // end process options
- if( cl_opts.program_mode != m_compress && cl_opts.output_filename.size() )
+ if( cl_opts.program_mode != m_compress && o_pn )
{
if( verbosity >= 0 )
std::fprintf( stderr, "%s: Option '%s' can only be used with "
@@ -611,6 +679,14 @@ int main( const int argc, const char * const argv[] )
program_name, f_pn, z_pn );
return 1;
}
+ if( cl_opts.program_mode == m_compress &&
+ ( cl_opts.level < 0 || cl_opts.level > 9 ) )
+ {
+ if( verbosity >= 0 )
+ std::fprintf( stderr, "%s: Option '--uncompressed' can't be used with '%s'.\n",
+ program_name, z_pn );
+ return 1;
+ }
#if !defined LZ_API_VERSION || LZ_API_VERSION < 1012 // compile-time test
#error "lzlib 1.12 or newer needed."
@@ -624,12 +700,17 @@ int main( const int argc, const char * const argv[] )
setmode( STDOUT_FILENO, O_BINARY );
#endif
+ if( cl_opts.data_size <= 0 && cl_opts.level >= 0 && cl_opts.level <= 9 )
+ {
+ if( cl_opts.level == 0 ) cl_opts.data_size = 1 << 20;
+ else cl_opts.data_size = 2 * option_mapping[cl_opts.level].dictionary_size;
+ }
if( cl_opts.num_workers < 0 ) // 0 disables multi-threading
cl_opts.num_workers = std::min( num_online, max_workers );
switch( cl_opts.program_mode )
{
- case m_none: show_error( "Missing operation.", 0, true ); return 1;
+ case m_none: show_error( "Missing operation.", 0, true ); return 1;
case m_append:
case m_create: return encode( cl_opts );
case m_compress: return compress( cl_opts );