Datasets:
HDTenEightyP
/
CompuServe

Dataset card Data Studio Files
xet
Community
text
stringlengths
0
99.6k
ARC VERSION 2.20 PAGE - 40
To read in a Huffman code, reset your test code to all
zeros, start at the beginning of the table, and shift in one
bit at a time from the squeezed file into the test code. The
Huffman input routine would be something like this:
test = 00000000 00000000 00000000
offset = 0
1.0 shift bit into test
1.1 if length(offset) < length of test then error
(can't happen)
if length(offset) > length of test then 1.0
if test=code(offset) then output ascii(offset)
and exit
offset=offset+1
if offset > number of codes then error
(can't happen)
goto 1.1
The routine works because no code can be a substring
of another code. That is, if one of the codes is "01" then
any other code starting with the bit sequence "01" cannot
occur.
This isn't the most efficient way to do it, but it
works! When a Huffman code that has the maximum length comes
in, the routine has to test each and every one of the codes
in the table! However, the most frequently occuring
characters in the file will occur near the beginning of the
table, so it isn't all that bad.
When un-squashing a file, you proceed exactly the same
way, except once you unsqueeze it you have to un-pack it as
well. The control character for packing is always $fe in
squashed or crunched files.
After the header, the crunched file appears as a
sequence of codes. The codes are originally 9 bits in length
and grow to a length of 12 bits. The special reserved code
of 256 appears as the end of file marker. Files that have
been crunched in one pass are identical to files that have
been crunched in two passes, except that the files length
and checksum will appear to be zero in the header for 1 pass
crunched files. If the file was crunched in one pass, then
the original length and checksum follow the end of file
marker.
** Preliminary Instructions **
For
D U A L L O G
By
Steve $ Shubitz
[75125,233]
This program is for use on the 4040
drive or it's dual drive eqivalent. It
is public domain and was originally
wriiten for the PET by an unknown
person. Dual Log is Public Domain and
may not be sold. This program is
designed to keep track of what disks
contain a certain file,disk ID's in use
and the ability to "crunch" your disk
collection.
Please start by Headering a new disk
and then copy Dual Log to this disk
under any name you wish. This disk
should be dicated to Dual Log and not
have any other programs on it. Next
just insert the disk in drive 0 (that's
zero) and Load"DUAL LOG",8. For this
program to work properly all file names
should be in uppercase only and disks
with "funny" directories should be
avoided.
You will know see the first of two
menues. If you select (1) you should
insert the disk to be logged in drive 1
and follow the prompts. The index file
and master file will be written to
drive 0. If you select (3) you may step
through all the disks on file by
tering "*" at the Enter Disk Name
prompt. If you select (4) and you are
looking for a program by the name of
"MR T'S SUPPORT" when you are prompted
for the file name to search for you