I am sorry, but this is not what I have meant.
Somebody (not me, I immediatly fell on my knees and sweared that
that I will look at my fingers while I am root) removed a big portion
of the entire filesystem by entering
rm -r /some_path/ *
The white space causes rm to remove all files recursivly under the current
directory tree. If there were only one or two files I could do it
from the command line by modifying the inodes. This would be a
a better (forgive me, gods of perl!) way, because the files would be
undeleted as they are (I am not to good in reading binary files).
Unfortunatly we are talking about 4000 files. Finding borders
takes more time than removing the deletion time of inodes.
So the question was:
Is there any way to manipulate inode-information with perl?
I am looking for something like
use CodeFromTheGodsofExt2fs;
my %inode = getInode($InodeNumber);
$inode{"DeletionTime"}=0;
$inode{"Link"}=1;
writeInode(%Inode);
This could be done several (4000) times to recover all
files that were deleted until that guy (not me I swear it)
realized that Control-C would be a good idea.
The harddisk stayed untouched, so the last command which accessed
the hdd was the rm -r - command.
Regards ...
P.S.: I wrote my previous posting anonymous NOT because I am that guy
(nope, it was not me - I have aliased rm to rm -i on my machines) but because
the enemy-machine did not accept cookies at all. | [reply]
[d/l]
[select] |
One way could be to patch the debugfs-sourcecode, but this
is too dirty for me.
Regards...
| [reply] |