Bump version.

[ntk.git] / documentation / src / drawing.dox

/**

 \page  drawing Drawing Things in FLTK

This chapter covers the drawing functions that are provided with FLTK.

\section sect_WhenCanYouDraw When Can You Draw Things in FLTK?

There are only certain places you can execute drawing code in FLTK.
Calling these functions at other places will result in undefined behavior!

\li The most common place is inside the virtual Fl_Widget::draw() method.
    To write code here, you must subclass one of the existing Fl_Widget
    classes and implement your own version of draw().

\li You can also create custom \ref common_boxtypes "boxtypes" and
    \ref common_labeltype "labeltypes". These involve writing small
    procedures that can be called by existing Fl_Widget::draw() methods.
    These "types" are identified by an 8-bit index that is stored in the
    widget's \p box(), \p labeltype(), and possibly other properties.

\li You can call Fl_Window::make_current() to do incremental update of a
    widget. Use Fl_Widget::window() to find the window.


\section sect_DrawingFunctions Drawing Functions

To use the drawing functions you must first include the <FL/fl_draw.H>
header file. FLTK provides the following types of drawing functions:

\li \ref ssect_Boxes
\li \ref ssect_Clipping
\li \ref drawing_colors
\li \ref ssect_Lines
\li \ref ssect_Fast
\li \ref ssect_Complex
\li \ref ssect_Text
\li \ref ssect_Fonts
\li \ref ssect_CharacterEncoding
\li \ref ssect_Overlay
\li \ref drawing_images
\li \ref ssect_DirectImageDrawing
\li \ref ssect_DirectImageReading
\li \ref ssect_Fl_Image
\li \ref ssect_Offscreen

\subsection ssect_Boxes Boxes

FLTK provides three functions that can be used to draw boxes for buttons
and other UI controls. Each function uses the supplied upper-lefthand corner
and width and height to determine where to draw the box.

void fl_draw_box(Fl_Boxtype b, int x, int y, int w, int h, Fl_Color c);

\par
The \p %fl_draw_box() function draws a standard boxtype \p b
in the specified color \p c.

\anchor drawing_fl_frame
void fl_frame(const char *s, int x, int y, int w, int h) <br>
void fl_frame2(const char *s, int x, int y, int w, int h)

\par
The \p %fl_frame() and \p %fl_frame2() functions draw a series of
line segments around the given box. The string \p s must contain groups
of 4 letters which specify one of 24 standard grayscale values,
where 'A' is black and 'X' is white.
The results of calling these functions with a string that is not a
multiple of 4 characters in length are undefined.

\par
The only difference between \p %fl_frame() and \p %fl_frame2()
is the order of the line segments:
  - For \p %fl_frame() the order of each set of 4 characters is:
    top, left, bottom, right.
  - For \p %fl_frame2() the order of each set of 4 characters is:
    bottom, right, top, left.

\par
Note that
\ref common_fl_frame "fl_frame(Fl_Boxtype b)" 
is described in the \ref common_boxtypes section.


\subsection ssect_Clipping Clipping

You can limit all your drawing to a rectangular region by calling
\p %fl_push_clip(), and put the drawings back by using
\p %fl_pop_clip().
This rectangle is measured in pixels and is unaffected by the current
transformation matrix.

In addition, the system may provide clipping when updating windows
which may be more complex than a simple rectangle.

void fl_push_clip(int x, int y, int w, int h) <br>
void fl_clip(int x, int y, int w, int h)

\par
Intersect the current clip region with a rectangle and push this new
region onto the stack.

\par
The \p %fl_clip() version is deprecated and
will be removed from future releases.

void fl_push_no_clip()

\par
Pushes an empty clip region on the stack so nothing will be clipped.

void fl_pop_clip()

\par
Restore the previous clip region.

\par
\b Note:
You must call \p %fl_pop_clip() once for every time you call
\p %fl_push_clip().
If you return to FLTK with the clip stack not empty unpredictable results
occur.

int fl_not_clipped(int x, int y, int w, int h)

\par
Returns non-zero if any of the rectangle intersects the current clip
region. If this returns 0 you don't have to draw the object.

\par
\b Note:
Under X this returns 2 if the rectangle is partially clipped,
and 1 if it is entirely inside the clip region.

int fl_clip_box(int x, int y, int w, int h, int &X, int &Y, int &W, int &H)

\par
Intersect the rectangle <tt>x,y,w,h</tt> with the current
clip region and returns the bounding box of the result in
<tt>X,Y,W,H</tt>. Returns non-zero if the resulting rectangle is
different than the original. This can be used to limit the
necessary drawing to a rectangle. \c W and \c H are
set to zero if the rectangle is completely outside the region.

void fl_clip_region(Fl_Region r) <br>
Fl_Region fl_clip_region()

\par
Replace the top of the clip stack with a clipping region of any shape.
Fl_Region is an operating system specific type. The second form returns 
the current clipping region.


\section drawing_colors Colors

FLTK manages colors as 32-bit unsigned integers, encoded as RGBI.
When the RGB bytes are non-zero, the value is treated as RGB.
If these bytes are zero, the I byte will be used as an index
into the colormap.

Values from 0 to 255, i.e. the I index value, represent
colors from the FLTK 1.3.x standard colormap
and are allocated as needed on screens without TrueColor support.
The \b Fl_Color enumeration type defines the
standard colors and color cube for the first 256 colors. All of
these are named with symbols in
\ref enumerations "<FL/Enumerations.H>".

Color values greater than 255 are treated as 24-bit RGB
values. These are mapped to the closest color supported by the
screen, either from one of the 256 colors in the FLTK 1.3.x
colormap or a direct RGB value on TrueColor screens.

Fl_Color fl_rgb_color(uchar r, uchar g, uchar b) <br>
Fl_Color fl_rgb_color(uchar grayscale)

\par
Generate Fl_Color out of specified
8-bit RGB values or one 8-bit grayscale value.

void fl_color(Fl_Color c) <br>
void fl_color(int c)

\par
Sets the color for all subsequent drawing operations.
Please use the first form:
the second form is only provided for back compatibility.

\par
For colormapped displays, a color cell will be allocated out
of \p fl_colormap the first time you use a color. If the
colormap fills up then a least-squares algorithm is used to find
the closest color.

Fl_Color fl_color()

\par
Returns the last color that was set using \p %fl_color().
This can be used for state save/restore.

void fl_color(uchar r, uchar g, uchar b)

\par
Set the color for all subsequent drawing operations. The
closest possible match to the RGB color is used. The RGB color
is used directly on TrueColor displays. For colormap visuals the
nearest index in the gray ramp or color cube is used.

unsigned Fl::get_color(Fl_Color i) <br>
void Fl::get_color(Fl_Color i, uchar &red, uchar &green, uchar &blue)

\par
Generate RGB values from a colormap index value \p i.
The first returns the RGB as a 32-bit unsigned integer,
and the second decomposes the RGB into three 8-bit values.
\todo work out why Fl::get_color() does not give links!

Fl::get_system_colors() <br>
Fl::foreground() <br>
Fl::background() <br>
Fl::background2()

\par
The first gets color values from the user preferences or the system,
and the other routines are used to apply those values.

Fl::own_colormap() <br>
Fl::free_color(Fl_Color i, int overlay) <br>
Fl::set_color(Fl_Color i, unsigned c)

\par
\p Fl::own_colormap() is used to install a local colormap [X11 only].
\par
\p Fl::free_color() and \p Fl::set_color() are used to remove and replace
entries from the colormap.
\todo work out why these do not give links!

There are two predefined graphical interfaces for choosing colors.
The function fl_show_colormap() shows a table of colors and returns an
Fl_Color index value.
The Fl_Color_Chooser widget provides a standard RGB color chooser.

As the Fl_Color encoding maps to a 32-bit unsigned integer representing
RGBI, it is also possible to specify a color using a hex constant as a
color map index:
<pre>
// COLOR MAP INDEX
color(0x000000II)
        ------ |
           |   |
           |   Color map index (8 bits)
           Must be zero
</pre>
\code
button->color(0x000000ff);                  // colormap index #255 (FL_WHITE)
\endcode

or specify a color using a hex constant for the RGB components:
<pre>
// RGB COLOR ASSIGNMENTS
color(0xRRGGBB00)
         | | | |
         | | | Must be zero
         | | Blue (8 bits)
         | Green (8 bits)
         Red (8 bits)
</pre>
\code
button->color(0xff000000);                  // RGB: red
button->color(0x00ff0000);                  // RGB: green
button->color(0x0000ff00);                  // RGB: blue
button->color(0xffffff00);                  // RGB: white
\endcode

\note
If TrueColor is not available, any RGB colors will be set to
the nearest entry in the colormap.

\subsection ssect_Lines Line Dashes and Thickness

FLTK supports drawing of lines with different styles and
widths. Full functionality is not available under Windows 95, 98,
and Me due to the reduced drawing functionality these operating
systems provide.

void fl_line_style(int style, int width, char* dashes)

\par
Set how to draw lines (the "pen").  If you change this it is your
responsibility to set it back to the default with
\p fl_line_style(0).

\par
\b Note:
Because of how line styles are implemented on MS Windows systems, you
\e must set the line style \e after setting the drawing color.
If you set the
color after the line style you will lose the line style settings!

\par
\p style is a bitmask which is a bitwise-OR of the following
values. If you don't specify a dash type you will get a solid
line. If you don't specify a cap or join type you will get a
system-defined default of whatever value is fastest.

\par
\li <tt>FL_SOLID&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; -------</tt>
\li <tt>FL_DASH&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; - - - -</tt>
\li <tt>FL_DOT&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; .......</tt>
\li <tt>FL_DASHDOT&nbsp;&nbsp;&nbsp; - . - .</tt>
\li <tt>FL_DASHDOTDOT - .. -</tt>
\li <tt>FL_CAP_FLAT</tt>
\li <tt>FL_CAP_ROUND</tt>
\li <tt>FL_CAP_SQUARE</tt> (extends past end point 1/2 line width)
\li <tt>FL_JOIN_MITER</tt> (pointed)
\li <tt>FL_JOIN_ROUND</tt>
\li <tt>FL_JOIN_BEVEL</tt> (flat)

\par
\p width is the number of pixels thick to draw the lines.
Zero results in the system-defined default, which on both X and
Windows is somewhat different and nicer than 1.

\par
\p dashes is a pointer to an array of dash lengths, measured in
pixels.  The first location is how long to draw a solid portion, the
next is how long to draw the gap, then the solid, etc.  It is
terminated with a zero-length entry. A \p NULL pointer or a zero-length
array results in a solid line. Odd array sizes are not supported and
result in undefined behavior.

\par
\b Note:
The dashes array does not work under Windows 95, 98, or Me, since those
operating systems do not support complex line styles.


\subsection ssect_Fast Drawing Fast Shapes

These functions are used to draw almost all the FLTK widgets.
They draw on exact pixel boundaries and are as fast as possible.
Their behavior is duplicated exactly on all platforms FLTK is
ported. It is undefined whether these are affected by the
\ref ssect_Complex "transformation matrix",
so you should only call these while the matrix is set to the
identity matrix (the default).

void fl_point(int x, int y)

\par
Draw a single pixel at the given coordinates.

void fl_rectf(int x, int y, int w, int h) <br>
void fl_rectf(int x, int y, int w, int h)

\par
Color a rectangle that exactly fills the given bounding box.

void fl_rectf(int x, int y, int w, int h, uchar r, uchar g, uchar b)

\par
Color a rectangle with "exactly" the passed
<tt>r,g,b</tt> color. On screens with less than 24 bits of
color this is done by drawing a solid-colored block using
\ref drawing_fl_draw_image "fl_draw_image()"
so that the correct color shade is produced.

void fl_rect(int x, int y, int w, int h) <br>
void fl_rect(int x, int y, int w, int h, Fl_Color c)

\par
Draw a 1-pixel border \e inside this bounding box.

void fl_line(int x, int y, int x1, int y1) <br>
void fl_line(int x, int y, int x1, int y1, int x2, int y2)

\par
Draw one or two lines between the given points.

void fl_loop(int x, int y, int x1, int y1, int x2, int y2) <br>
void fl_loop(int x, int y, int x1, int y1, int x2, int y2, int x3, int y3)

\par
Outline a 3 or 4-sided polygon with lines.

void fl_polygon(int x, int y, int x1, int y1, int x2, int y2) <br>
void fl_polygon(int x, int y, int x1, int y1, int x2, int y2, int x3, int y3)

\par
Fill a 3 or 4-sided polygon. The polygon must be convex.

void fl_xyline(int x, int y, int x1) <br>
void fl_xyline(int x, int y, int x1, int y2) <br>
void fl_xyline(int x, int y, int x1, int y2, int x3)

\par
Draw horizontal and vertical lines. A horizontal line is
drawn first, then a vertical, then a horizontal.

void fl_yxline(int x, int y, int y1) <br>
void fl_yxline(int x, int y, int y1, int x2) <br>
void fl_yxline(int x, int y, int y1, int x2, int y3)

\par
Draw vertical and horizontal lines. A vertical line is drawn
first, then a horizontal, then a vertical.

void fl_arc(int x, int y, int w, int h, double a1, double a2) <br>
void fl_pie(int x, int y, int w, int h, double a1, double a2)

\par
Draw ellipse sections using integer coordinates. These
functions match the rather limited circle drawing code provided
by X and MS Windows. The advantage over using
\ref drawing_fl_arc "fl_arc()"
with floating point
coordinates is that they are faster because they often use the
hardware, and they draw much nicer small circles, since the
small sizes are often hard-coded bitmaps.

\par
If a complete circle is drawn it will fit inside the passed bounding
box. The two angles are measured in degrees counter-clockwise from
3'oclock and are the starting and ending angle of the arc, \p a2
must be greater or equal to \p a1.

\par
\p %fl_arc() draws a series of lines to approximate the arc.
Notice that the integer version of \p %fl_arc() has a different
number of arguments to the other
\ref drawing_fl_arc "fl_arc()"
function described later in this chapter.

\par
\p %fl_pie() draws a filled-in pie slice. This slice may
extend outside the line drawn by \p %fl_arc(); to avoid this
use \p w-1 and \p h-1.

\todo
add an Fl_Draw_Area_Cb typedef to allow fl_scroll(...) to be doxygenated?

void fl_scroll(int X, int Y, int W, int H, int dx, int dy, void (*draw_area)(void*, int,int,int,int), void* data)

\par
Scroll a rectangle and draw the newly exposed portions. The contents
of the rectangular area is first shifted by \p dx and 
\p dy pixels. The callback is then called for every newly 
exposed rectangular area,


\subsection ssect_Complex Drawing Complex Shapes

The complex drawing functions let you draw arbitrary shapes
with 2-D linear transformations. The functionality matches that
found in the Adobe&reg; PostScript&tm; language. The
exact pixels that are filled are less defined than for the fast
drawing functions so that FLTK can take advantage of drawing
hardware. On both X and MS Windows the transformed vertices are
rounded to integers before drawing the line segments: this
severely limits the accuracy of these functions for complex
graphics, so use OpenGL when greater accuracy and/or performance
is required.

void fl_push_matrix() <br>
void fl_pop_matrix()

\par
Save and restore the current transformation.  The maximum
depth of the stack is 32 entries.

void fl_scale(double x,double y) <br>
void fl_scale(double x) <br>
void fl_translate(double x,double y) <br>
void fl_rotate(double d) <br>
void fl_mult_matrix(double a,double b,double c,double d,double x,double y)

\par
Concatenate another transformation onto the current one. The rotation
angle is in degrees (not radians) and is counter-clockwise.

double fl_transform_x(double x, double y) <br>
double fl_transform_y(double x, double y) <br>
double fl_transform_dx(double x, double y) <br>
double fl_transform_dy(double x, double y) <br>
void fl_transformed_vertex(double xf, double yf)

\par
Transform a coordinate or a distance using the current transformation matrix.
After transforming a coordinate pair, it can be added to the vertex
list without any further translations using \p %fl_transformed_vertex().

void fl_begin_points() <br>
void fl_end_points()

\par
Start and end drawing a list of points. Points are added to
the list with \p %fl_vertex().

void fl_begin_line() <br>
void fl_end_line()

\par
Start and end drawing lines.

void fl_begin_loop() <br>
void fl_end_loop()

\par
Start and end drawing a closed sequence of lines.

void fl_begin_polygon() <br>
void fl_end_polygon()

\par
Start and end drawing a convex filled polygon.

void fl_begin_complex_polygon() <br>
void fl_gap() <br>
void fl_end_complex_polygon()

\par
Start and end drawing a complex filled polygon. This polygon
may be concave, may have holes in it, or may be several
disconnected pieces. Call \p %fl_gap() to separate loops of
the path. It is unnecessary but harmless to call
\p %fl_gap() before the first vertex, after the last one,
or several times in a row.

\par
\p %fl_gap() should only be called between 
\p %fl_begin_complex_polygon() and
\p %fl_end_complex_polygon().
To outline the polygon, use
\p %fl_begin_loop() and replace each
\p %fl_gap() with a
\p %fl_end_loop();%fl_begin_loop() pair.

\par
\b Note:
For portability, you should only draw polygons that appear the same whether
"even/odd" or "non-zero" winding rules are used to fill them. Holes should
be drawn in the opposite direction of the outside loop.

void fl_vertex(double x,double y)

\par
Add a single vertex to the current path.

void fl_curve(double X0, double Y0, double X1, double Y1, double X2, double Y2, double X3, double Y3)

\par
Add a series of points on a Bezier curve to the path.  The curve ends
(and two of the points) are at <tt>X0,Y0</tt> and <tt>X3,Y3</tt>.

\anchor drawing_fl_arc
void fl_arc(double x, double y, double r, double start, double end)

\par
Add a series of points to the current path on the arc of a
circle; you can get elliptical paths by using scale and rotate
before calling \p %fl_arc().
The center of the circle is given by \p x and \p y,
and \p r is its radius.
\p %fl_arc()
takes \p start and \p end angles that are measured
in degrees counter-clockwise from 3 o'clock.
If \p end is less than \p start then it draws the arc in a clockwise
direction.

void fl_circle(double x, double y, double r)

\par
\p fl_circle(...) is equivalent to \p fl_arc(...,0,360) but may
be faster. It must be the \e only thing in the path: if you want
a circle as part of a complex polygon you must use \p %fl_arc().

\par
\b Note:
\p %fl_circle() draws incorrectly if the transformation is both rotated and
non-square scaled.

\subsection ssect_Text Drawing Text

All text is drawn in the
\ref drawing_fl_font "current font".
It is undefined whether this location or the characters are
modified by the current transformation.

void fl_draw(const char *, int x, int y) <br>
void fl_draw(const char *, int n, int x, int y)

\par
Draw a nul-terminated string or an array of \p n characters
starting at the given location. Text is aligned to the left and to
the baseline of the font. To align to the bottom, subtract
\p %fl_descent() from \p y.
To align to the top, subtract \p %fl_descent() and add \p %fl_height().
This version of \p %fl_draw()  provides direct access to
the text drawing function of the underlying OS. It does not apply any 
special handling to control characters. 

void fl_draw(const char* str, int x, int y, int w, int h, Fl_Align align, Fl_Image* img, int draw_symbols)

\par
Fancy string drawing function which is used to draw all the
labels. The string is formatted and aligned inside the passed
box.  Handles '\\t' and '\\n', expands all other control
characters to ^X, and aligns inside or against the edges of the
box described by \p x, \p y, \p w and \p h.
See Fl_Widget::align() for values for \p align.
The value \p FL_ALIGN_INSIDE is ignored, as this function always
prints inside the box.

\par
If \p img is provided and is not \p NULL, the
image is drawn above or below the text as specified by the
\p align value.

\par
The \p draw_symbols argument specifies whether or not
to look for symbol names starting with the "@" character.

\par
The text length is limited to 1024 characters per line.

void fl_measure(const char *str, int& w, int& h, int draw_symbols)

\par
Measure how wide and tall the string will be when printed by
the \p fl_draw(...align) function.
If the incoming \p w is non-zero it will wrap to that width.

int fl_height()

\par
Recommended minimum line spacing for the current font.  You
can also just use the value of \p size passed to
\ref drawing_fl_font "fl_font()".

int fl_descent()

\par
Recommended distance above the bottom of a \p %fl_height() tall box to draw
the text at so it looks centered vertically in that box.

double fl_width(const char* txt) <br>
double fl_width(const char* txt, int n) <br>
double fl_width(unsigned int unicode_char)

\par
Return the pixel width of a nul-terminated string, a sequence of \p n
characters, or a single character in the current font.

const char* fl_shortcut_label(int shortcut)

\par
Unparse a shortcut value as used by Fl_Button or Fl_Menu_Item
into a human-readable string like "Alt+N".  This only
works if the shortcut is a character key or a numbered function
key. If the shortcut is zero an empty string is returned. The
return value points at a static buffer that is overwritten with
each call.

\subsection ssect_Fonts Fonts

FLTK supports a set of standard fonts based on the Times,
Helvetica/Arial, Courier, and Symbol typefaces, as well as
custom fonts that your application may load. Each font is
accessed by an index into a font table.

Initially only the first 16 faces are filled in. There are
symbolic names for them: FL_HELVETICA,
FL_TIMES, FL_COURIER, and modifier values
FL_BOLD and FL_ITALIC which can be added to
these, and FL_SYMBOL and FL_ZAPF_DINGBATS.
Faces greater than 255 cannot be used in Fl_Widget
labels, since Fl_Widget stores the index as a byte.

\anchor drawing_fl_font
void fl_font(int face, int size)

\par
Set the current font, which is then used by the routines
described above. You may call this outside a draw context if
necessary to call fl_width(), but on X this will open
the display.

\par
The font is identified by a \p face and a \p size.
The size of the font is measured in \p pixels and not "points".
Lines should be spaced \p size pixels apart or more.

int fl_font() <br>
int fl_size()

\par
Returns the face and size set by the most recent call to
\p fl_font(a,b). This can be used to save/restore the font.

\subsection ssect_CharacterEncoding Character Encoding

FLTK 1.3 expects all text in Unicode UTF-8 encoding. UTF-8 is 
ASCII compatible for the first 128 characters. International 
characters are encoded in multibyte sequences. 

FLTK expects individual characters, characters that are not part of 
a string, in UCS-4 encoding, which is also ASCII compatible, but 
requires 4 bytes to store a Unicode character.

For more information about character encodings, see the chapter on
\ref unicode.

\subsection ssect_Overlay Drawing Overlays

These functions allow you to draw interactive selection rectangles
without using the overlay hardware. FLTK will XOR a single rectangle
outline over a window.

void fl_overlay_rect(int x, int y, int w, int h); <br>
void fl_overlay_clear();

\par
\p %fl_overlay_rect() draws a selection rectangle, erasing any
previous rectangle by XOR'ing it first. \p %fl_overlay_clear()
will erase the rectangle without drawing a new one.

\par
Using these functions is tricky. You should make a widget
with both a \p handle() and \p draw() method.
\p draw() should call \p %fl_overlay_clear() before
doing anything else.  Your \p handle() method should call
<tt>window()->make_current()</tt> and then
\p %fl_overlay_rect() after FL_DRAG events, and
should call \p %fl_overlay_clear() after a
FL_RELEASE event.


\section drawing_images Drawing Images

To draw images, you can either do it directly from data in
your memory, or you can create a Fl_Image object. The advantage of
drawing directly is that it is more intuitive, and it is faster
if the image data changes more often than it is redrawn. The
advantage of using the object is that FLTK will cache translated
forms of the image (on X it uses a server pixmap) and thus
redrawing is \e much faster.

\subsection ssect_DirectImageDrawing Direct Image Drawing

The behavior when drawing images when the current
transformation matrix is not the identity is not defined, so you
should only draw images when the matrix is set to the identity.

\anchor drawing_fl_draw_image
void fl_draw_image(const uchar *buf,int X,int Y,int W,int H,int D,int L)<br>
void fl_draw_image_mono(const uchar *buf,int X,int Y,int W,int H,int D,int L)

\par
Draw an 8-bit per color RGB or luminance image.  The pointer
points at the "r" data of the top-left pixel. Color
data must be in <tt>r,g,b</tt> order.
The top left corner is given by \p X and \p Y
and the size of the image is given by \p W and \p H.
\p D is the delta to add to the pointer between pixels,
it may be any value greater or equal to \p 3,
or it can be negative to flip the image horizontally.
\p L is the delta to add to the pointer between lines
(if 0 is passed it uses \p W*D).
and may be larger than \p W*D to crop data,
or negative to flip the image vertically.

\par
It is highly recommended that you put the following code before the
first show() of \e any window in your program to get rid
of the dithering if possible: 

\code
Fl::visual(FL_RGB);
\endcode

\par
Gray scale (1-channel) images may be drawn. This is done if
<tt>abs(D)</tt> is less than 3, or by calling
\p %fl_draw_image_mono(). Only one 8-bit sample is used for
each pixel, and on screens with different numbers of bits for
red, green, and blue only gray colors are used. Setting
\p D greater than 1 will let you display one channel of a
color image.

\par
\b Note:
The X version does not support all possible visuals.
If FLTK cannot draw the image in the current visual it
will abort. FLTK supports any visual of 8 bits or less,
and all common TrueColor visuals up to 32 bits.

typedef void (*Fl_Draw_Image_Cb)(void *data,int x,int y,int w,uchar *buf) <br>
void fl_draw_image(Fl_Draw_Image_Cb cb,void *data,int X,int Y,int W,int H,int D) <br>
void fl_draw_image_mono(Fl_Draw_Image_Cb cb,void *data,int X,int Y,int W,int H,int D)

\par
Call the passed function to provide each scan line of the
image.  This lets you generate the image as it is being drawn,
or do arbitrary decompression of stored data, provided it can be
decompressed to individual scan lines easily.

\par
The callback is called with the \p void* user data
pointer which can be used to point at a structure of information
about the image, and the \p x, \p y, and \p w
of the scan line desired from the image. 0,0 is the upper-left
corner of the image, <I>not <tt>X,Y</tt></I>. A pointer to a
buffer to put the data into is passed. You must copy \p w
pixels from scanline \p y, starting at pixel \p x,
to this buffer.

\par
Due to cropping, less than the whole image may be requested.
So \p x may be greater than zero, the first \p y may
be greater than zero, and \p w may be less than \p W.
The buffer is long enough to store the entire \p W*D
pixels, this is for convenience with some decompression
schemes where you must decompress the entire line at once:
decompress it into the buffer, and then if \p x is not
zero, copy the data over so the \p x'th pixel is at the
start of the buffer.

\par
You can assume the \p y's will be consecutive, except
the first one may be greater than zero.

\par
If \p D is 4 or more, you must fill in the unused bytes
with zero.

int fl_draw_pixmap(char* const* data, int x, int y, Fl_Color bg) <br>
int fl_draw_pixmap(const char* const* cdata, int x, int y, Fl_Color bg)

\par
Draws XPM image data, with the top-left corner at the given position.
The image is dithered on 8-bit displays so you won't lose color space
for programs displaying both images and pixmaps. This function returns
zero if there was any error decoding the XPM data.

\par
To use an XPM, do:

\code
#include "foo.xpm"
...
fl_draw_pixmap(foo, X, Y);
\endcode

\par
Transparent colors are replaced by the optional
Fl_Color argument. To draw with true transparency you must
use the Fl_Pixmap class.

int fl_measure_pixmap(char* const* data, int &w, int &h) <br>
int fl_measure_pixmap(const char* const* cdata, int &w, int &h)

\par
An XPM image contains the dimensions in its data. This
function finds and returns the width and height. The return
value is non-zero if the dimensions were parsed ok and zero if
there was any problem.

\subsection ssect_DirectImageReading Direct Image Reading

FLTK provides a single function for reading from the current
window or off-screen buffer into a RGB(A) image buffer.

uchar* fl_read_image(uchar *p, int X, int Y, int W, int H, int alpha)

\par
Read a RGB(A) image from the current window or off-screen
buffer. The \p p argument points to a buffer that can hold
the image and must be at least \p W*H*3 bytes when reading
RGB images and \p W*H*4 bytes when reading RGBA images. If
\p NULL, \p %fl_read_image() will create an array of
the proper size which can be freed using \p delete[].

\par
The \p alpha parameter controls whether an alpha
channel is created and the value that is placed in the alpha
channel. If 0, no alpha channel is generated.

\subsection ssect_Fl_Image Image Classes

FLTK provides a base image class called Fl_Image which supports
creating, copying, and drawing images of various kinds, along
with some basic color operations. Images can be used as labels
for widgets using the \p image() and \p deimage() methods or drawn directly.

The Fl_Image class does almost nothing by itself, but is instead 
supported by three basic image types:

\li Fl_Bitmap
\li Fl_Pixmap
\li Fl_RGB_Image

The Fl_Bitmap class encapsulates a mono-color bitmap image.
The \p draw() method draws the image using the current drawing
color.

The Fl_Pixmap class encapsulates a colormapped image.
The \p draw() method draws the image using the colors in the
file, and masks off any transparent colors automatically.

The Fl_RGB_Image class encapsulates a full-color 
(or grayscale) image with 1 to 4 color components. Images with 
an even number of components are assumed to contain an 
alpha channel that is used for transparency. The transparency 
provided by the draw() method is either a 24-bit 
blend against the existing window contents or a "screen door" 
transparency mask, depending on the platform and screen color depth. 

char fl_can_do_alpha_blending()

\par
\p %fl_can_do_alpha_blending() will return 1, if your
platform supports true alpha blending for RGBA images, or 0, 
if FLTK will use screen door transparency.

FLTK also provides several image classes based on the three
standard image types for common file formats:

\li Fl_GIF_Image 
\li Fl_JPEG_Image 
\li Fl_PNG_Image 
\li Fl_PNM_Image 
\li Fl_XBM_Image 
\li Fl_XPM_Image

Each of these image classes load a named file of the
corresponding format. The Fl_Shared_Image class
can be used to load any type of image file - the class examines
the file and constructs an image of the appropriate type.

Finally, FLTK provides a special image class called Fl_Tiled_Image to
tile another image object in the specified area. This class can be
used to tile a background image in a Fl_Group widget, for example.

virtual void Fl_Tiled_Image::copy(); <br>
virtual Fl_Image* Fl_Tiled_Image::copy(int w, int h);

\par
The \p copy() method creates a copy of the image. The second form
specifies the new size of the image - the image is resized using the
nearest-neighbor algorithm.

void Fl_Tiled_Image::draw(int x, int y, int w, int h, int ox, int oy);

\par
The \p draw() method draws the image object.
<tt>x,y,w,h</tt> indicates a destination rectangle.
<tt>ox,oy,w,h</tt> is a source rectangle. This source rectangle
is copied to the destination. The source rectangle may extend
outside the image, i.e. \p ox and \p oy may be
negative and \p w and \p h may be bigger than the
image, and this area is left unchanged.

void Fl_Tiled_Image::draw(int x, int y)

\par
Draws the image with the upper-left corner at <tt>x,y</tt>.
This is the same as doing \p draw(x,y,img->w(),img->h(),0,0).

\subsection ssect_Offscreen Offscreen Drawing

Sometimes it can be very useful to generate a complex drawing
in memory first and copy it to the screen at a later point in 
time. This technique can significantly reduce the amount of
repeated drawing. Offscreen drawing functions are declared in <FL/x.H>.
Fl_Double_Window uses offscreen rendering
to avoid flickering on systems that don't support 
double-buffering natively. 

Fl_Offscreen fl_create_offscreen(int w, int h)

\par
Create an RGB offscreen buffer with \p w*h pixels.

void fl_delete_offscreen(Fl_Offscreen)

\par
Delete a previously created offscreen buffer. All drawings are lost.

void fl_begin_offscreen(Fl_Offscreen)

\par
Send all subsequent drawing commands to this offscreen buffer. 
FLTK can draw into a buffer at any time. There is no need to wait for 
an Fl_Widget::draw() to occur.

void fl_end_offscreen()

\par
Quit sending drawing commands to this offscreen buffer.

void fl_copy_offscreen(int x, int y, int w, int h, Fl_Offscreen osrc, int srcx, int srcy)

\par
Copy a rectangular area of the size \p w*h from \p srcx,srcy
in the offscreen buffer into the current buffer at \p x,y.


\htmlonly
<hr>
<table summary="navigation bar" width="100%" border="0">
<tr>
  <td width="45%" align="LEFT">
    <a class="el" href="editor.html">
    [Prev]
    Designing a Simple Text Editor
    </a>
  </td>
  <td width="10%" align="CENTER">
    <a class="el" href="index.html">[Index]</a>
  </td>
  <td width="45%" align="RIGHT">
    <a class="el" href="events.html">
    Handling Events
    [Next]
    </a>
  </td>
</tr>
</table>
\endhtmlonly

*/