Closed GoogleCodeExporter closed 8 years ago
Original comment by penguin.lin
on 18 Jan 2011 at 6:31
Original comment by penguin.lin
on 3 Mar 2011 at 10:10
For this enhancement, I am thinking about the solution to extract a hardlink
file which underlying object is NOT in the same fileset. Is there any
suggestions?
Original comment by penguin.lin
on 3 Mar 2011 at 2:24
For hardlinks, I suggest extracting the target file with the name of the
requested file. For softlinks, I'd extract the soft-link as is. Do you think
this is feasible?
Original comment by HIGH...@gmail.com
on 4 Mar 2011 at 8:17
In my current implementation, all the hardlinks will be listed into a list, and
they will be extracted after all others files are extracted.
For hardlinks, there are some method can be done when the target file is NOT
existed:
1) Linking the files and ignoring the error.
2) Extracting the content of target file.
3) Extracting the content of target file, and scan & update the hardlink object
in the list (update their equivalent object to the one which was extracted
successfully).
4) Extracting the target contest into the temporary directory, linking them,
then delete these temporary files after all hardlinks are processed.
Original comment by penguin.lin
on 7 Mar 2011 at 6:22
Original comment by penguin.lin
on 8 Mar 2011 at 10:10
Consider the following case with 2 files a and b which are hardlinked.
If a was the original file and b was hardlinked afterwards or if a is a
hardlink to b, the result for extracting only 'a' would be different although
conceptually (2 directory entries, 1 inode, 1 file) they shouldn't be.
So I believe option 2 is the only sane option.
Original comment by HIGH...@gmail.com
on 8 Mar 2011 at 7:16
I will start committing the codes into the trunk,
and these enhancement will be set as fixed in the several days.
Original comment by penguin.lin
on 4 May 2011 at 6:39
Original comment by penguin.lin
on 7 Jul 2011 at 10:19
Original issue reported on code.google.com by
HIGH...@gmail.com
on 13 Jan 2011 at 3:24