home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
The Devil's Doorknob BBS Capture (1996-2003)
/
devilsdoorknobbbscapture1996-2003.iso
/
Dloads
/
PROGRAMM
/
FGL112B.ZIP
/
USER06.DOC
< prev
next >
Wrap
Text File
|
1992-10-05
|
55KB
|
1,221 lines
Chapter 6
Graphics Fundamentals
76 Fastgraph User's Guide
Overview
This chapter describes Fastgraph's fundamental graphics routines,
sometimes called graphics primitives. These routines perform such functions
as clearing the screen, drawing points, drawing solid and dashed lines,
drawing closed shapes (polygons, circles, and ellipses), drawing rectangles
(solid and dithered), and filling arbitrary regions. Most of these routines
have no effect in text video modes, but there are a few exceptions, and they
will be noted in the descriptions of those routines.
Clearing the Screen
The Fastgraph routine fg_erase clears the entire screen in any video
mode. In text modes, fg_erase clears the screen by storing a space character
(ASCII 32) with a gray foreground attribute in each character cell. In
graphics modes, fg_erase clears the screen by setting each pixel to zero.
This of course causes each pixel to be displayed its background color. The
fg_erase routine has no arguments.
Clipping
The suppression of graphics outside a pre-defined area is called
clipping. Many of Fastgraph's graphics-oriented routines provide clipping,
either automatically or through a special version of the routine.
Fastgraph includes two routines, fg_setclip and fg_setclipw, to define a
rectangular clipping region. The fg_setclip routine defines the clipping
region in screen space, while the fg_setclipw routine performs the same
function in world space. Each routine takes four arguments: the minimum x,
the maximum x, the minimum y, and the maximum y coordinate of the clipping
region. The arguments are integer quantities for fg_setclip and floating
point quantities for fg_setclipw. For example, the statement
fg_setclip(0,159,0,99);
would define the upper left quadrant of the screen as the clipping region in
a 320 by 200 graphics mode.
An implicit clipping region equal to the entire screen is defined as
part of the fg_setmode routine's initializations. Clipping is not supported
for text modes.
Points
The Fastgraph routine fg_point provides the most basic graphics
operation -- setting a pixel to a specified color. The fg_point routine has
two integer arguments. The first specifies the pixel's x coordinate, and the
second its y coordinate. The pixel is drawn using the current color value,
as specified in the most recent call to fg_setcolor. There is also a world
space version of this routine, fg_pointw, that uses floating point arguments.
Another Fastgraph routine is available for reading a pixel's color
value. The fg_getpixel routine has two integer arguments that specify the
(x,y) coordinates for the pixel of interest. There is no world space version
Chapter 6: Graphics Fundamentals 77
of the fg_getpixel routine, but you can obtain a pixel's color value in world
space by applying the fg_xscreen and fg_yscreen functions to the world space
coordinates and passing the resulting values to fg_getpixel.
Example 6-1 uses the fg_point routine to draw 100 random points in
random colors. It also uses the fg_getpixel routine to insure no two points
are adjacent. The program establishes a graphics video mode with the
fg_automode and fg_setmode routines. Next, it determines the maximum color
value for the selected video mode; note if we used virtual colors (color
indices above the maximum color value), some of the colors would be the
background color and would thus produce invisible points. The main part of
the program is a while loop that first generates a random pair of (x,y)
screen coordinates. It then calls the fg_getpixel routine to check the
pixels at (x,y) and the eight adjacent positions. If none of these pixels
are set, the program generates a random color value and draws a point in that
color. After doing this 100 times, the program waits for a keystroke,
restores the original video mode and screen attributes, and then returns to
DOS.
Example 6-1.
#include <fastgraf.h>
#include <stdlib.h>
void main(void);
void main()
{
int area;
int color, old_color;
int left;
int max_color, max_x, max_y;
int new_mode, old_mode;
int x, y;
old_mode = fg_getmode();
new_mode = fg_automode();
fg_setmode(new_mode);
78 Fastgraph User's Guide
if (new_mode == 4)
max_color = 3;
else if (new_mode == 11 || new_mode == 17)
max_color = 1;
else if (new_mode == 19)
max_color = 255;
else
max_color = 15;
left = 100;
max_x = fg_getmaxx() - 1;
max_y = fg_getmaxy() - 1;
while (left > 0) {
x = rand() % max_x + 1;
y = rand() % max_y + 1;
area = fg_getpixel(x-1,y-1) + fg_getpixel(x,y-1) + fg_getpixel(x+1,y-1)
+ fg_getpixel(x-1,y) + fg_getpixel(x,y) + fg_getpixel(x+1,y)
+ fg_getpixel(x-1,y+1) + fg_getpixel(x,y+1) + fg_getpixel(x+1,y+1);
if (area == 0) {
color = rand() % max_color + 1;
fg_setcolor(color);
fg_point(x,y);
left--;
}
}
fg_waitkey();
fg_setmode(old_mode);
fg_reset();
}
Chapter 6: Graphics Fundamentals 79
The Graphics Cursor
Many of Fastgraph's graphics routines depend on the position of the
graphics cursor as a reference point. For example, Fastgraph includes
routines to draw lines from the graphics cursor position to a specified
position, and the image display routines discussed in Chapter 9 display or
retrieve an image relative to the graphics cursor position. The graphics
cursor is not a cursor in the true sense; it is simply a pair of (x,y)
coordinates with a special meaning. The fg_setmode routine sets the graphics
cursor position to the screen space coordinates (0,0), and the fg_initw
routine sets it to the world space coordinates (0.0,0.0).
Fastgraph includes four routines for changing the graphics cursor
position. The fg_move routine sets it to an absolute screen space position,
while the fg_movew routine sets it to an absolute world space position. The
fg_moverel routine sets the graphics cursor position to a screen space
position relative to its current position. The fg_moverw routine does the
same in world space.
Each of these routines has two arguments that specify the (x,y)
coordinates of the new position. For the screen space routines, the
arguments are integer quantities. For the world space routines, the
arguments are floating point quantities.
You can obtain the screen space coordinates of the graphics cursor
position with the fg_getxpos and fg_getypos routines. These routines have no
arguments and respectively return the x and y coordinates of the graphics
cursor position as the function value. To obtain the world space coordinates
of the graphics cursor position, you can apply the fg_xworld and fg_yworld
functions to the return values of fg_getxpos and fg_getypos.
Solid Lines
Fastgraph includes four routines for drawing solid lines. All four
routines draw lines in the current color value (as determined by the most
recent call to fg_setcolor) and observe the clipping limits. The fg_draw
routine draws a line from the current graphics cursor position to an absolute
screen space position, while the fg_draww routine draws a line to an absolute
world space position. The fg_drawrel routine draws a line from the current
graphics cursor position to a screen space position relative to it. The
fg_drawrw routine does the same in world space.
80 Fastgraph User's Guide
Each of these routines has two arguments that specify the (x,y)
coordinates of the destination position. For the screen space routines, the
arguments are integer quantities. For the world space routines, the
arguments are floating point quantities. In either case the destination
position becomes the new graphics cursor position. This makes it possible to
draw connected lines without calling a graphics cursor movement routine
between successive calls to a line drawing routine.
Examples 6-2 and 6-3 each draw a pair of crossed lines that divide the
screen into quadrants. Example 6-2 does this using the fg_move and fg_draw
routines, while example 6-3 uses the fg_moverel and fg_drawrel routines.
Both examples draw the lines in bright white, the default for color 15 in all
graphics video modes.
Example 6-2. Example 6-3.
#include <fastgraf.h> #include <fastgraf.h>
void main(void); void main(void);
void main() void main()
{ {
int max_x, max_y; int max_x, max_y;
int mid_x, mid_y; int mid_x, mid_y;
int new_mode, old_mode; int new_mode, old_mode;
old_mode = fg_getmode(); old_mode = fg_getmode();
new_mode = fg_automode(); new_mode = fg_automode();
fg_setmode(new_mode); fg_setmode(new_mode);
max_x = fg_getmaxx(); max_x = fg_getmaxx();
max_y = fg_getmaxy(); max_y = fg_getmaxy();
mid_x = max_x / 2; mid_x = max_x / 2;
mid_y = max_y / 2; mid_y = max_y / 2;
fg_setcolor(15); fg_setcolor(15);
fg_move(mid_x,0); fg_move(mid_x,0);
fg_draw(mid_x,max_y); fg_drawrel(0,max_y);
fg_move(0,mid_y); fg_moverel(-mid_x,-mid_y);
fg_draw(max_x,mid_y); fg_drawrel(max_x,0);
fg_waitkey(); fg_waitkey();
fg_setmode(old_mode); fg_setmode(old_mode);
fg_reset(); fg_reset();
} }
Examples 6-4 and 6-5 are variations of example 6-2. Example 6-4 uses
world space rather than screen space to draw the crossed lines. Example 6-5
is the same as example 6-2 except it defines a clipping area to restrict
drawing to the upper left quadrant of the screen. The clipping suppresses
the right half of the horizontal line and the lower half of the vertical
line.
Chapter 6: Graphics Fundamentals 81
Example 6-4. Example 6-5.
#include <fastgraf.h> #include <fastgraf.h>
void main(void); void main(void);
void main() void main()
{ {
int new_mode, old_mode; int max_x, max_y;
int mid_x, mid_y;
old_mode = fg_getmode(); int new_mode, old_mode;
new_mode = fg_automode();
fg_setmode(new_mode); old_mode = fg_getmode();
fg_initw(); new_mode = fg_automode();
fg_setworld(-10.0,10.0,-10.0,10.0); fg_setmode(new_mode);
fg_setcolor(15); max_x = fg_getmaxx();
fg_movew(0.0,10.0); max_y = fg_getmaxy();
fg_draww(0.0,-10.0); mid_x = max_x / 2;
fg_movew(-10.0,0.0); mid_y = max_y / 2;
fg_draww(10.0,0.0);
fg_waitkey(); fg_setclip(0,mid_x,0,mid_y);
fg_setmode(old_mode); fg_setcolor(15);
fg_reset(); fg_move(mid_x,0);
} fg_draw(mid_x,max_y);
fg_move(0,mid_y);
fg_draw(max_x,mid_y);
fg_waitkey();
fg_setmode(old_mode);
fg_reset();
}
82 Fastgraph User's Guide
Dashed Lines
Fastgraph includes four routines for drawing dashed lines. All four
routines draw lines in the current color value (as determined by the most
recent call to fg_setcolor) and observe the clipping limits. The fg_dash
routine draws a dashed line from the current graphics cursor position to an
absolute screen space position, while the fg_dashw routine draws a dashed
line to an absolute world space position. The fg_dashrel routine draws a
dashed line from the current graphics cursor position to a screen space
position relative to it. The fg_dashrw routine does the same in world space.
Each of these routines has three arguments. The first two specify the
(x,y) coordinates of the destination position. For the screen space
routines, these arguments are integer quantities. For the world space
routines, these arguments are floating point quantities. The third argument
is a 16-bit pattern that defines the appearance of the dashed line. Bits
that are set in the pattern produce the visible part of the line, while bits
that are reset produce the invisible part. This pattern is repeated as
necessary to draw the entire line. For example, the pattern value 3333 hex
would produce a dashed line with the first two pixels off, the next two on,
the next two off, and so forth. Similarly, the pattern value FFFF hex would
produce a solid line.
The destination position passed to any of the dashed line routines
becomes the new graphics cursor position. This makes it possible to draw
connected dashed lines without calling a graphics cursor movement routine
between successive calls to a line drawing routine.
Example 6-6 draws a pair of crossed dashed lines that divide the screen
into quadrants. It does this using the fg_move and fg_dash routines and
draws the lines in bright white, the default for color 15 in all graphics
video modes. The dash pattern for each line is 3333 hex, which alternates
two pixels off and on.
Example 6-6.
#include <fastgraf.h>
void main(void);
void main()
{
int max_x, max_y;
int mid_x, mid_y;
int new_mode, old_mode;
old_mode = fg_getmode();
new_mode = fg_automode();
fg_setmode(new_mode);
max_x = fg_getmaxx();
max_y = fg_getmaxy();
mid_x = max_x / 2;
Chapter 6: Graphics Fundamentals 83
mid_y = max_y / 2;
fg_setcolor(15);
fg_move(mid_x,0);
fg_dash(mid_x,max_y,0x3333);
fg_move(0,mid_y);
fg_dash(max_x,mid_y,0x3333);
fg_waitkey();
fg_setmode(old_mode);
fg_reset();
}
Polygons, Circles, and Ellipses
Fastgraph includes routines for drawing three types of closed shapes:
polygons, circles, and ellipses. Both screen space and world space versions
of these routines are available. All the routines for drawing closed shapes
observe the clipping limits.
The fg_polygon routine draws an unfilled polygon in screen space.
Fg_polygon requires an array of integer x coordinates as its first argument,
and an array of integer y coordinates as its second argument. Each (x,y)
coordinate pair from the two arrays is treated as a polygon vertex. In other
words, the x coordinate of the first polygon vertex is the first element of
the x coordinate array, and the y coordinate of the first vertex is the first
element of the y coordinate array. Similarly, the second elements of each
array define the second vertex, and so forth. The third argument for
fg_polygon is an integer quantity that specifies the number of elements in
the two coordinate arrays.
Another routine, fg_polygonw, draws an unfilled polygon in world space.
The fg_polygonw routine is the same as the fg_polygon routine, except its x
and y coordinate arrays must contain floating point values instead of
integers.
The drawing of the polygon begins at the graphics cursor position. The
routine then draws a solid line in the current color to the first vertex,
then to the second vertex, and continues in this manner up to the last
vertex. If necessary, fg_polygon and fg_polygonw draw a line connecting the
last vertex and the original graphics cursor position. This last operation
closes the polygon and effectively leaves the graphics cursor position
unchanged.
Example 6-7 illustrates the use of the fg_polygon routine in the EGA
monochrome or enhanced modes (modes 15 and 16). The program exits if neither
of these video modes are available.
Example 6-7.
#include <fastgraf.h>
#include <stdio.h>
#include <stdlib.h>
void main(void);
#define VERTICES 10
84 Fastgraph User's Guide
int x[] = {200,300,400,400,300,240,160,160,200,200};
int y[] = {100, 80,100,220,320,320,240,200,160,160};
void main()
{
int old_mode;
old_mode = fg_getmode();
if (fg_testmode(16,1))
fg_setmode(16);
else if (fg_testmode(15,1))
fg_setmode(15);
else {
printf("This program requires a 640 x 350 ");
printf("EGA graphics mode.\n");
exit(1);
}
fg_setcolor(1);
fg_polygon(x,y,VERTICES);
fg_waitkey();
fg_setmode(old_mode);
fg_reset();
}
The fg_circle routine draws an unfilled circle in screen space, centered
at the graphics cursor position, using the current color. The routine's only
argument specifies the circle's radius in horizontal screen space units.
Another routine, fg_circlew, draws an unfilled circle where the radius is
measured in horizontal world space units. Both routines leave the graphics
cursor position unchanged.
The fg_ellipse routine draws an unfilled ellipse in screen space,
centered at the graphics cursor position, using the current color. The
routine requires two arguments that respectively specify the length of its
horizontal and vertical semi-axes. In other words, the first argument is the
absolute distance from the center of the ellipse to its horizontal extremity,
and the second argument is the absolute distance from the center of the
ellipse to its vertical extremity. Another routine, fg_ellipsew, draws an
unfilled ellipse in world space. Both routines leave the graphics cursor
position unchanged.
Example 6-8 illustrates the use of the fg_circlew and fg_ellipsew
routines. The program first uses the fg_automode routine to propose a
graphics video mode and then uses the fg_setmode routine to select that video
mode. It then makes color 15 the current color, which by default is bright
white in all color graphics modes and "on" in the monochrome graphics modes.
Next, it establishes a 200-unit by 200-unit world space coordinate system.
The program then uses fg_ellipsew to draw an ellipse and fg_circlew to draw a
circle, both centered at the middle of the screen (which is the origin of our
world space coordinate system). The circle has a radius of 1/16 the width of
the screen (12.5 horizontal world space units), and the ellipse is
horizontally inscribed within the circle.
Chapter 6: Graphics Fundamentals 85
Example 6-9 illustrates the use of the fg_circle and fg_ellipse
routines. It is functionally identical to the program of example 6-8, but it
uses screen space rather than world space coordinates to draw the circle and
ellipse. Note the arguments to the fg_circle and fg_ellipse routines are
dependent on the maximum x and y coordinates of the selected video mode. If
we didn't compute these arguments in this manner, the actual size of the
circle and ellipse would be proportional to the pixel resolution of the video
mode. No such dependency exists when using world space, but we pay a price
for this feature in slightly slower execution speed.
Example 6-8. Example 6-9.
#include <fastgraf.h> #include <fastgraf.h>
void main(void); void main(void);
void main() void main()
{ {
int old_mode; int mid_x, mid_y;
int old_mode;
old_mode = fg_getmode(); int x,y;
fg_setmode(fg_automode());
fg_setcolor(15); old_mode = fg_getmode();
fg_setmode(fg_automode());
fg_initw(); fg_setcolor(15);
fg_setworld(-100.0,100.0,-100.0,100.0);
mid_x = fg_getmaxx() / 2;
fg_movew(0.0,0.0); mid_y = fg_getmaxy() / 2;
fg_ellipsew(12.5,12.5); x = mid_x / 8;
fg_circlew(12.5); y = mid_y / 8;
fg_waitkey();
fg_move(mid_x,mid_y);
fg_setmode(old_mode); fg_ellipse(x,y);
fg_reset(); fg_circle(x);
} fg_waitkey();
fg_setmode(old_mode);
fg_reset();
}
86 Fastgraph User's Guide
Solid Rectangles
Fastgraph includes four routines for drawing solid rectangles, two for
screen space and two for world space, with and without clipping. None of
these routines affect the graphics cursor position.
The fg_rect routine draws a solid rectangle in screen space, without
regard to the clipping limits, using the current color. Fg_rect requires
four integer arguments that respectively define the minimum x, maximum x,
minimum y, and maximum y screen space coordinates of the rectangle. The
minimum coordinates must be less than or equal to the maximum coordinates, or
else the results are unpredictable. The fg_clprect routine is identical in
all respects to the fg_rect routine, except it observes the clipping limits.
The world space versions of the solid rectangle drawing routines are
fg_rectw and fg_clprectw. Like fg_rect and fg_clprect, they require four
arguments that define the extremes of the rectangle, but the arguments are
floating point world space coordinates.
You also can use the fg_rect routine in text modes. When used in a text
mode, fg_rect expects its four arguments to be expressed in character space
(that is, rows and columns) rather than screen space. This means the four
arguments respectively specify the minimum column, maximum column, minimum
row, and maximum row of the rectangle. Fastgraph constructs the rectangle by
storing the solid block character (ASCII decimal value 219) in each character
cell comprising the rectangle. The rectangle is drawn using the current
character attribute, but because the solid block character occupies the
entire character cell, the background component of the attribute is
essentially meaningless.
Example 6-10 demonstrates the use of the fg_rect routine by drawing 200
random-size rectangles in random colors. The program first uses the
fg_automode routine to propose a graphics video mode and then uses the
fg_setmode routine to select that video mode. Next, it determines the
horizontal and vertical screen resolution for the selected video mode, using
the fg_getmaxx and fg_getmaxy routines. The main part of the program is a
for loop that generates a random rectangle in each iteration. Inside the
loop, the C library function rand is used to generate the extremes of the
rectangle. If necessary, the program then exchanges the coordinates to make
the minimum coordinates less than or equal to the maximum coordinates.
Finally, it again uses rand to generate a random color number between 0 and
15, and then draws the rectangle in that color. After drawing all 200
rectangles, the program restores the original video mode and screen
attributes before returning to DOS.
Chapter 6: Graphics Fundamentals 87
Example 6-10.
#include <fastgraf.h>
void main(void);
#define RECTANGLES 200
#define SWAP(a,b,temp) { temp = a; a = b; b = temp; }
void main()
{
int i;
int minx, maxx, miny, maxy;
int old_mode;
int temp;
int xres, yres;
old_mode = fg_getmode();
fg_setmode(fg_automode());
xres = fg_getmaxx() + 1;
yres = fg_getmaxy() + 1;
for (i = 0; i < RECTANGLES; i++) {
minx = rand() % xres;
maxx = rand() % xres;
miny = rand() % yres;
maxy = rand() % yres;
if (minx > maxx)
SWAP(minx,maxx,temp);
if (miny > maxy)
SWAP(miny,maxy,temp);
fg_setcolor(rand()%16);
fg_rect(minx,maxx,miny,maxy);
}
fg_setmode(old_mode);
fg_reset();
}
Unfilled Rectangles
Fastgraph includes a routine for drawing unfilled rectangles. The
fg_box routine draws an unfilled rectangle in screen space, with regard to
the clipping limits, using the current color. The arguments to fg_box are
the same as those for fg_rect. The depth of the rectangle's edges is one
pixel by default, but you can change the depth by calling fg_boxdepth. The
fg_boxdepth routine expects two arguments. The first argument is the width
of the rectangle's left and right sides, while the second is the height of
its top and bottom sides. Once you call fg_boxdepth, fg_box draws all
unfilled rectangles using the depth values specified in the most recent call
to fg_boxdepth. Unlike fg_rect, the fg_box routine has no effect in text
video modes.
Example 6-11 is the same as example 6-10, but it draws unfilled instead
of solid rectangles. The program uses fg_box to draw the each rectangle and
fg_boxdepth to define the rectangle depth at three pixels in each direction.
88 Fastgraph User's Guide
Note that you don't need to call fg_boxdepth for each rectangle if you want
all of them to have the same depth.
Example 6-11.
#include <fastgraf.h>
void main(void);
#define RECTANGLES 200
#define SWAP(a,b,temp) { temp = a; a = b; b = temp; }
void main()
{
int i;
int minx, maxx, miny, maxy;
int old_mode;
int temp;
int xres, yres;
old_mode = fg_getmode();
fg_setmode(fg_automode());
fg_boxdepth(3,3);
xres = fg_getmaxx() + 1;
yres = fg_getmaxy() + 1;
for (i = 0; i < RECTANGLES; i++) {
minx = rand() % xres;
maxx = rand() % xres;
miny = rand() % yres;
maxy = rand() % yres;
if (minx > maxx)
SWAP(minx,maxx,temp);
if (miny > maxy)
SWAP(miny,maxy,temp);
fg_setcolor(rand()%16);
fg_box(minx,maxx,miny,maxy);
}
fg_setmode(old_mode);
fg_reset();
}
Dithered Rectangles
The process of alternating different color pixels across a region of the
display area is called dithering. This technique is especially useful in the
graphics modes with few colors, such as CGA and Hercules modes, because you
can simulate additional colors through effective uses of dithering.
Fastgraph includes two routines for drawing dithered rectangles, one for
screen space and one for world space. Neither routine observes the clipping
limits, nor do they affect the graphics cursor position.
The fg_drect routine draws a dithered rectangle in screen space. Like
the fg_rect routine, fg_drect requires four integer arguments that
respectively define the minimum x, maximum x, minimum y, and maximum y screen
Chapter 6: Graphics Fundamentals 89
space coordinates of the rectangle. The minimum coordinates must be less
than or equal to the maximum coordinates, or else the results are
unpredictable. However, fg_drect also requires a fifth argument that defines
the dithering matrix, which in turn determines the pixel pattern used to
build the dithered rectangle. The size and format of the dithering matrix
are dependent on the video mode.
The world space version of the dithered rectangle drawing routine is
fg_drectw. Like fg_drect, it requires four arguments that define the
extremes of the rectangle, and a fifth argument that defines the dithering
matrix.
As mentioned earlier, the size and format of the dithering matrix are
dependent on the video mode. The dithering matrix is a four byte array in
all video modes except the 256 color graphics modes (modes 19 through 23),
where it is an eight byte array. This array contains a pixel pattern that
fg_drect or fg_drectw replicates across the rectangle's area. The structure
of the dithering matrix closely mimics the structure of video memory in each
graphics mode.
The remainder of this section will present some simple mode-specific
examples to illustrate the structure of the dithering matrix in the different
graphics modes. Suppose we would like to produce a "checkerboard" of light
blue and bright white pixels. That is, in a given row of a rectangle,
consecutive pixels will alternate between these two colors. Additionally,
the pattern for adjacent rows will be shifted such that there will always be
a bright white pixel above and below a light blue pixel, and vice versa.
Hence this pixel pattern would look something like
B W B W
W B W B
B W B W
W B W B
where each B represents a light blue pixel, and each W represents a bright
white pixel. The following examples describe the dithering matrix that could
be used to produce such a pixel pattern in each graphics mode.
CGA Four-Color Graphics Modes
The CGA four-color graphics modes (modes 4 and 5) use a four-byte
dithering matrix that Fastgraph treats as a four-row by one-column array.
Since each pixel in these modes requires two bits of video memory, each byte
of the dithering matrix holds four pixels. Thus, the pixel representation of
the dithering matrix would appear as shown below on the left; its translation
to numeric values appears on the right.
[3] B W B W [3] 01 11 01 11
[2] W B W B [2] 11 01 11 01
[1] B W B W [1] 01 11 01 11
[0] W B W B [0] 11 01 11 01
90 Fastgraph User's Guide
Because these modes do not offer a light blue color, we've used light cyan
(color value 1 in palette 1) to approximate light blue. The B pixels thus
translate to color value 1, or 01 binary. Bright white is available as color
value 3 in palette 1, so the W pixels translate to color value 3, or 11
binary. The hexadecimal equivalent of the binary value 11011101 (for array
elements [0] and [2]) is DD, and the hexadecimal equivalent of the binary
value 01110111 (for array elements [1] and [3]) is 77. As shown in example
6-12, these are precisely the values assigned to the elements of the
dithering matrix.
Example 6-12 uses mode 4 to display a 50-pixel by 50-pixel dithered
rectangle in the upper left corner of the screen. The dithering matrix
represents the blue and white checkerboard pattern discussed in the preceding
paragraph.
Example 6-12.
#include <fastgraf.h>
void main(void);
void main()
{
char matrix[4];
int old_mode;
old_mode = fg_getmode();
fg_setmode(4);
matrix[0] = matrix[2] = 0xDD;
matrix[1] = matrix[3] = 0x77;
fg_drect(0,49,0,49,matrix);
fg_waitkey();
fg_setmode(old_mode);
fg_reset();
}
CGA Two-Color Graphics Mode
The CGA two-color graphics mode (mode 6) uses a four-byte dithering
matrix that Fastgraph treats as a four-row by one-column array, as in the
other four-color CGA modes. However, each pixel in this mode only requires
one bit of video memory, so each byte of the dithering matrix holds eight
pixels. Thus, the pixel representation of the dithering matrix would appear
as shown below on the left; its translation to numeric values appears on the
right.
[3] B W B W B W B W [3] 0 1 0 1 0 1 0 1
[2] W B W B W B W B [2] 1 0 1 0 1 0 1 0
[1] B W B W B W B W [1] 0 1 0 1 0 1 0 1
[0] W B W B W B W B [0] 1 0 1 0 1 0 1 0
Chapter 6: Graphics Fundamentals 91
Mode 6 obviously does not offer a light blue color, so we've used black
(color value 0) in its place. The B pixels thus translate to color value 0.
Bright white is available as color value 1, so the W pixels translate to
color value 1. The hexadecimal equivalent of the binary value 10101010 (for
array elements [0] and [2]) is AA, and the hexadecimal equivalent of the
binary value 01010101 (for array elements [1] and [3]) is 55. Thus, to make
example 6-12 run in mode 6, we only need to change the fg_setmode argument
from 4 to 6 and change the dithering matrix values as shown below.
matrix[0] = matrix[2] = 0xAA;
matrix[1] = matrix[3] = 0x55;
Tandy/PCjr 16-Color Graphics Mode
The Tandy/PCjr 16-color graphics mode (mode 9) also uses a four-byte
dithering matrix that Fastgraph treats as a four-row by one-column array.
Each pixel in this mode requires four bits of video memory, so each byte of
the dithering matrix only holds two pixels. Thus, the pixel representation
of the dithering matrix would appear as shown below on the left; its
translation to numeric values appears on the right.
[3] B W [3] 1001 1111
[2] W B [2] 1111 1001
[1] B W [1] 1001 1111
[0] W B [0] 1111 1001
The B pixels translate to color value 9 (light blue), or 1001 binary, and the
W pixels translate to color value 15 (bright white), or 1111 binary. The
hexadecimal equivalent of the binary value 11111001 (for array elements [0]
and [2]) is F9, and the hexadecimal equivalent of the binary value 10011111
(for array elements [1] and [3]) is 9F. Thus, to make example 6-12 run in
mode 9, we only need to change the fg_setmode argument from 4 to 9 and change
the dithering matrix values as shown below.
matrix[0] = matrix[2] = 0xF9;
matrix[1] = matrix[3] = 0x9F;
Hercules Graphics Mode
The size and format of the dithering matrix in the Hercules graphics
mode (mode 11) are the same as in the CGA two-color mode (mode 6). Please
refer to page 90 for a discussion of mode 6 dithering.
Hercules Low-Resolution Graphics Mode
The size and format of the dithering matrix in the Hercules low-
resolution graphics mode (mode 12) are the same as in the CGA four-color
modes (modes 4 and 5). As far as our checkerboard example goes, we'll use
black (color value 0) in place of light blue, and bold (color value 3)
92 Fastgraph User's Guide
instead of bright white. Thus, the B pixels translate to 00 binary, while
the W pixels translate to 11 binary. The hexadecimal equivalent of the
binary value 11001100 (for array elements [0] and [2]) is CC, and the
hexadecimal equivalent of the binary value 00110011 (for array elements [1]
and [3]) is 33. Thus, to make example 6-12 run in mode 12, we only need to
change the fg_setmode argument from 4 to 12 and change the dithering matrix
values as shown below.
matrix[0] = matrix[2] = 0xCC;
matrix[1] = matrix[3] = 0x33;
EGA and VGA Graphics Modes
The native EGA and VGA graphics modes (modes 13 through 18) use a four-
byte dithering matrix that Fastgraph treats as a four-row by one-column
array. Unlike the other graphics modes, which allow you to store pixels of
several colors in the dithering matrix, the EGA and VGA modes treat the
dithering matrix as a bit map for a specific color. Since each color in the
dither pattern must be stored in a separate bit map (that is, in a separate
dithering matrix), you must call fg_drect once for each color. Furthermore,
you must use the fg_setcolor routine before each call to fg_drect to define
the color used with the dithering matrix.
In all EGA and VGA graphics modes, each byte of the dithering matrix is
a bit map that represents eight pixels. Using our familiar checkerboard
example, the pixel representation of the dithering matrix would appear as
shown below.
[3] B W B W B W B W
[2] W B W B W B W B
[1] B W B W B W B W
[0] W B W B W B W B
Translating this pattern to numeric values is simple. Just construct one
dithering matrix for each color in the pattern (there are two colors in this
example), where pixels of the current color translate to 1, and other pixels
translate to 0. Following our example, the translation for the B pixels
appears below on the left, while the translation for the W pixels appears on
the right.
[3] 1 0 1 0 1 0 1 0 [3] 0 1 0 1 0 1 0 1
[2] 0 1 0 1 0 1 0 1 [2] 1 0 1 0 1 0 1 0
[1] 1 0 1 0 1 0 1 0 [1] 0 1 0 1 0 1 0 1
[0] 0 1 0 1 0 1 0 1 [0] 1 0 1 0 1 0 1 0
Chapter 6: Graphics Fundamentals 93
The hexadecimal equivalent of the binary value 01010101 is 55, and the
hexadecimal equivalent of the binary value 10101010 is AA. As shown in
example 6-13, these are precisely the values assigned to the elements of the
dithering matrices.
Example 6-13 uses mode 13 to display our light blue and bright white
checkerboard pattern. Note you must call fg_drect twice -- once for the
light blue pixels (color value 9), and again for the bright white pixels
(color value 15). Note also how fg_setcolor is used before each call to
fg_drect to define the color of the pixels fg_drect will display.
Example 6-13.
#include <fastgraf.h>
void main(void);
void main()
{
char matrix[4];
int old_mode;
old_mode = fg_getmode();
fg_setmode(13);
matrix[0] = matrix[2] = 0x55;
matrix[1] = matrix[3] = 0xAA;
fg_setcolor(9);
fg_drect(0,49,0,49,matrix);
matrix[0] = matrix[2] = 0xAA;
matrix[1] = matrix[3] = 0x55;
fg_setcolor(15);
fg_drect(0,49,0,49,matrix);
fg_waitkey();
fg_setmode(old_mode);
fg_reset();
}
MCGA and VGA 256-Color Graphics Modes
The MCGA and VGA 256-color graphics modes (modes 19 through 23) use an
eight-byte dithering matrix that Fastgraph treats as a four-row by two-column
array. Each pixel in these modes requires eight bits of video memory, so
each byte of the dithering matrix only holds a single pixel. We therefore
need the two column dithering matrix to produce any type of dither pattern.
The pixel representation of the dithering matrix would appear as shown below
on the left; its translation to numeric values appears on the right.
[6] B W [7] [6] 9 15 [7]
[4] W B [5] [4] 15 9 [5]
[2] B W [3] [2] 9 15 [3]
[0] W B [1] [0] 15 9 [1]
94 Fastgraph User's Guide
The B pixels translate to color value 9 (light blue), and the W pixels
translate to color value 15 (bright white). Example 6-14 uses mode 19 to
draw our light blue and bright white checkerboard pattern.
Example 6-14.
#include <fastgraf.h>
void main(void);
void main()
{
char matrix[8];
int old_mode;
old_mode = fg_getmode();
fg_setmode(19);
matrix[0] = matrix[3] = matrix[4] = matrix[7] = 15;
matrix[1] = matrix[2] = matrix[5] = matrix[6] = 9;
fg_drect(0,49,0,49,matrix);
fg_waitkey();
fg_setmode(old_mode);
fg_reset();
}
Closing Remarks
There are two other important items pertaining to the fg_drect and
fg_drectw routines. These items apply regardless of which graphics video
mode is being used.
First, the dithering matrix may not contain virtual color values. That
is, the pixel color values stored in the dithering matrix must be between 0
and the maximum color value for the current video mode. If any color value
is redefined using the fg_defcolor routine, Fastgraph always ignores the
redefinition and instead uses the actual color value. Note this does not
apply to palette registers or video DAC registers, because in these cases we
are redefining the color associated with a color value and not the color
value itself.
Second, Fastgraph aligns the dithering matrix to specific pixel rows.
Fastgraph draws the dithered rectangle starting with the pixel row specified
by the rectangle's lower limit (the maximum y coordinate for fg_drect, or the
minimum y coordinate for fg_drectw) and proceeds upward until reaching the
rectangle's upper limit. In all cases the dithering matrix used by fg_drect
and fg_drectw contains four rows. If we let r represent the pixel row
corresponding to the rectangle's lower limit, then the first row used in the
dithering matrix is r modulo 4 (assuming the dithering matrix rows are
numbered 0 to 3). This alignment enables you to use the same dithering
matrix in multiple calls to fg_drect and fg_drectw for building an object of
adjacent dithered rectangles (for example, an L-shaped area) and still have
the dither pattern match where the rectangles intersect.
Chapter 6: Graphics Fundamentals 95
Region Fill
Fastgraph includes routines for filling arbitrary regions. The fg_paint
routine fills a region with the current color value by specifying a screen
space point in the region's interior. The fg_paintw routine also fills a
region, but it requires the interior point to be expressed in world space.
Neither routine changes the graphics cursor position.
Each of these routines has two arguments that specify the (x,y)
coordinates of the interior point. For the fg_paint routine, the arguments
are integer quantities. For the fg_paintw routine, they are floating point
quantities.
The region being filled must be a closed polygon whose boundary color is
different from that of the specified interior point. The region may contain
holes (interior areas that will not be filled). Fastgraph fills the region
by changing every interior pixel whose color is the same as the specified
interior point, to the current color. If the interior point is already the
current color, the region fill routines do nothing. It is important to note
the screen edges are not considered polygon boundaries, and filling an open
polygon will cause the fg_paint and fg_paintw routines to behave
unpredictably.
Example 6-15 illustrates a simple use of the fg_paint routine in a 320
by 200 graphics mode. The program uses fg_bestmode to select an available
video mode (if no 320 by 200 graphics mode is available, the program exits).
After establishing the selected video mode, the program uses the fg_move and
fg_drawrel routines to draw a hollow rectangle in color 10 and a hollow
diamond in color 9. The diamond is drawn in the middle of the rectangle,
thus making it a hole with respect to the rectangle. The program leaves
these shapes on the screen until a key is pressed. At that time, it calls
the fg_paint routine to fill that part of the rectangle outside the diamond
with color 10. After waiting for another keystroke, the program again uses
fg_paint to fill the interior of the diamond with color 15. Finally, the
program waits for another keystroke, restores the original video mode and
screen attributes, and returns to DOS.
Example 6-15.
#include <fastgraf.h>
#include <stdio.h>
#include <stdlib.h>
void main(void);
void main()
{
int old_mode, new_mode;
new_mode = fg_bestmode(320,200,1);
if (new_mode < 0) {
printf("This program requires a 320 x 200 ");
printf("graphics mode.\n");
exit(1);
}
old_mode = fg_getmode();
96 Fastgraph User's Guide
fg_setmode(new_mode);
fg_setcolor(10);
fg_move(100,50);
fg_drawrel(120,0);
fg_drawrel(0,100);
fg_drawrel(-120,0);
fg_drawrel(0,-100);
fg_setcolor(9);
fg_move(160,80);
fg_drawrel(30,20);
fg_drawrel(-30,20);
fg_drawrel(-30,-20);
fg_drawrel(30,-20);
fg_waitkey();
fg_setcolor(10);
fg_paint(160,70);
fg_waitkey();
fg_setcolor(15);
fg_paint(160,100);
fg_waitkey();
fg_setmode(old_mode);
fg_reset();
}
Summary of Fundamental Graphics Routines
This section summarizes the functional descriptions of the Fastgraph
routines presented in this chapter. More detailed information about these
routines, including their arguments and return values, may be found in the
Fastgraph Reference Manual.
FG_BOX draws an unfilled rectangle in screen space, with respect to the
clipping region. The width of the rectangle's edges is one pixel unless
changed with the fg_boxdepth routine.
FG_BOXDEPTH defines the depth of rectangles drawn with the fg_box
routine. The fg_setmode routine initializes the box depth to one pixel.
FG_CIRCLE draws an unfilled circle in screen space. The circle is
centered at the graphics cursor position.
FG_CIRCLEW draws an unfilled circle in world space. The circle is
centered at the graphics cursor position.
FG_CLPRECT draws a solid (filled) rectangle in screen space, with
respect to the clipping region.
FG_CLPRECTW draws a solid (filled) rectangle in world space, with
respect to the clipping region.
Chapter 6: Graphics Fundamentals 97
FG_DASH draws a dashed line from the graphics cursor position to an
absolute screen space position. It also makes the destination position the
new graphics cursor position.
FG_DASHREL draws a dashed line from the graphics cursor position to a
screen space position relative to it. It also makes the destination position
the new graphics cursor position.
FG_DASHRW draws a dashed line from the graphics cursor position to a
world space position relative to it. It also makes the destination position
the new graphics cursor position.
FG_DASHW draws a dashed line from the graphics cursor position to an
absolute world space position. It also makes the destination position the
new graphics cursor position.
FG_DRAW draws a solid line from the graphics cursor position to an
absolute screen space position. It also makes the destination position the
new graphics cursor position.
FG_DRAWREL draws a solid line from the graphics cursor position to a
screen space position relative to it. It also makes the destination position
the new graphics cursor position.
FG_DRAWRW draws a solid line from the graphics cursor position to a
world space position relative to it. It also makes the destination position
the new graphics cursor position.
FG_DRAWW draws a solid line from the graphics cursor position to an
absolute world space position. It also makes the destination position the
new graphics cursor position.
FG_DRECT draws a dithered rectangle in screen space, without regard to
the clipping region. The rectangle's dither pattern is defined through a
dithering matrix whose format is dependent on the video mode being used.
FG_DRECTW draws a dithered rectangle in world space, without regard to
the clipping region. The rectangle's dither pattern is defined through a
dithering matrix whose format is dependent on the video mode being used.
FG_ELLIPSE draws an unfilled ellipse in screen space. The ellipse is
centered at the graphics cursor position, and its size is determined by the
specified lengths of the semi-axes.
FG_ELLIPSEW draws an unfilled ellipse in world space. The ellipse is
centered at the graphics cursor position, and its size is determined by the
specified lengths of the semi-axes.
FG_ERASE clears the screen in either text or graphics modes.
FG_GETPIXEL returns the color value of a specified pixel.
FG_GETXPOS returns the screen space x coordinate of the graphics cursor
position.
FG_GETYPOS returns the screen space y coordinate of the graphics cursor
position.
98 Fastgraph User's Guide
FG_MOVE establishes the graphics cursor position at an absolute screen
space point.
FG_MOVEREL establishes the graphics cursor position at a screen space
point relative to the current position.
FG_MOVERW establishes the graphics cursor position at a world space
point relative to the current position.
FG_MOVEW establishes the graphics cursor position at an absolute world
space point.
FG_PAINT fills an arbitrary closed region with the current color value.
The region is defined by specifying a screen space point within its interior.
FG_PAINTW fills an arbitrary closed region with the current color value.
The region is defined by specifying a world space point within its interior.
FG_POINT draws a point (that is, displays a pixel) in screen space.
FG_POINTW draws a point in world space.
FG_POLYGON draws an unfilled polygon in screen space, using two
coordinate arrays to define the polygon vertices. The drawing of the polygon
begins at the graphics cursor position, through the vertices defined by the
coordinate arrays, and finally back to the original cursor position if
necessary.
FG_POLYGONW draws an unfilled polygon in world space. It is the same as
the fg_polygon routine, except the coordinate arrays contain world space
values.
FG_RECT draws a solid (filled) rectangle in screen space or character
space, without regard to the clipping region.
FG_RECTW draws a solid (filled) rectangle in world space, without regard
to the clipping region.
FG_SETCLIP defines the clipping region in screen space. The clipping
region is a rectangular area outside of which graphics are suppressed.
FG_SETCLIPW defines the clipping region in world space.
