text
stringlengths
0
99.6k
GET previous track,sector and change
link to 0,255
PRINT "archive has been repaired."
ENDIF
ENDIF
DONE fixing archive.
ARC VERSION 2.20 PAGE - 48
ARC TRICKS AND TIPS.
1) Manipulating Archives.
If you were to use a disk doctor to look at an archive,
you would notice that every file in the archive begins at
the beginning of a CBM disk sector and ends at the end of a
sector. This fact can be used to advantage, and will be in
future ARCs, in a number of ways.
First of all, it makes it possible to break an archive
up into a number of smaller archives simply by manipulating
the directory and a few track,sector links. Since it is not
necessary to move any data, this should be a relatively fast
proceedure. You could split an archive in two with a disk
doctor as follows:
1) create a dummy file on the same disk as the archive.
2) scratch the dummy file.
3) locate the archive and follow the sector links until
you see the filename of the archive entry which is
to be the first filename in the second archive.
4) jot down the track and sector you are looking at.
5) backup to the previous sector and change the link
to 0,255
6) get back to the directory and locate the dummy
file from (1)
7) change the filetype from (6) to program (130)
8) change the track,sector to the ones you jotted
down in (4)
You now have two separate archives. The block count for
these files will not be correct in the directory, and should
be fixed, but no errors should arise if they are not.
A similar proceedure could be used to delete an archive
entry. Simply manipulate the sector links to point around
the entry to be deleted and then validate the disk.
You could also re-arrange the order of the archive
entries in an archive, append one archive to another,
transfer an archive entry to another archive and so on and
so forth... All of this could be done without having to move
a single byte of data!
ARC VERSION 2.20 PAGE - 49
2) Random Access Archives.
An archive could be made random access in each of two
ways.
The first method would be to follow the track,sector
links and read the number of CBM disk blocks in each archive
entry. This information could then be used to calculate
where each archive entry begins.
A second method would be to convert the archive to a
relative file. Simply create the archive by normal
proceedures, and then use ARC to archive the archive you
just created. Now load and run your favorite disk sector
editor and view the first sector of the archive of the
archive. Change the file's type to "R" for RELative in the
archive header. (You will see the filetype as a "P" two
bytes before the filename) Now all you have to do is
unarchive it and it will become a relative file with record
length 254 and can be accessed randomly.
But why would you want a random access archive
anyway???
We plan to write a simple un-squash routine that can be
accessed as a subroutine by a BASIC or Machine Language
programmer. Unfortuneately the CRUNCH routines require a
great deal of workspace (about 28K) and would not be of much
use to most applications. However, a small routine could
reside inside about 2K bytes which would un-pack, un-