Merging upstream version 0.17.

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

1 files changed, 116 insertions, 32 deletions

diff --git a/delete_lz.cc b/delete_lz.cc
index 6888520..961645f 100644
--- a/delete_lz.cc
+++ b/delete_lz.cc
@@ -1,18 +1,18 @@
-/*  Tarlz - Archiver with multimember lzip compression
-    Copyright (C) 2013-2019 Antonio Diaz Diaz.
+/* Tarlz - Archiver with multimember lzip compression
+   Copyright (C) 2013-2020 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 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.
+   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/>.
+   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
@@ -31,18 +31,103 @@
 #include <lzlib.h>
 
 #include "arg_parser.h"
-#include "lzip_index.h"
 #include "tarlz.h"
+#include "lzip_index.h"
+
+
+/* Read 'size' decompressed bytes from the archive.
+   Return value: 0 = OK, 1 = damaged member, 2 = fatal error. */
+int archive_read_lz( LZ_Decoder * const decoder, const int infd,
+                     long long & file_pos, const long long member_end,
+                     const long long cdata_size, uint8_t * const buf,
+                     const int size, const char ** msg )
+  {
+  int sz = 0;
+
+  while( sz < size )
+    {
+    const int rd = LZ_decompress_read( decoder, buf + sz, size - sz );
+    if( rd < 0 )
+      { *msg = LZ_strerror( LZ_decompress_errno( decoder ) ); return 1; }
+    if( rd == 0 && LZ_decompress_finished( decoder ) == 1 )
+      { *msg = end_msg; return 2; }
+    sz += rd;
+    if( sz < size && LZ_decompress_write_size( decoder ) > 0 )
+      {
+      const long long ibuf_size = 16384;
+      uint8_t ibuf[ibuf_size];
+      const long long rest = ( file_pos < member_end ) ?
+        member_end - file_pos : cdata_size - file_pos;
+      const int rsize = std::min( LZ_decompress_write_size( decoder ),
+                                  (int)std::min( ibuf_size, rest ) );
+      if( rsize <= 0 ) LZ_decompress_finish( decoder );
+      else
+        {
+        const int rd = preadblock( infd, ibuf, rsize, file_pos );
+        if( LZ_decompress_write( decoder, ibuf, rd ) != rd )
+          internal_error( "library error (LZ_decompress_write)." );
+        file_pos += rd;
+        if( rd < rsize )
+          {
+          LZ_decompress_finish( decoder );
+          if( errno ) { *msg = "Error reading archive"; return 2; }
+          }
+        }
+      }
+    }
+  return 0;
+  }
+
+
+int parse_records_lz( LZ_Decoder * const decoder, const int infd,
+                      long long & file_pos, const long long member_end,
+                      const long long cdata_size, long long & data_pos,
+                      Extended & extended, const Tar_header header,
+                      Resizable_buffer & rbuf, const char ** msg,
+                      const bool permissive )
+  {
+  const long long edsize = parse_octal( header + size_o, size_l );
+  const long long bufsize = round_up( edsize );
+  if( edsize <= 0 || edsize >= 1LL << 33 || bufsize >= INT_MAX )
+    return 1;					// overflow or no extended data
+  if( !rbuf.resize( bufsize ) ) return 1;	// extended records buffer
+  int retval = archive_read_lz( decoder, infd, file_pos, member_end,
+                                cdata_size, (uint8_t *)rbuf(), bufsize, msg );
+  if( retval == 0 )
+    { if( extended.parse( rbuf(), edsize, permissive ) ) data_pos += bufsize;
+      else retval = 2; }
+  return retval;
+  }
+
+
+int skip_member_lz( LZ_Decoder * const decoder, const int infd,
+                    long long & file_pos, const long long member_end,
+                    const long long cdata_size, long long & data_pos,
+                    long long rest, const char ** msg )
+  {
+  const int bufsize = 32 * header_size;
+  uint8_t buf[bufsize];
+  while( rest > 0 )				// skip tar member
+    {
+    const int rsize = ( rest >= bufsize ) ? bufsize : rest;
+    const int ret = archive_read_lz( decoder, infd, file_pos, member_end,
+                                     cdata_size, buf, rsize, msg );
+    if( ret != 0 ) return ret;
+    data_pos += rsize;
+    rest -= rsize;
+    }
+  return 0;
+  }
 
 
 /* Deleting from a corrupt archive must not worsen the corruption. Stop and
-   tail-copy as soon as corruption is found. */
-int delete_members_lz( const char * const archive_namep,
-                       const Arg_parser & parser,
+   tail-copy as soon as corruption is found.
+*/
+int delete_members_lz( const Cl_options & cl_opts,
+                       const char * const archive_namep,
                        std::vector< char > & name_pending,
                        const Lzip_index & lzip_index,
-                       const int filenames, const int infd, const int outfd,
-                       const bool missing_crc, const bool permissive )
+                       const int infd, const int outfd )
   {
   Resizable_buffer rbuf;
   LZ_Decoder * const decoder = LZ_decompress_open();
@@ -87,7 +172,7 @@ int delete_members_lz( const char * const archive_namep,
         {
         if( block_is_zero( header, header_size ) )		// EOF
           {
-          if( prev_extended && !permissive )
+          if( prev_extended && !cl_opts.permissive )
             { show_file_error( archive_namep, fv_msg1 ); retval = 2; }
           goto done;
           }
@@ -100,48 +185,47 @@ int delete_members_lz( const char * const archive_namep,
       const Typeflag typeflag = (Typeflag)header[typeflag_o];
       if( typeflag == tf_global )
         {
-        if( prev_extended && !permissive )
+        if( prev_extended && !cl_opts.permissive )
           { show_file_error( archive_namep, fv_msg2 ); retval = 2; goto done; }
         Extended dummy;		// global headers are parsed and ignored
         retval = parse_records_lz( decoder, infd, file_pos, member_end,
                                    cdata_size, data_pos, dummy, header,
                                    rbuf, &msg, true );
         if( retval == 0 ) continue;
-        show_file_error( archive_namep, msg ? msg : gblrec_msg );
+        show_file_error( archive_namep, gblrec_msg );
         goto done;
         }
       if( typeflag == tf_extended )
         {
-        if( prev_extended && !permissive ) { msg = fv_msg3; retval = 2; }
+        if( prev_extended && !cl_opts.permissive )
+          { msg = fv_msg3; retval = 2; }
         else retval = parse_records_lz( decoder, infd, file_pos, member_end,
                                         cdata_size, data_pos, extended, header,
-                                        rbuf, &msg, permissive );
-        if( retval == 0 && !extended.crc_present() && missing_crc )
+                                        rbuf, &msg, cl_opts.permissive );
+        if( retval == 0 && !extended.crc_present() && cl_opts.missing_crc )
           { msg = mcrc_msg; retval = 2; }
         if( retval == 0 ) { prev_extended = true; continue; }
-        show_file_error( archive_namep, msg ? msg : extrec_msg );
+        show_file_error( archive_namep, extrec_msg );
         goto done;
         }
       prev_extended = false;
 
       extended.fill_from_ustar( header );	// copy metadata from header
 
-      long long rest = extended.file_size();
-      const int rem = rest % header_size;
-      if( rem ) rest += header_size - rem;		// padding
+      long long rest = round_up( extended.file_size() );  // size + padding
       if( data_pos + rest >= mdata_end ) data_pos += rest;
       else						// skip tar member
         if( ( retval = skip_member_lz( decoder, infd, file_pos, member_end,
                          cdata_size, data_pos, rest, &msg ) ) != 0 )
           goto done;
 
-      if( !check_skip_filename( parser, name_pending, extended.path().c_str(),
-                                filenames ) )		// delete tar member
+      // delete tar member
+      if( !check_skip_filename( cl_opts, name_pending, extended.path().c_str() ) )
         {
         // verify that members match
         if( member_begin != mdata_pos || data_pos != mdata_end )
           { show_file_error( extended.path().c_str(),
-                             "Can't delete: not individually compressed." );
+                             "Can't delete: not compressed individually." );
             retval2 = 2; extended.reset(); continue; }
         if( !show_member_name( extended, header, 1, rbuf ) )
           { retval = 1; goto done; }
@@ -163,6 +247,6 @@ done:
   if( LZ_decompress_close( decoder ) < 0 && !retval )
     { show_error( "LZ_decompress_close failed." ); retval = 1; }
   // tail copy keeps trailing data
-  return tail_copy( archive_namep, parser, name_pending, lzip_index,
+  return tail_copy( archive_namep, cl_opts.parser, name_pending, lzip_index,
                     istream_pos, infd, outfd, retval );
   }