by Eric S. Raymond and Zeyd M. Ben-Halim
updates since release 1.9.9e by Thomas Dickey
curses. It is not an exhaustive reference for
the curses Application Programming Interface (API); that role is filled by the
curses manual pages. Rather, it is intended to help C programmers
ease into using the package.
This document is aimed at C applications programmers not yet specifically
familiar with ncurses. If you are already an experienced curses
programmer, you should nevertheless read the sections on Mouse
Interfacing, Debugging,
Compatibility
with Older Versions, and Hints, Tips,
and Tricks. These will bring you up to speed on the special features and
quirks of the ncurses implementation. If you are not so
experienced, keep reading.
The curses package is a subroutine library for
terminal-independent screen-painting and input-event handling which presents a
high level screen model to the programmer, hiding differences between terminal
types and doing automatic optimization of output to change one screen full of
text into another. Curses uses terminfo, which is a database format
that can describe the capabilities of thousands of different terminals.
The curses API may seem something of an archaism on UNIX
desktops increasingly dominated by X, Motif, and Tcl/Tk. Nevertheless, UNIX
still supports tty lines and X supports xterm(1); the
curses API has the advantage of (a) back-portability to
character-cell terminals, and (b) simplicity. For an application that does not
require bit-mapped graphics and multiple fonts, an interface implementation
using curses will typically be a great deal simpler and less
expensive than one using an X toolkit.
curses was the routines written to provide
screen-handling for the game rogue; these used the already-existing
termcap database facility for describing terminal capabilities.
These routines were abstracted into a documented library and first released with
the early BSD UNIX versions.
System III UNIX from Bell Labs featured a rewritten and much-improved
curses library. It introduced the terminfo format. Terminfo is
based on Berkeley's termcap database, but contains a number of improvements and
extensions. Parameterized capabilities strings were introduced, making it
possible to describe multiple video attributes, and colors and to handle far
more unusual terminals than possible with termcap. In the later AT&T System
V releases, curses evolved to use more facilities and offer more
capabilities, going far beyond BSD curses in power and flexibility.
ncurses, a free implementation of the System V curses
API with some clearly marked extensions. It includes the following System V
curses features:
The ncurses package can also capture and use event reports from
a mouse in some environments (notably, xterm under the X window system). This
document includes tips for using the mouse.
The ncurses package was originated by Pavel Curtis. The original
maintainer of this package is Zeyd
Ben-Halim <zmbenhal@netcom.com>. Eric S. Raymond
<esr@snark.thyrsus.com> wrote many of the new features in versions after
1.8.1 and wrote most of this introduction. Jürgen Pfeifer wrote all of the menu
and forms code as well as the Ada95
binding. Ongoing work is being done by Thomas Dickey (maintainer).
Contact the current maintainers at bug-ncurses@gnu.org.
This document also describes the panels extension library, similarly modeled on the SVr4 panels facility. This library allows you to associate backing store with each of a stack or deck of overlapping windows, and provides operations for moving windows around in the stack that change their visibility in the natural way (handling window overlaps).
Finally, this document describes in detail the menus and forms extension libraries, also cloned from System V, which support easy construction and sequences of menus and fill-in forms.
stdscr, is automatically provided for the
programmer.
#include <curses.h>at the top of the program source. The screen package uses the Standard I/O library, so
<curses.h> includes <stdio.h>.
<curses.h> also includes <termios.h>,
<termio.h>, or <sgtty.h> depending on your
system. It is redundant (but harmless) for the programmer to do these includes,
too. In linking with curses you need to have -lncurses
in your LDFLAGS or on the command line. There is no need for any other
libraries.
curscr, for current screen) is a screen image of what the terminal
currently looks like. Another screen (called stdscr, for standard
screen) is provided by default to make changes on.
A window is a purely internal representation. It is used to build and store a potential image of a portion of the terminal. It doesn't bear any necessary relation to what is really on the terminal screen; it's more like a scratchpad or write buffer.
To make the section of physical screen corresponding to a window reflect the
contents of the window structure, the routine refresh() (or
wrefresh() if the window is not stdscr) is called.
A given physical screen section may be within the scope of any number of overlapping windows. Also, changes can be made to windows in any order, without regard to motion efficiency. Then, at will, the programmer can effectively say ``make it look like this,'' and let the package implementation determine the most efficient way to repaint the screen.
curscr, which knows what the terminal looks like, and
stdscr, which is what the programmer wants the terminal to look
like next. The user should never actually access curscr directly.
Changes should be made to through the API, and then the routine
refresh() (or wrefresh()) called.
Many functions are defined to use stdscr as a default screen.
For example, to add a character to stdscr, one calls
addch() with the desired character as argument. To write to a
different window. use the routine waddch() (for `w'indow-specific
addch()) is provided. This convention of prepending function names with a `w'
when they are to be applied to specific windows is consistent. The only routines
which do not follow it are those for which a window must always be specified.
In order to move the current (y, x) coordinates from one point to another,
the routines move() and wmove() are provided. However,
it is often desirable to first move and then perform some I/O operation. In
order to avoid clumsiness, most I/O routines can be preceded by the prefix 'mv'
and the desired (y, x) coordinates prepended to the arguments to the function.
For example, the calls
move(y, x); addch(ch);can be replaced by
mvaddch(y, x, ch);and
wmove(win, y, x); waddch(win, ch);can be replaced by
mvwaddch(win, y, x, ch);Note that the window description pointer (win) comes before the added (y, x) coordinates. If a function requires a window pointer, it is always the first parameter passed.
curses library sets
some variables describing the terminal capabilities. type name description
------------------------------------------------------------------
int LINES number of lines on the terminal
int COLS number of columns on the terminal
The curses.h also introduces some #define
constants and types of general usefulness:
bool
bool doneit;)
TRUE
FALSE
ERR
OK
stdscr. These instructions will work on any window, providing you
change the function names and parameters as mentioned above.
Here is a sample program to motivate the discussion:
#include <curses.h>
#include <signal.h>
static void finish(int sig);
int
main(int argc, char *argv[])
{
int num = 0;
/* initialize your non-curses data structures here */
(void) signal(SIGINT, finish); /* arrange interrupts to terminate */
(void) initscr(); /* initialize the curses library */
keypad(stdscr, TRUE); /* enable keyboard mapping */
(void) nonl(); /* tell curses not to do NL->CR/NL on output */
(void) cbreak(); /* take input chars one at a time, no wait for \n */
(void) echo(); /* echo input - in color */
if (has_colors())
{
start_color();
/*
* Simple color assignment, often all we need. Color pair 0 cannot
* be redefined. This example uses the same value for the color
* pair as for the foreground color, though of course that is not
* necessary:
*/
init_pair(1, COLOR_RED, COLOR_BLACK);
init_pair(2, COLOR_GREEN, COLOR_BLACK);
init_pair(3, COLOR_YELLOW, COLOR_BLACK);
init_pair(4, COLOR_BLUE, COLOR_BLACK);
init_pair(5, COLOR_CYAN, COLOR_BLACK);
init_pair(6, COLOR_MAGENTA, COLOR_BLACK);
init_pair(7, COLOR_WHITE, COLOR_BLACK);
}
for (;;)
{
int c = getch(); /* refresh, accept single keystroke of input */
attrset(COLOR_PAIR(num % 8));
num++;
/* process the command keystroke */
}
finish(0); /* we're done */
}
static void finish(int sig)
{
endwin();
/* do your non-curses wrapup here */
exit(0);
}
curscr and stdscr must be allocated. These function
initscr() does both these things. Since it must allocate space for
the windows, it can overflow memory when attempting to do so. On the rare
occasions this happens, initscr() will terminate the program with
an error message. initscr() must always be called before any of the
routines which affect windows are used. If it is not, the program will core dump
as soon as either curscr or stdscr are referenced.
However, it is usually best to wait to call it until after you are sure you will
need it, like after checking for startup errors. Terminal status changing
routines like nl() and cbreak() should be called after
initscr().
Once the screen windows have been allocated, you can set them up for your
program. If you want to, say, allow a screen to scroll, use
scrollok(). If you want the cursor to be left in place after the
last change, use leaveok(). If this isn't done,
refresh() will move the cursor to the window's current (y, x)
coordinates after updating it.
You can create new windows of your own using the functions
newwin(), derwin(), and subwin(). The
routine delwin() will allow you to get rid of old windows. All the
options described above can be applied to any window.
addch() and move().
addch() adds a character at the current (y, x) coordinates.
move() changes the current (y, x) coordinates to whatever you want
them to be. It returns ERR if you try to move off the window. As
mentioned above, you can combine the two into mvaddch() to do both
things at once.
The other output functions, such as addstr() and
printw(), all call addch() to add characters to the
window.
After you have put on the window what you want there, when you want the
portion of the terminal covered by the window to be made to look like it, you
must call refresh(). In order to optimize finding changes,
refresh() assumes that any part of the window not changed since the
last refresh() of that window has not been changed on the terminal,
i.e., that you have not refreshed a portion of the terminal with an overlapping
window. If this is not the case, the routine touchwin() is provided
to make it look like the entire window has been changed, thus making
refresh() check the whole subsection of the terminal for changes.
If you call wrefresh() with curscr as its argument,
it will make the screen look like curscr thinks it looks like. This
is useful for implementing a command which would redraw the screen in case it
get messed up.
addch() is getch() which, if echo is set, will call
addch() to echo the character. Since the screen package needs to
know what is on the terminal at all times, if characters are to be echoed, the
tty must be in raw or cbreak mode. Since initially the terminal has echoing
enabled and is in ordinary ``cooked'' mode, one or the other has to changed
before calling getch(); otherwise, the program's output will be
unpredictable.
When you need to accept line-oriented input in a window, the functions
wgetstr() and friends are available. There is even a
wscanw() function that can do scanf()(3)-style
multi-field parsing on window input. These pseudo-line-oriented functions turn
on echoing while they execute.
The example code above uses the call keypad(stdscr, TRUE) to
enable support for function-key mapping. With this feature, the
getch() code watches the input stream for character sequences that
correspond to arrow and function keys. These sequences are returned as
pseudo-character values. The #define values returned are listed in
the curses.h The mapping from sequences to #define
values is determined by key_ capabilities in the terminal's
terminfo entry.
addch()
function (and some others, including box() and
border()) can accept some pseudo-character arguments which are
specially defined by ncurses. These are #define values
set up in the curses.h header; see there for a complete list (look
for the prefix ACS_).
The most useful of the ACS defines are the forms-drawing characters. You can
use these to draw boxes and simple graphs on the screen. If the terminal does
not have such characters, curses.h will map them to a recognizable
(though ugly) set of ASCII defaults.
ncurses package supports screen highlights including standout,
reverse-video, underline, and blink. It also supports color, which is treated as
another kind of highlight.
Highlights are encoded, internally, as high bits of the pseudo-character type
(chtype) that curses.h uses to represent the contents
of a screen cell. See the curses.h header file for a complete list
of highlight mask values (look for the prefix A_).
There are two ways to make highlights. One is to logical-or the value of the
highlights you want into the character argument of an addch() call,
or any other output call that takes a chtype argument.
The other is to set the current-highlight value. This is logical-or'ed with
any highlight you specify the first way. You do this with the functions
attron(), attroff(), and attrset(); see
the manual pages for details. Color is a special kind of highlight. The package
actually thinks in terms of color pairs, combinations of foreground and
background colors. The sample code above sets up eight color pairs, all of the
guaranteed-available colors on black. Note that each color pair is, in effect,
given the name of its foreground color. Any other range of eight non-conflicting
values could have been used as the first arguments of the
init_pair() values.
Once you've done an init_pair() that creates color-pair N, you
can use COLOR_PAIR(N) as a highlight that invokes that particular
color combination. Note that COLOR_PAIR(N), for constant N, is
itself a compile-time constant and can be used in initializers.
ncurses library
also provides a mouse interface.
NOTE: this facility is specific to
ncurses, it is not part of either the XSI Curses standard, nor of
System V Release 4, nor BSD curses. System V Release 4 curses contains code
with similar interface definitions, however it is not documented. Other than
by disassembling the library, we have no way to determine exactly how that
mouse code works. Thus, we recommend that you wrap mouse-related code in an
#ifdef using the feature macro NCURSES_MOUSE_VERSION so it will not be
compiled and linked on non-ncurses systems. Presently, mouse event
reporting works in the following environments:
gpm(1), Alessandro
Rubini's mouse server.
The mouse interface is very simple. To activate it, you use the function
mousemask(), passing it as first argument a bit-mask that specifies
what kinds of events you want your program to be able to see. It will return the
bit-mask of events that actually become visible, which may differ from the
argument if the mouse device is not capable of reporting some of the event types
you specify.
Once the mouse is active, your application's command loop should watch for a
return value of KEY_MOUSE from wgetch(). When you see
this, a mouse event report has been queued. To pick it off the queue, use the
function getmouse() (you must do this before the next
wgetch(), otherwise another mouse event might come in and make the
first one inaccessible).
Each call to getmouse() fills a structure (the address of which
you'll pass it) with mouse event data. The event data includes zero-origin,
screen-relative character-cell coordinates of the mouse pointer. It also
includes an event mask. Bits in this mask will be set, corresponding to the
event type being reported.
The mouse structure contains two additional fields which may be significant in the future as ncurses interfaces to new kinds of pointing device. In addition to x and y coordinates, there is a slot for a z coordinate; this might be useful with touch-screens that can return a pressure or duration parameter. There is also a device ID field, which could be used to distinguish between multiple pointing devices.
The class of visible events may be changed at any time via
mousemask(). Events that can be reported include presses, releases,
single-, double- and triple-clicks (you can set the maximum button-down time for
clicks). If you don't make clicks visible, they will be reported as
press-release pairs. In some environments, the event mask may include bits
reporting the state of shift, alt, and ctrl keys on the keyboard during the
event.
A function to check whether a mouse event fell within a given window is also supplied. You can use this to see whether a given window should consider a mouse event relevant to it.
Because mouse event reporting will not be available in all environments, it
would be unwise to build ncurses applications that require
the use of a mouse. Rather, you should use the mouse as a shortcut for
point-and-shoot commands your application would normally accept from the
keyboard. Two of the test games in the ncurses distribution
(bs and knight) contain code that illustrates how this
can be done.
See the manual page curs_mouse(3X) for full details of the
mouse-interface functions.
ncurses routines, the routine endwin() is provided. It
restores tty modes to what they were when initscr() was first
called, and moves the cursor down to the lower-left corner. Thus, anytime after
the call to initscr, endwin() should be called before exiting.
initscr()
initscr().
This will determine the terminal type and initialize curses data structures.
initscr() also arranges that the first call to
refresh() will clear the screen. If an error occurs a message is
written to standard error and the program exits. Otherwise it returns a
pointer to stdscr. A few functions may be called before initscr
(slk_init(), filter(), ripoffline(),
use_env(), and, if you are using multiple terminals,
newterm().)
endwin()
endwin() before exiting or
shelling out of the program. This function will restore tty modes, move the
cursor to the lower left corner of the screen, reset the terminal into the
proper non-visual mode. Calling refresh() or
doupdate() after a temporary escape from the program will restore
the ncurses screen from before the escape.
newterm(type, ofp, ifp)
newterm() instead of initscr().
newterm() should be called once for each terminal. It returns a
variable of type SCREEN * which should be saved as a reference to
that terminal. (NOTE: a SCREEN variable is not a screen in the sense
we are describing in this introduction, but a collection of parameters used to
assist in optimizing the display.) The arguments are the type of the terminal
(a string) and FILE pointers for the output and input of the
terminal. If type is NULL then the environment variable $TERM is
used. endwin() should called once at wrapup time for each
terminal opened using this function.
set_term(new)
newterm(). The screen reference for the new terminal is passed
as the parameter. The previous terminal is returned by the function. All other
calls affect only the current terminal.
delscreen(sp)
newterm(); deallocates the data structures
associated with a given SCREEN reference. refresh() and wrefresh(win)
wrefresh()
copies the named window to the physical terminal screen, taking into account
what is already there in order to do optimizations. refresh()
does a refresh of stdscr. Unless leaveok() has been
enabled, the physical cursor of the terminal is left at the location of the
window's cursor.
doupdate() and wnoutrefresh(win)
wnoutrefresh()), and then calling the
routine to update the screen (doupdate()). If the programmer
wishes to output several windows at once, a series of calls to
wrefresh will result in alternating calls to
wnoutrefresh() and doupdate(), causing several
bursts of output to the screen. By calling wnoutrefresh() for
each window, it is then possible to call doupdate() once,
resulting in only one burst of output, with fewer total characters transmitted
(this also avoids a visually annoying flicker at each update). setupterm(term, filenum, errret)
term is the character string representing the name of the
terminal being used. filenum is the UNIX file descriptor of the
terminal to be used for output. errret is a pointer to an
integer, in which a success or failure indication is returned. The values
returned can be 1 (all is well), 0 (no such terminal), or -1 (some problem
locating the terminfo database).
The value of term can be given as NULL, which will cause the
value of TERM in the environment to be used. The
errret pointer can also be given as NULL, meaning no error code
is wanted. If errret is defaulted, and something goes wrong,
setupterm() will print an appropriate error message and exit,
rather than returning. Thus, a simple program can call setupterm(0, 1, 0) and
not worry about initialization errors.
After the call to setupterm(), the global variable
cur_term is set to point to the current structure of terminal
capabilities. By calling setupterm() for each terminal, and
saving and restoring cur_term, it is possible for a program to
use two or more terminals at once. Setupterm() also stores the
names section of the terminal description in the global character array
ttytype[]. Subsequent calls to setupterm() will
overwrite this array, so you'll have to save it yourself if need be.
NOTE: These functions are not part of the standard curses API!
trace()
TRACE_ defines in the
curses.h file for details. (It is also possible to set a trace
level by assigning a trace level value to the environment variable
NCURSES_TRACE).
_tracef()
printf(), only it outputs a newline after the end of
arguments. The output goes to a file called trace in the current
directory. ncurses distribution
that can alleviate this problem somewhat; it compacts long sequences of similar
operations into more succinct single-line pseudo-operations. These pseudo-ops
can be distinguished by the fact that they are named in capital letters.
ncurses
manual pages are a complete reference for this library. In the remainder of this
document, we discuss various useful methods that may not be obvious from the
manual page descriptions.
noraw() or nocbreak(), think again and
move carefully. It's probably better design to use getstr() or one
of its relatives to simulate cooked mode. The noraw() and
nocbreak() functions try to restore cooked mode, but they may end
up clobbering some control bits set before you started your application. Also,
they have always been poorly documented, and are likely to hurt your
application's usability with other curses libraries.
Bear in mind that refresh() is a synonym for
wrefresh(stdscr). Don't try to mix use of stdscr with
use of windows declared by newwin(); a refresh() call
will blow them off the screen. The right way to handle this is to use
subwin(), or not touch stdscr at all and tile your
screen with declared windows which you then wnoutrefresh()
somewhere in your program event loop, with a single doupdate() call
to trigger actual repainting.
You are much less likely to run into problems if you design your screen
layouts to use tiled rather than overlapping windows. Historically, curses
support for overlapping windows has been weak, fragile, and poorly documented.
The ncurses library is not yet an exception to this rule.
There is a panels library included in the ncurses distribution
that does a pretty good job of strengthening the overlapping-windows facilities.
Try to avoid using the global variables LINES and COLS. Use
getmaxyx() on the stdscr context instead. Reason: your
code may be ported to run in an environment with window resizes, in which case
several screens could be open with different sizes.
ncurses.
To leave ncurses mode, call endwin() as you would
if you were intending to terminate the program. This will take the screen back
to cooked mode; you can do your shell-out. When you want to return to
ncurses mode, simply call refresh() or
doupdate(). This will repaint the screen.
There is a boolean function, isendwin(), which code can use to
test whether ncurses screen mode is active. It returns
TRUE in the interval between an endwin() call and the
following refresh(), FALSE otherwise.
Here is some sample code for shellout:
addstr("Shelling out...");
def_prog_mode(); /* save current tty modes */
endwin(); /* restore original tty modes */
system("sh"); /* run shell */
addstr("returned.\n"); /* prepare return message */
refresh(); /* restore save modes, repaint screen */
SIGWINCH to the application running under xterm. The easiest
way to handle SIGWINCH is to do an endwin, followed by
an refresh and a screen repaint you code yourself. The
refresh will pick up the new screen size from the xterm's
environment.
That is the standard way, of course (it even works with some vendor's curses
implementations). Its drawback is that it clears the screen to reinitialize the
display, and does not resize subwindows which must be shrunk.
Ncurses provides an extension which works better, the
resizeterm function. That function ensures that all windows are
limited to the new screen dimensions, and pads stdscr with blanks
if the screen is larger.
The ncurses library provides a SIGWINCH signal handler, which
pushes a KEY_RESIZE via the wgetch() calls. When
ncurses returns that code, it calls resizeterm to
update the size of the standard screen's window, repainting that (filling with
blanks or truncating as needed). It also resizes other windows, but its effect
may be less satisfactory because it cannot know how you want the screen
re-painted. You will usually have to write special-purpose code to handle
KEY_RESIZE yourself.
initscr() function actually calls a function named
newterm() to do most of its work. If you are writing a program that
opens multiple terminals, use newterm() directly.
For each call, you will have to specify a terminal type and a pair of file
pointers; each call will return a screen reference, and stdscr will
be set to the last one allocated. You will switch between screens with the
set_term call. Note that you will also have to call
def_shell_mode and def_prog_mode on each tty yourself.
ncurses mode. An easy way to do this is
to call setupterm(), then use the functions
tigetflag(), tigetnum(), and tigetstr()
to do your testing.
A particularly useful case of this often comes up when you want to test
whether a given terminal type should be treated as `smart' (cursor-addressable)
or `stupid'. The right way to test this is to see if the return value of
tigetstr("cup") is non-NULL. Alternatively, you can include the
term.h file and test the value of the macro
cursor_address.
addchstr()
family of functions for fast screen-painting of text when you know the text
doesn't contain any control characters. Try to make attribute changes infrequent
on your screens. Don't use the immedok() option!
wresize() function allows you to resize a window in place. The
associated resizeterm() function simplifies the construction of SIGWINCH
handlers, for resizing all windows.
The define_key() function allows you to define at runtime
function-key control sequences which are not in the terminal description. The
keyok() function allows you to temporarily enable or disable
interpretation of any function-key control sequence.
The use_default_colors() function allows you to construct
applications which can use the terminal's default foreground and background
colors as an additional "default" color. Several terminal emulators support this
feature, which is based on ISO 6429.
Ncurses supports up 16 colors, unlike SVr4 curses which defines only 8. While most terminals which provide color allow only 8 colors, about a quarter (including XFree86 xterm) support 16 colors.
ncurses and the
(undocumented!) behavior of older curses implementations. These arise from
ambiguities or omissions in the documentation of the API.
curses
versions were often not documented precisely.
To understand why this is a problem, remember that screen updates are calculated between two representations of the entire display. The documentation says that when you refresh a window, it is first copied to the virtual screen, and then changes are calculated to update the physical screen (and applied to the terminal). But "copied to" is not very specific, and subtle differences in how copying works can produce different behaviors in the case where two overlapping windows are each being refreshed at unpredictable intervals.
What happens to the overlapping region depends on what
wnoutrefresh() does with its argument -- what portions of the
argument window it copies to the virtual screen. Some implementations do "change
copy", copying down only locations in the window that have changed (or been
marked changed with wtouchln() and friends). Some implementations
do "entire copy", copying all window locations to the virtual screen
whether or not they have changed.
The ncurses library itself has not always been consistent on
this score. Due to a bug, versions 1.8.7 to 1.9.8a did entire copy. Versions
1.8.6 and older, and versions 1.9.9 and newer, do change copy.
For most commercial curses implementations, it is not documented and not
known for sure (at least not to the ncurses maintainers) whether
they do change copy or entire copy. We know that System V release 3 curses has
logic in it that looks like an attempt to do change copy, but the surrounding
logic and data representations are sufficiently complex, and our knowledge
sufficiently indirect, that it's hard to know whether this is reliable. It is
not clear what the SVr4 documentation and XSI standard intend. The XSI Curses
standard barely mentions wnoutrefresh(); the SVr4 documents seem to be
describing entire-copy, but it is possible with some effort and straining to
read them the other way.
It might therefore be unwise to rely on either behavior in programs that
might have to be linked with other curses implementations. Instead, you can do
an explicit touchwin() before the wnoutrefresh() call
to guarantee an entire-contents copy anywhere.
The really clean way to handle this is to use the panels library. If, when
you want a screen update, you do update_panels(), it will do all
the necessary wnoutrefresh() calls for whatever panel stacking
order you have defined. Then you can do one doupdate() and there
will be a single burst of physical I/O that will do all your updates.
ncurses (1.8.7 or older) you may be surprised by the
behavior of the erase functions. In older versions, erased areas of a window
were filled with a blank modified by the window's current attribute (as set by
wattrset(), wattron(),
wattroff() and friends).
In newer versions, this is not so. Instead, the attribute of erased blanks is
normal unless and until it is modified by the functions bkgdset()
or wbkgdset().
This change in behavior conforms ncurses to System V Release 4
and the XSI Curses standard.
ncurses
library is intended to be base-level conformant with the XSI Curses standard
from X/Open. Many extended-level features (in fact, almost all features not
directly concerned with wide characters and internationalization) are also
supported.
One effect of XSI conformance is the change in behavior described under "Background Erase -- Compatibility with Old Versions".
Also, ncurses meets the XSI requirement that every macro entry
point have a corresponding function which may be linked (and will be
prototype-checked) if the macro definition is disabled with #undef.
ncurses library
by itself provides good support for screen displays in which the windows are
tiled (non-overlapping). In the more general case that windows may overlap, you
have to use a series of wnoutrefresh() calls followed by a
doupdate(), and be careful about the order you do the window
refreshes in. It has to be bottom-upwards, otherwise parts of windows that
should be obscured will show through.
When your interface design is such that windows may dive deeper into the visibility stack or pop to the top at runtime, the resulting book-keeping can be tedious and difficult to get right. Hence the panels library.
The panel library first appeared in AT&T System V. The
version documented here is the panel code distributed with
ncurses.
#include <panel.h>and must be linked explicitly with the panels library using an
-lpanel argument. Note that they must also link the
ncurses library with -lncurses. Many linkers are
two-pass and will accept either order, but it is still good practice to put
-lpanel first and -lncurses second.
refresh()) that
displays all panels in the deck in the proper order to resolve overlaps. The
standard window, stdscr, is considered below all panels.
Details on the panels functions are available in the man pages. We'll just hit the highlights here.
You create a panel from a window by calling new_panel() on a
window pointer. It then becomes the top of the deck. The panel's window is
available as the value of panel_window() called with the panel
pointer as argument.
You can delete a panel (removing it from the deck) with
del_panel. This will not deallocate the associated window; you have
to do that yourself. You can replace a panel's window with a different window by
calling replace_window. The new window may be of different size;
the panel code will re-compute all overlaps. This operation doesn't change the
panel's position in the deck.
To move a panel's window, use move_panel(). The
mvwin() function on the panel's window isn't sufficient because it
doesn't update the panels library's representation of where the windows are.
This operation leaves the panel's depth, contents, and size unchanged.
Two functions (top_panel(), bottom_panel()) are
provided for rearranging the deck. The first pops its argument window to the top
of the deck; the second sends it to the bottom. Either operation leaves the
panel's screen location, contents, and size unchanged.
The function update_panels() does all the
wnoutrefresh() calls needed to prepare for doupdate()
(which you must call yourself, afterwards).
Typically, you will want to call update_panels() and
doupdate() just before accepting command input, once in each cycle
of interaction with the user. If you call update_panels() after
each and every panel write, you'll generate a lot of unnecessary refresh
activity and screen flicker.
wnoutrefresh() or wrefresh() operations with
panels code; this will work only if the argument window is either in the top
panel or unobscured by any other panels.
The stsdcr window is a special case. It is considered below all
panels. Because changes to panels may obscure parts of stdscr,
though, you should call update_panels() before
doupdate() even when you only change stdscr.
Note that wgetch automatically calls wrefresh.
Therefore, before requesting input from a panel window, you need to be sure that
the panel is totally unobscured.
There is presently no way to display changes to one obscured panel without repainting all panels.
hide_panel for this. Use
show_panel() to render it visible again. The predicate function
panel_hidden tests whether or not a panel is hidden.
The panel_update code ignores hidden panels. You cannot do
top_panel() or bottom_panel on a hidden panel(). Other
panels operations are applicable.
panel_above() and
panel_below. Handed a panel pointer, they return the panel above or
below that panel. Handed NULL, they return the bottom-most or
top-most panel.
Every panel has an associated user pointer, not used by the panel code, to
which you can attach application data. See the man page documentation of
set_panel_userptr() and panel_userptr for details.
menu library is a curses extension that supports easy programming
of menu hierarchies with a uniform but flexible interface.
The menu library first appeared in AT&T System V. The
version documented here is the menu code distributed with
ncurses.
#include <menu.h>and must be linked explicitly with the menus library using an
-lmenu argument. Note that they must also link the
ncurses library with -lncurses. Many linkers are
two-pass and will accept either order, but it is still good practice to put
-lmenu first and -lncurses second.
The menu can then by posted, that is written to an associated window. Actually, each menu has two associated windows; a containing window in which the programmer can scribble titles or borders, and a subwindow in which the menu items proper are displayed. If this subwindow is too small to display all the items, it will be a scrollable viewport on the collection of items.
A menu may also be unposted (that is, undisplayed), and finally freed to make the storage associated with it and its items available for re-use.
The general flow of control of a menu program looks like this:
curses.
new_item().
new_menu().
post_menu().
unpost_menu().
free_menu().
free_item().
curses. menu_opts(3x) to see
how to change the default). Both types always have a current item.
From a single-valued menu you can read the selected value simply by looking
at the current item. From a multi-valued menu, you get the selected set by
looping through the items applying the item_value() predicate
function. Your menu-processing code can use the function
set_item_value() to flag the items in the select set.
Menu items can be made unselectable using set_item_opts() or
item_opts_off() with the O_SELECTABLE argument. This
is the only option so far defined for menus, but it is good practice to code as
though other option bits might be on.
set_menu_format() allows you to set the
maximum size of the viewport or menu page that will be used to
display menu items. You can retrieve any format associated with a menu with
menu_format(). The default format is rows=16, columns=1.
The actual menu page may be smaller than the format size. This depends on the item number and size and whether O_ROWMAJOR is on. This option (on by default) causes menu items to be displayed in a `raster-scan' pattern, so that if more than one item will fit horizontally the first couple of items are side-by-side in the top row. The alternative is column-major display, which tries to put the first several items in the first column.
As mentioned above, a menu format not large enough to allow all items to fit on-screen will result in a menu display that is vertically scrollable.
You can scroll it with requests to the menu driver, which will be described in the section on menu input handling.
Each menu has a mark string used to visually tag selected items;
see the menu_mark(3x) manual page for details. The mark string
length also influences the menu page size.
The function scale_menu() returns the minimum display size that
the menu code computes from all these factors. There are other menu display
attributes including a select attribute, an attribute for selectable items, an
attribute for unselectable items, and a pad character used to separate item name
text from description text. These have reasonable defaults which the library
allows you to change (see the menu_attribs(3x) manual page.
The outer or frame window is not otherwise touched by the menu routines. It exists so the programmer can associate a title, a border, or perhaps help text with the menu and have it properly refreshed or erased at post/unpost time. The inner window or subwindow is where the current menu page is displayed.
By default, both windows are stdscr. You can set them with the
functions in menu_win(3x).
When you call post_menu(), you write the menu to its subwindow.
When you call unpost_menu(), you erase the subwindow, However,
neither of these actually modifies the screen. To do that, call
wrefresh() or some equivalent.
menu_driver() repeatedly. The
first argument of this routine is a menu pointer; the second is a menu command
code. You should write an input-fetching routine that maps input characters to
menu command codes, and pass its output to menu_driver(). The menu
command codes are fully documented in menu_driver(3x).
The simplest group of command codes is REQ_NEXT_ITEM,
REQ_PREV_ITEM, REQ_FIRST_ITEM,
REQ_LAST_ITEM, REQ_UP_ITEM,
REQ_DOWN_ITEM, REQ_LEFT_ITEM,
REQ_RIGHT_ITEM. These change the currently selected item. These
requests may cause scrolling of the menu page if it only partially displayed.
There are explicit requests for scrolling which also change the current item
(because the select location does not change, but the item there does). These
are REQ_SCR_DLINE, REQ_SCR_ULINE,
REQ_SCR_DPAGE, and REQ_SCR_UPAGE.
The REQ_TOGGLE_ITEM selects or deselects the current item. It is
for use in multi-valued menus; if you use it with O_ONEVALUE on,
you'll get an error return (E_REQUEST_DENIED).
Each menu has an associated pattern buffer. The menu_driver()
logic tries to accumulate printable ASCII characters passed in in that buffer;
when it matches a prefix of an item name, that item (or the next matching item)
is selected. If appending a character yields no new match, that character is
deleted from the pattern buffer, and menu_driver() returns
E_NO_MATCH.
Some requests change the pattern buffer directly:
REQ_CLEAR_PATTERN, REQ_BACK_PATTERN,
REQ_NEXT_MATCH, REQ_PREV_MATCH. The latter two are
useful when pattern buffer input matches more than one item in a multi-valued
menu.
Each successful scroll or item navigation request clears the pattern buffer.
It is also possible to set the pattern buffer explicitly with
set_menu_pattern().
Finally, menu driver requests above the constant MAX_COMMAND are
considered application-specific commands. The menu_driver() code
ignores them and returns E_UNKNOWN_COMMAND.
menu_opts(3x) for details.
It is possible to change the current item from application code; this is
useful if you want to write your own navigation requests. It is also possible to
explicitly set the top row of the menu display. See
mitem_current(3x). If your application needs to change the menu
subwindow cursor for any reason, pos_menu_cursor() will restore it
to the correct location for continuing menu driver processing.
It is possible to set hooks to be called at menu initialization and wrapup
time, and whenever the selected item changes. See menu_hook(3x).
Each item, and each menu, has an associated user pointer on which you can
hang application data. See mitem_userptr(3x) and
menu_userptr(3x).
form library is a
curses extension that supports easy programming of on-screen forms for data
entry and program control.
The form library first appeared in AT&T System V. The
version documented here is the form code distributed with
ncurses.
#include <form.h>and must be linked explicitly with the forms library using an
-lform argument. Note that they must also link the
ncurses library with -lncurses. Many linkers are
two-pass and will accept either order, but it is still good practice to put
-lform first and -lncurses second.
To make forms, you create groups of fields and connect them with form frame objects; the form library makes this relatively simple.
Once defined, a form can be posted, that is written to an associated window. Actually, each form has two associated windows; a containing window in which the programmer can scribble titles or borders, and a subwindow in which the form fields proper are displayed.
As the form user fills out the posted form, navigation and editing keys
support movement between fields, editing keys support modifying field, and plain
text adds to or changes data in a current field. The form library allows you
(the forms designer) to bind each navigation and editing key to any keystroke
accepted by curses Fields may have validation conditions on them,
so that they check input data for type and value. The form library supplies a
rich set of pre-defined field types, and makes it relatively easy to define new
ones.
Once its transaction is completed (or aborted), a form may be unposted (that is, undisplayed), and finally freed to make the storage associated with it and its items available for re-use.
The general flow of control of a form program looks like this:
curses.
new_field().
new_form().
post_form().
unpost_form().
free_form().
free_field().
curses. In forms programs, however, the `process user requests' is somewhat more complicated than for menus. Besides menu-like navigation operations, the menu driver loop has to support field editing and data validation.
new_field(): FIELD *new_field(int height, int width, /* new field size */
int top, int left, /* upper left corner */
int offscreen, /* number of offscreen rows */
int nbuf); /* number of working buffers */
Menu items always occupy a single row, but forms fields may have multiple
rows. So new_field() requires you to specify a width and height
(the first two arguments, which mist both be greater than zero).
You must also specify the location of the field's upper left corner on the
screen (the third and fourth arguments, which must be zero or greater). Note
that these coordinates are relative to the form subwindow, which will coincide
with stdscr by default but need not be stdscr if
you've done an explicit set_form_win() call.
The fifth argument allows you to specify a number of off-screen rows. If this
is zero, the entire field will always be displayed. If it is nonzero, the form
will be scrollable, with only one screen-full (initially the top part) displayed
at any given time. If you make a field dynamic and grow it so it will no longer
fit on the screen, the form will become scrollable even if the
offscreen argument was initially zero.
The forms library allocates one working buffer per field; the size of each
buffer is ((height + offscreen)*width + 1, one character for each
position in the field plus a NUL terminator. The sixth argument is the number of
additional data buffers to allocate for the field; your application can use them
for its own purposes.
FIELD *dup_field(FIELD *field, /* field to copy */
int top, int left); /* location of new copy */
The function dup_field() duplicates an existing field at a
new location. Size and buffering information are copied; some attribute flags
and status bits are not (see the form_field_new(3X) for details). FIELD *link_field(FIELD *field, /* field to copy */
int top, int left); /* location of new copy */
The function link_field() also duplicates an existing field
at a new location. The difference from dup_field() is that it
arranges for the new field's buffer to be shared with the old one.
Besides the obvious use in making a field editable from two different form pages, linked fields give you a way to hack in dynamic labels. If you declare several fields linked to an original, and then make them inactive, changes from the original will still be propagated to the linked fields.
As with duplicated fields, linked fields have attribute bits separate from the original.
As you might guess, all these field-allocations return NULL if
the field allocation is not possible due to an out-of-memory error or
out-of-bounds arguments.
To connect fields to a form, use
FORM *new_form(FIELD **fields);This function expects to see a NULL-terminated array of field pointers. Said fields are connected to a newly-allocated form object; its address is returned (or else NULL if the allocation fails).
Note that new_field() does not copy the pointer array
into private storage; if you modify the contents of the pointer array during
forms processing, all manner of bizarre things might happen. Also note that any
given field may only be connected to one form.
The functions free_field() and free_form are
available to free field and form objects. It is an error to attempt to free a
field connected to a form, but not vice-versa; thus, you will generally free
your form objects first.
O_STATIC bit) involve sufficient complications to
be covered in sections of their own later on. We cover the functions used to get
and set several basic attributes here.
When a field is created, the attributes not specified by the
new_field function are copied from an invisible system default
field. In attribute-setting and -fetching functions, the argument NULL is taken
to mean this field. Changes to it persist as defaults until your forms
application terminates.
int field_info(FIELD *field, /* field from which to fetch */
int *height, *int width, /* field size */
int *top, int *left, /* upper left corner */
int *offscreen, /* number of offscreen rows */
int *nbuf); /* number of working buffers */
This function is a sort of inverse of new_field(); instead of
setting size and location attributes of a new field, it fetches them from an
existing one.
int move_field(FIELD *field, /* field to alter */
int top, int left); /* new upper-left corner */
You can, of course. query the current location through
field_info().
int set_field_just(FIELD *field, /* field to alter */
int justmode); /* mode to set */
int field_just(FIELD *field); /* fetch mode of field */
The mode values accepted and returned by this functions are preprocessor
macros NO_JUSTIFICATION, JUSTIFY_RIGHT,
JUSTIFY_LEFT, or JUSTIFY_CENTER.
This group of four field attributes controls the visual appearance of the field on the screen, without affecting in any way the data in the field buffer.
int set_field_fore(FIELD *field, /* field to alter */
chtype attr); /* attribute to set */
chtype field_fore(FIELD *field); /* field to query */
int set_field_back(FIELD *field, /* field to alter */
chtype attr); /* attribute to set */
chtype field_back(FIELD *field); /* field to query */
int set_field_pad(FIELD *field, /* field to alter */
int pad); /* pad character to set */
chtype field_pad(FIELD *field);
int set_new_page(FIELD *field, /* field to alter */
int flag); /* TRUE to force new page */
chtype new_page(FIELD *field); /* field to query */
The attributes set and returned by the first four functions are normal
curses(3x) display attribute values (A_STANDOUT,
A_BOLD, A_REVERSE etc). The page bit of a field
controls whether it is displayed at the start of a new form screen.
int set_field_opts(FIELD *field, /* field to alter */
int attr); /* attribute to set */
int field_opts_on(FIELD *field, /* field to alter */
int attr); /* attributes to turn on */
int field_opts_off(FIELD *field, /* field to alter */
int attr); /* attributes to turn off */
int field_opts(FIELD *field); /* field to query */
By default, all options are on. Here are the available option bits:
REQ_PREV_CHOICE and
REQ_NEXT_CHOICE will fail. Such read-only fields may be useful
for help messages.
The option values are bit-masks and can be composed with logical-or in the obvious way.
int set_field_status(FIELD *field, /* field to alter */
int status); /* mode to set */
int field_status(FIELD *field); /* fetch mode of field */
Setting this flag under program control can be useful if you use the same
form repeatedly, looking for modified fields each time.
Calling field_status() on a field not currently selected for
input will return a correct value. Calling field_status() on a
field that is currently selected for input may not necessarily give a correct
field status value, because entered data isn't necessarily copied to buffer zero
before the exit validation check. To guarantee that the returned status value
reflects reality, call field_status() either (1) in the field's
exit validation check routine, (2) from the field's or form's initialization or
termination hooks, or (3) just after a REQ_VALIDATION request has
been processed by the forms driver.
int set_field_userptr(FIELD *field, /* field to alter */
char *userptr); /* mode to set */
char *field_userptr(FIELD *field); /* fetch mode of field */
(Properly, this user pointer field ought to have (void *)
type. The (char *) type is retained for System V compatibility.)
It is valid to set the user pointer of the default field (with a
set_field_userptr() call passed a NULL field pointer.) When a new
field is created, the default-field user pointer is copied to initialize the new
field's user pointer.
A one-line dynamic field will have a fixed height (1) but variable width, scrolling horizontally to display data within the field area as originally dimensioned and located. A multi-line dynamic field will have a fixed width, but variable height (number of rows), scrolling vertically to display data within the field area as originally dimensioned and located.
Normally, a dynamic field is allowed to grow without limit. But it is possible to set an upper limit on the size of a dynamic field. You do it with this function:
int set_max_field(FIELD *field, /* field to alter (may not be NULL) */
int max_size); /* upper limit on field size */
If the field is one-line, max_size is taken to be a column
size limit; if it is multi-line, it is taken to be a line size limit. To disable
any limit, use an argument of zero. The growth limit can be changed whether or
not the O_STATIC bit is on, but has no effect until it is.
The following properties of a field change when it becomes dynamic:
O_AUTOSKIP and O_NL_OVERLOAD are ignored.
dup_field() and link_field() calls copy
dynamic-buffer sizes. If the O_STATIC option is set on one of a
collection of links, buffer resizing will occur only when the field is edited
through that link.
field_info() will retrieve the original static size
of the field; use dynamic_field_info() to get the actual dynamic
size. A field's validation check (if any) is not called when
set_field_buffer() modifies the input buffer, nor when that buffer
is changed through a linked field.
The form library provides a rich set of pre-defined validation
types, and gives you the capability to define custom ones of your own. You can
examine and change field validation attributes with the following functions:
int set_field_type(FIELD *field, /* field to alter */
FIELDTYPE *ftype, /* type to associate */
...); /* additional arguments*/
FIELDTYPE *field_type(FIELD *field); /* field to query */
The validation type of a field is considered an attribute of the field. As
with other field attributes, Also, doing set_field_type() with a
NULL field default will change the system default for validation of
newly-created fields.
Here are the pre-defined validation types:
int set_field_type(FIELD *field, /* field to alter */
TYPE_ALPHA, /* type to associate */
int width); /* maximum width of field */
The width argument sets a minimum width of data. Typically
you'll want to set this to the field width; if it's greater than the field
width, the validation check will always fail. A minimum width of zero makes
field completion optional.
int set_field_type(FIELD *field, /* field to alter */
TYPE_ALNUM, /* type to associate */
int width); /* maximum width of field */
The width argument sets a minimum width of data. As with
TYPE_ALPHA, typically you'll want to set this to the field width; if it's
greater than the field width, the validation check will always fail. A minimum
width of zero makes field completion optional.
int set_field_type(FIELD *field, /* field to alter */
TYPE_ENUM, /* type to associate */
char **valuelist; /* list of possible values */
int checkcase; /* case-sensitive? */
int checkunique); /* must specify uniquely? */
The valuelist parameter must point at a NULL-terminated list
of valid strings. The checkcase argument, if true, makes comparison
with the string case-sensitive.
When the user exits a TYPE_ENUM field, the validation procedure tries to complete the data in the buffer to a valid entry. If a complete choice string has been entered, it is of course valid. But it is also possible to enter a prefix of a valid string and have it completed for you.
By default, if you enter such a prefix and it matches more than one value in
the string list, the prefix will be completed to the first matching value. But
the checkunique argument, if true, requires prefix matches to be
unique in order to be valid.
The REQ_NEXT_CHOICE and REQ_PREV_CHOICE input
requests can be particularly useful with these fields.
int set_field_type(FIELD *field, /* field to alter */
TYPE_INTEGER, /* type to associate */
int padding, /* # places to zero-pad to */
int vmin, int vmax); /* valid range */
Valid characters consist of an optional leading minus and digits. The
range check is performed on exit. If the range maximum is less than or equal to
the minimum, the range is ignored.
If the value passes its range check, it is padded with as many leading zero digits as necessary to meet the padding argument.
A TYPE_INTEGER value buffer can conveniently be interpreted with
the C library function atoi(3).
int set_field_type(FIELD *field, /* field to alter */
TYPE_NUMERIC, /* type to associate */
int padding, /* # places of precision */
double vmin, double vmax); /* valid range */
Valid characters consist of an optional leading minus and digits. possibly
including a decimal point. If your system supports locale's, the decimal point
character used must be the one defined by your locale. The range check is
performed on exit. If the range maximum is less than or equal to the minimum,
the range is ignored.
If the value passes its range check, it is padded with as many trailing zero digits as necessary to meet the padding argument.
A TYPE_NUMERIC value buffer can conveniently be interpreted with
the C library function atof(3).
int set_field_type(FIELD *field, /* field to alter */
TYPE_REGEXP, /* type to associate */
char *regexp); /* expression to match */
The syntax for regular expressions is that of regcomp(3). The
check for regular-expression match is performed on exit.
char *field_buffer(FIELD *field, /* field to query */
int bufindex); /* number of buffer to query */
Normally, the state of the zero-numbered buffer for each field is set by
the user's editing actions on that field. It's sometimes useful to be able to
set the value of the zero-numbered (or some other) buffer from your application:
int set_field_buffer(FIELD *field, /* field to alter */
int bufindex, /* number of buffer to alter */
char *value); /* string value to set */
If the field is not large enough and cannot be resized to a sufficiently
large size to contain the specified value, the value will be truncated to fit.
Calling field_buffer() with a null field pointer will raise an
error. Calling field_buffer() on a field not currently selected for
input will return a correct value. Calling field_buffer() on a
field that is currently selected for input may not necessarily give a correct
field buffer value, because entered data isn't necessarily copied to buffer zero
before the exit validation check. To guarantee that the returned buffer value
reflects on-screen reality, call field_buffer() either (1) in the
field's exit validation check routine, (2) from the field's or form's
initialization or termination hooks, or (3) just after a
REQ_VALIDATION request has been processed by the forms driver.
NULL.
The principal attribute of a form is its field list. You can query and change this list with:
int set_form_fields(FORM *form, /* form to alter */
FIELD **fields); /* fields to connect */
char *form_fields(FORM *form); /* fetch fields of form */
int field_count(FORM *form); /* count connect fields */
The second argument of set_form_fields() may be a
NULL-terminated field pointer array like the one required by
new_form(). In that case, the old fields of the form are
disconnected but not freed (and eligible to be connected to other forms), then
the new fields are connected.
It may also be null, in which case the old fields are disconnected (and not freed) but no new ones are connected.
The field_count() function simply counts the number of fields
connected to a given from. It returns -1 if the form-pointer argument is NULL.
stdscr.
By making this step explicit, you can associate a form with a declared frame window on your screen display. This can be useful if you want to adapt the form display to different screen sizes, dynamically tile forms on the screen, or use a form as part of an interface layout managed by panels.
The two windows associated with each form have the same functions as their analogues in the menu library. Both these windows are painted when the form is posted and erased when the form is unposted.
The outer or frame window is not otherwise touched by the form routines. It exists so the programmer can associate a title, a border, or perhaps help text with the form and have it properly refreshed or erased at post/unpost time. The inner window or subwindow is where the current form page is actually displayed.
In order to declare your own frame window for a form, you'll need to know the size of the form's bounding rectangle. You can get this information with:
int scale_form(FORM *form, /* form to query */
int *rows, /* form rows */
int *cols); /* form cols */
The form dimensions are passed back in the locations pointed to by the
arguments. Once you have this information, you can use it to declare of windows,
then use one of these functions: int set_form_win(FORM *form, /* form to alter */
WINDOW *win); /* frame window to connect */
WINDOW *form_win(FORM *form); /* fetch frame window of form */
int set_form_sub(FORM *form, /* form to alter */
WINDOW *win); /* form subwindow to connect */
WINDOW *form_sub(FORM *form); /* fetch form subwindow of form */
Note that curses operations, including refresh(), on the
form, should be done on the frame window, not the form subwindow.
It is possible to check from your application whether all of a scrollable field is actually displayed within the menu subwindow. Use these functions:
int data_ahead(FORM *form); /* form to be queried */ int data_behind(FORM *form); /* form to be queried */The function
data_ahead() returns TRUE if (a) the current
field is one-line and has undisplayed data off to the right, (b) the current
field is multi-line and there is data off-screen below it.
The function data_behind() returns TRUE if the first (upper left
hand) character position is off-screen (not being displayed).
Finally, there is a function to restore the form window's cursor to the value expected by the forms driver:
int pos_form_cursor(FORM *) /* form to be queried */If your application changes the form window cursor, call this function before handing control back to the forms driver in order to re-synchronize it.
form_driver() handles virtualized input requests for form
navigation, editing, and validation requests, just as menu_driver
does for menus (see the section on menu input
handling). int form_driver(FORM *form, /* form to pass input to */
int request); /* form request code */
Your input virtualization function needs to take input and then convert it
to either an alphanumeric character (which is treated as data to be entered in
the currently-selected field), or a forms processing request.
The forms driver provides hooks (through input-validation and field-termination functions) with which your application code can check that the input taken by the driver matched what was expected.
REQ_NEXT_PAGE
REQ_PREV_PAGE
REQ_FIRST_PAGE
REQ_LAST_PAGE
REQ_NEXT_PAGE from the last page goes to the
first, and REQ_PREV_PAGE from the first page goes to the last.
REQ_NEXT_FIELD
REQ_PREV_FIELD
REQ_FIRST_FIELD
REQ_LAST_FIELD
REQ_SNEXT_FIELD
REQ_SPREV_FIELD
REQ_SFIRST_FIELD
REQ_SLAST_FIELD
REQ_LEFT_FIELD
REQ_RIGHT_FIELD
REQ_UP_FIELD
REQ_DOWN_FIELD
REQ_NEXT_FIELD from the last field goes to
the first, and REQ_PREV_FIELD from the first field goes to the
last. The order of the fields for these (and the REQ_FIRST_FIELD
and REQ_LAST_FIELD requests) is simply the order of the field
pointers in the form array (as set up by new_form() or
set_form_fields()
It is also possible to traverse the fields as if they had been sorted in screen-position order, so the sequence goes left-to-right and top-to-bottom. To do this, use the second group of four sorted-movement requests.
Finally, it is possible to move between fields using visual directions up, down, right, and left. To accomplish this, use the third group of four requests. Note, however, that the position of a form for purposes of these requests is its upper-left corner.
For example, suppose you have a multi-line field B, and two single-line
fields A and C on the same line with B, with A to the left of B and C to the
right of B. A REQ_MOVE_RIGHT from A will go to B only if A, B, and
C all share the same first line; otherwise it will skip over B to C.
REQ_NEXT_CHAR
REQ_PREV_CHAR
REQ_NEXT_LINE
REQ_PREV_LINE
REQ_NEXT_WORD
REQ_PREV_WORD
REQ_BEG_FIELD
REQ_END_FIELD
REQ_BEG_LINE
REQ_END_LINE
REQ_LEFT_CHAR
REQ_RIGHT_CHAR
REQ_UP_CHAR
REQ_DOWN_CHAR
REQ_SCR_FLINE
REQ_SCR_BLINE
REQ_SCR_FPAGE
REQ_SCR_BPAGE
REQ_SCR_FHPAGE
REQ_SCR_BHPAGE
REQ_SCR_FCHAR
REQ_SCR_BCHAR
REQ_SCR_HFLINE
REQ_SCR_HBLINE
REQ_SCR_HFHALF
REQ_SCR_HBHALF
The following requests support editing the field and changing the edit mode:
REQ_INS_MODE
REQ_OVL_MODE
REQ_NEW_LINE
REQ_INS_CHAR
REQ_INS_LINE
REQ_DEL_CHAR
REQ_DEL_PREV
REQ_DEL_LINE
REQ_DEL_WORD
REQ_CLR_EOL
REQ_CLR_EOF
REQ_CLEAR_FIELD
REQ_NEW_LINE
and REQ_DEL_PREV requests is complicated and partly controlled by a
pair of forms options. The special cases are triggered when the cursor is at the
beginning of a field, or on the last line of the field.
First, we consider REQ_NEW_LINE:
The normal behavior of REQ_NEW_LINE in insert mode is to break
the current line at the position of the edit cursor, inserting the portion of
the current line after the cursor as a new line following the current and moving
the cursor to the beginning of that new line (you may think of this as inserting
a newline in the field buffer).
The normal behavior of REQ_NEW_LINE in overlay mode is to clear
the current line from the position of the edit cursor to end of line. The cursor
is then moved to the beginning of the next line.
However, REQ_NEW_LINE at the beginning of a field, or on the
last line of a field, instead does a REQ_NEXT_FIELD.
O_NL_OVERLOAD option is off, this special action is disabled.
Now, let us consider REQ_DEL_PREV:
The normal behavior of REQ_DEL_PREV is to delete the previous
character. If insert mode is on, and the cursor is at the start of a line, and
the text on that line will fit on the previous one, it instead appends the
contents of the current line to the previous one and deletes the current line
(you may think of this as deleting a newline from the field buffer).
However, REQ_DEL_PREV at the beginning of a field is instead
treated as a REQ_PREV_FIELD.
If the O_BS_OVERLOAD option is off, this special action is
disabled and the forms driver just returns E_REQUEST_DENIED.
See Form Options for discussion of how to set and clear the overload options.
REQ_NEXT_CHOICE
REQ_PREV_CHOICE
TYPE_ENUM has built-in successor and
predecessor functions. When you define a field type of your own (see Custom
Validation Types), you can associate our own ordering functions.
curses value greater than
KEY_MAX and less than or equal to the constant
MAX_COMMAND. If your input-virtualization routine returns a value
above MAX_COMMAND, the forms driver will ignore it.
typedef void (*HOOK)(); /* pointer to function returning void */
int set_form_init(FORM *form, /* form to alter */
HOOK hook); /* initialization hook */
HOOK form_init(FORM *form); /* form to query */
int set_form_term(FORM *form, /* form to alter */
HOOK hook); /* termination hook */
HOOK form_term(FORM *form); /* form to query */
int set_field_init(FORM *form, /* form to alter */
HOOK hook); /* initialization hook */
HOOK field_init(FORM *form); /* form to query */
int set_field_term(FORM *form, /* form to alter */
HOOK hook); /* termination hook */
HOOK field_term(FORM *form); /* form to query */
These functions allow you to either set or query four different hooks. In
each of the set functions, the second argument should be the address of a hook
function. These functions differ only in the timing of the hook call.
set_current_field() call
set_form_page() call
You can set a default hook for all fields by passing one of the set functions a NULL first argument.
You can disable any of these hooks by (re)setting them to NULL, the default value.
int set_current_field(FORM *form, /* form to alter */
FIELD *field); /* field to shift to */
FIELD *current_field(FORM *form); /* form to query */
int field_index(FORM *form, /* form to query */
FIELD *field); /* field to get index of */
The function field_index() returns the index of the given
field in the given form's field array (the array passed to
new_form() or set_form_fields()).
The initial current field of a form is the first active field on the first
page. The function set_form_fields() resets this.
It is also possible to move around by pages.
int set_form_page(FORM *form, /* form to alter */
int page); /* page to go to (0-origin) */
int form_page(FORM *form); /* return form's current page */
The initial page of a newly-created form is 0. The function
set_form_fields() resets this.
int set_form_opts(FORM *form, /* form to alter */
int attr); /* attribute to set */
int form_opts_on(FORM *form, /* form to alter */
int attr); /* attributes to turn on */
int form_opts_off(FORM *form, /* form to alter */
int attr); /* attributes to turn off */
int form_opts(FORM *form); /* form to query */
By default, all options are on. Here are the available option bits:
REQ_NEW_LINE as described in Editing
Requests. The value of this option is ignored on dynamic fields that have
not reached their size limit; these have no last line, so the circumstances
for triggering a REQ_NEXT_FIELD never arise.
REQ_DEL_PREV as described in Editing
Requests. form
library gives you the capability to define custom validation types of your own.
Further, the optional additional arguments of set_field_type
effectively allow you to parameterize validation types. Most of the
complications in the validation-type interface have to do with the handling of
the additional arguments within custom validation functions.
FIELD *link_fieldtype(FIELDTYPE *type1,
FIELDTYPE *type2);
This function creates a field type that will accept any of the values
legal for either of its argument field types (which may be either predefined or
programmer-defined). If a set_field_type() call later requires
arguments, the new composite type expects all arguments for the first type, than
all arguments for the second. Order functions (see Order
Requests) associated with the component types will work on the composite;
what it does is check the validation function for the first type, then for the
second, to figure what type the buffer contents should be treated as.
typedef int (*HOOK)(); /* pointer to function returning int */
FIELDTYPE *new_fieldtype(HOOK f_validate, /* field validator */
HOOK c_validate) /* character validator */
int free_fieldtype(FIELDTYPE *ftype); /* type to free */
At least one of the arguments of new_fieldtype() must be
non-NULL. The forms driver will automatically call the new type's validation
functions at appropriate points in processing a field of the new type.
The function free_fieldtype() deallocates the argument
fieldtype, freeing all storage associated with it.
Normally, a field validator is called when the user attempts to leave the field. Its first argument is a field pointer, from which it can get to field buffer 0 and test it. If the function returns TRUE, the operation succeeds; if it returns FALSE, the edit cursor stays in the field.
A character validator gets the character passed in as a first argument. It too should return TRUE if the character is valid, FALSE otherwise.
set_field_type(). If no such arguments are defined for the field
type, this pile pointer argument will be NULL.
In order to arrange for such arguments to be passed to your validation
functions, you must associate a small set of storage-management functions with
the type. The forms driver will use these to synthesize a pile from the trailing
arguments of each set_field_type() argument, and a pointer to the
pile will be passed to the validation functions.
Here is how you make the association:
typedef char *(*PTRHOOK)(); /* pointer to function returning (char *) */
typedef void (*VOIDHOOK)(); /* pointer to function returning void */
int set_fieldtype_arg(FIELDTYPE *type, /* type to alter */
PTRHOOK make_str, /* make structure from args */
PTRHOOK copy_str, /* make copy of structure */
VOIDHOOK free_str); /* free structure storage */
Here is how the storage-management hooks are used:
make_str
set_field_type(). It gets one
argument, a va_list of the type-specific arguments passed to
set_field_type(). It is expected to return a pile pointer to a
data structure that encapsulates those arguments.
copy_str
free_str
make_str and copy_str
functions may return NULL to signal allocation failure. The library routines
will that call them will return error indication when this happens. Thus, your
validation functions should never see a NULL file pointer and need not check
specially for it.
TYPE_ENUM is. For such types, it is possible to define successor
and predecessor functions to support the REQ_NEXT_CHOICE and
REQ_PREV_CHOICE requests. Here's how: typedef int (*INTHOOK)(); /* pointer to function returning int */
int set_fieldtype_arg(FIELDTYPE *type, /* type to alter */
INTHOOK succ, /* get successor value */
INTHOOK pred); /* get predecessor value */
The successor and predecessor arguments will each be passed two arguments;
a field pointer, and a pile pointer (as for the validation functions). They are
expected to use the function field_buffer() to read the current
value, and set_field_buffer() on buffer 0 to set the next or
previous value. Either hook may return TRUE to indicate success (a legal next or
previous value was set) or FALSE to indicate failure.
Use that code as a model, and evolve it towards what you really want. You
will avoid many problems and annoyances that way. The code in the
ncurses library has been specifically exempted from the package
copyright to support this.
If your custom type defines order functions, have do something intuitive with a blank field. A useful convention is to make the successor of a blank field the types minimum value, and its predecessor the maximum.