Re: [Bug-tar] bug in pax_dump_header_1()?

[Sergey Poznyakoff]

Bernd Schubert <address@hidden> ha escrit:

> - pax_dump_header_1() does not check either to use a single block only. For 
> complex files with lots of different sparse blocks and so a large sparse map, 
Yes, this sounds reasonable.  Allow me some time to revise it.

> - May pax_dump_header_1() should pad the block with zeros?

No, in that particular case there's no need to.

> - Silent error on extracting the archive and leaving GNUSparseFile.* files 
> around.

That's not a silent error.  That's what extended header blocks are for.
The standard says that

   if this archive is read by a pax utility conforming to the ISO
   POSIX-2:1993 standard, the header block fields are used to create a
   regular file that contains the extended header records as data. 

> Does it not even detect those are sparse files?

Tar did not detect this because the archive was malformed and its
further actions were in compliance with the standard, which says to
extract the data as if they were usual ustar archive member.

Regards,
Sergey