Instruction manual

If the current keystroke cor-
responds not to a move command, but to
a Plot command, the program sets the
cursor disable flag, FLAG, calls the plot
subroutine and then branches back
to
get the next keystroke (all of this is done
in step
680). The Quit command forces a
branch
to a routine that closes out the
current byte (starting at step 1080), adds
a record mark (step
1170) and draws
thew completed shape (step
1170). At
this juncture, you are asked a series of
questions, the answers to which will
allow you to:
1) forget the current shape and go
back and try again
without re-accessing
the current shape file from disk;
2) keep the current shape, update
the shape file directory and start a new
shape;
3) forget the whole thing-add no
new shapes to the file and quit;
4)load an updated shape file
to
disk and quit.
These alternatives will help you to avoid
filling up
the shape table with unwanted
shapes, and
allow you to experiment
without being forced to save all of your
experiments.
The closing
out of the current byte
preparatory to ending the current shape
definition (step 1080) poses a problem if
the last keystroke is a Plot command
because a P command alone does not
generate a vector. There is nothing
to
store after a final P command, unless it
is followed by some sort of move. The
problem is handled in steps 1100-1140 by
adding an arbitrary up-move after a final
Plot command
to generate a plot-then-
move-up vector. (Note
that in the illustra-
tion Figure
2, the concluding vector is a
plot-then-move-down. This was done for
the sake
of clarity in drawing only. The
point is mentioned in case some
unusually perceptive reader notices that
the foregoing description does not tally
with the example in Figure
2). The final
vector is either added
to the current
byte, in which it will appear as the only
entry. If the last keystroke prior
to clos-
ing the current shape table is anything
other than a Plot command, the current
byte can
be closed out immediately
without further ado.
The erase command has the very
limited
capability of erasing the last Plot
command only. As discussed before, a
Plot command alone does
not result in
formation
of a vector until it is followed
by a command. Therefore, if a Plot com-
mand is issued in error and no move
command
follows it, no vector will be
generated and the shape table remains
unchanged at this point. It is therefore
possible
to undo the Plot command
simply,
without the complication of
19:16
analyzing the last byte for returning to
the
state that preceeded the mistaken
command (and it would be
complicated! i). At the point at which the
Plot command is mistakenly issued,
KSVE$ has a certain value. If we wish
to
go back to the condition prior to the
mistaken Plot command, we
must
restore that value to KSVE$ so that when
the correct command is issued it is pro-
perly interpreted when KSVE$ is examin-
ed subsequently. The character required
for this purpose lies waiting in KI$. Thus,
the erase command loads
this previous
value
into KSVE$ and "unplots" the in-
correct
plotting circle by re-plotting with
the color
"black" (HCOlOR = 0 in step
720). Note that because of these limita-
tions, no plot command can be undone
after a move has been made.
Byte assembly using the 3-bit codes
(stored currently in SYMBOL)
occurs in
780-980. The variable CYCLE keeps track
of the number of 3-bit codes entered into
the current byte (called BYTE in the pro-
gram).
After the second 3-bit code is
loaded
into BYTE (step 820) a check is
made (step 840)
to see if the byte is less
than
8; if it is, we know that the byte con-
tains an unrecognizable move-up vector
in the
left five bits. In that case, a dum-
my move-right 3-bit code is inserted
into
the byte, the byte is stored (step 860) and
a new byte
is formed consisting of the
required move-up
(000) followed by a
dummy move-left
(110) to compensate
for
the dummy move-right. The resulting
byte contains the
bit string 0001 1000,
decimal
24, generated in step 880.
Statements 950·980 take care of the
cases in which the third 3-bit code is a
plot-then-move code or a move-up only
code, which require that the current byte
be stored, and the current 3-bit code be
loaded into the next byte.
The Display Program
It is likely that your disk or tape will
be replete with shape files tailored to
various uses, now that creating shape
tables is so easy. A convenient display
program
will become essential in order
to find out which shapes are stored
where. The display program
that ac-
complishes this (Figure
8) is an example
of how shape files may be used is a pro-
gram. The program constructs a 6
x 6
grid on the high resolution screen and
displays one shape per grid cell. To iden-
tify
the location of the shapes in the
shape table, each occupied cell carries
the shape index in the upper left-hand
corner. The numerals required for plot-
ting these indices are extracted from a
shape table called NUMERALS
that you
will have
to create at storage location
20000 (decimal) by means
of the shape
creating program. The numerals are
restricted
to a 5 x 7 grid, and are format·
ted as illustrated
by the example in
MICRO
-- The 6502 Journal
Figure 1. Sufficient space is reserved in
the display squares
to accomodate
three-digit numerals from 1 through
255.
"Aha," you ask, "how can 255 shapes be
displayed in a 6 x 6 grid?" The program
provides for paging through the shape
table,
36 shapes at a time. The paging is
activated by
hitting any alphanumeric
key on the APPLE keyboard.
The display program opens
by get-
ting the shape files
that it needs-one
for numerals (step 50) and the table to be
displayed (step 90). Pointers to the
tables are set up (steps
70 and 120).
Starting at step 180, each shape I is ac-
cessed in a
FOR. .. NEXT loop. A grid-
specific index is calculated (step 190) by
taking the current shape index I modulo
36(step
190). For the first shape in each
group
of 36 (I modulo 36 = 1), the screen
is cleared (step
240) and the 6 x 6 grid is
displayed (steps 250-330). The row and
column
positions for the I-th shape in
the grid are found (steps
360, 370). The
shape index is
"unpacked" into its
separate
digits (steps 380-410) and these
digits are plotted in the correct grid cell
in
the upper left-hand corner (steps
430-480). The NUMERALS shape table is
accessed in step
420 by placing the
pointer
to the NUMERALS shape table in
(decimal) addresses 232 and
233, so that
subsequent DRAW commands will refer
to
this table. In similar fashion, when the
shapes to be plotted are required, the
address
of the shape table must be
entered into addresses 232, 233. This
program
illustrates how any number of
shape tables may be used inside a pro-
gram simply
by supplying the correct
pointers at the time that shapes are
to
be DRAWn or XDRAWn.
Parting Words
The 15 x 15 grid used for shape
creation is the largest practical size for
the APPLE screen with space provided
for text. A larger grid can
be accomodat-
ed by eliminating the text area, but this
will compromise the required starting
coordinate input. However, the number
of cells could be increased by decreas-
ing cell size and using a smaller plotting
figure. If you try this, it is convenient
to
select a plotting grid with odd numbers
of X and Y segments so that the central
plotting area falls on a grid square and
not at the intersection
of two grid lines.
This is
of help in centering shapes.
You should also be aware, if it is not
obvious by now,
that the location of a
shape on the grid has no bearing on
where
it plots in high resolution
graphics, except with regard to the in-
itial point
of the shape, which alone
determines
justification. You may use
any convenient subsection
of the full
grid for plotting, and it does not have to
be the same subsection for each shape.
continued on page
19
December, 1979