1 Writing Programs with NCURSES
3 by Eric S. Raymond and Zeyd M. Ben-Halim
4 updates since release 1.9.9e by Thomas Dickey
9 + A Brief History of Curses
10 + Scope of This Document
13 + An Overview of Curses
14 o Compiling Programs using Curses
16 o Standard Windows and Function Naming Conventions
22 o Using Forms Characters
23 o Character Attributes and Color
26 + Function Descriptions
27 o Initialization and Wrapup
28 o Causing Output to the Terminal
29 o Low-Level Capability Access
31 + Hints, Tips, and Tricks
32 o Some Notes of Caution
33 o Temporarily Leaving ncurses Mode
34 o Using ncurses under xterm
35 o Handling Multiple Terminal Screens
36 o Testing for Terminal Capabilities
38 o Special Features of ncurses
39 + Compatibility with Older Versions
40 o Refresh of Overlapping Windows
42 + XSI Curses Conformance
44 + Compiling With the Panels Library
46 + Panels, Input, and the Standard Screen
48 + Miscellaneous Other Facilities
50 + Compiling with the menu Library
55 + Processing Menu Input
56 + Miscellaneous Other Features
58 + Compiling with the forms Library
60 + Creating and Freeing Fields and Forms
61 + Fetching and Changing Field Attributes
62 o Fetching Size and Location Data
63 o Changing the Field Location
64 o The Justification Attribute
65 o Field Display Attributes
69 + Variable-Sized Fields
77 + Direct Field Buffer Manipulation
79 + Control of Form Display
80 + Input Processing in the Forms Driver
81 o Page Navigation Requests
82 o Inter-Field Navigation Requests
83 o Intra-Field Navigation Requests
85 o Field Editing Requests
87 o Application Commands
89 + Field Change Commands
91 + Custom Validation Types
94 o Validation Function Arguments
95 o Order Functions For Custom Types
97 _________________________________________________________________
101 This document is an introduction to programming with curses. It is not
102 an exhaustive reference for the curses Application Programming
103 Interface (API); that role is filled by the curses manual pages.
104 Rather, it is intended to help C programmers ease into using the
107 This document is aimed at C applications programmers not yet
108 specifically familiar with ncurses. If you are already an experienced
109 curses programmer, you should nevertheless read the sections on Mouse
110 Interfacing, Debugging, Compatibility with Older Versions, and Hints,
111 Tips, and Tricks. These will bring you up to speed on the special
112 features and quirks of the ncurses implementation. If you are not so
113 experienced, keep reading.
115 The curses package is a subroutine library for terminal-independent
116 screen-painting and input-event handling which presents a high level
117 screen model to the programmer, hiding differences between terminal
118 types and doing automatic optimization of output to change one screen
119 full of text into another. Curses uses terminfo, which is a database
120 format that can describe the capabilities of thousands of different
123 The curses API may seem something of an archaism on UNIX desktops
124 increasingly dominated by X, Motif, and Tcl/Tk. Nevertheless, UNIX
125 still supports tty lines and X supports xterm(1); the curses API has
126 the advantage of (a) back-portability to character-cell terminals, and
127 (b) simplicity. For an application that does not require bit-mapped
128 graphics and multiple fonts, an interface implementation using curses
129 will typically be a great deal simpler and less expensive than one
132 A Brief History of Curses
134 Historically, the first ancestor of curses was the routines written to
135 provide screen-handling for the game rogue; these used the
136 already-existing termcap database facility for describing terminal
137 capabilities. These routines were abstracted into a documented library
138 and first released with the early BSD UNIX versions.
140 System III UNIX from Bell Labs featured a rewritten and much-improved
141 curses library. It introduced the terminfo format. Terminfo is based
142 on Berkeley's termcap database, but contains a number of improvements
143 and extensions. Parameterized capabilities strings were introduced,
144 making it possible to describe multiple video attributes, and colors
145 and to handle far more unusual terminals than possible with termcap.
146 In the later AT&T System V releases, curses evolved to use more
147 facilities and offer more capabilities, going far beyond BSD curses in
148 power and flexibility.
150 Scope of This Document
152 This document describes ncurses, a free implementation of the System V
153 curses API with some clearly marked extensions. It includes the
154 following System V curses features:
155 * Support for multiple screen highlights (BSD curses could only
156 handle one `standout' highlight, usually reverse-video).
157 * Support for line- and box-drawing using forms characters.
158 * Recognition of function keys on input.
160 * Support for pads (windows of larger than screen size on which the
161 screen or a subwindow defines a viewport).
163 Also, this package makes use of the insert and delete line and
164 character features of terminals so equipped, and determines how to
165 optimally use these features with no help from the programmer. It
166 allows arbitrary combinations of video attributes to be displayed,
167 even on terminals that leave ``magic cookies'' on the screen to mark
168 changes in attributes.
170 The ncurses package can also capture and use event reports from a
171 mouse in some environments (notably, xterm under the X window system).
172 This document includes tips for using the mouse.
174 The ncurses package was originated by Pavel Curtis. The original
175 maintainer of this package is Zeyd Ben-Halim <zmbenhal@netcom.com>.
176 Eric S. Raymond <esr@snark.thyrsus.com> wrote many of the new features
177 in versions after 1.8.1 and wrote most of this introduction. Juergen
178 Pfeifer wrote all of the menu and forms code as well as the Ada95
179 binding. Ongoing work is being done by Thomas Dickey (maintainer).
180 Contact the current maintainers at bug-ncurses@gnu.org.
182 This document also describes the panels extension library, similarly
183 modeled on the SVr4 panels facility. This library allows you to
184 associate backing store with each of a stack or deck of overlapping
185 windows, and provides operations for moving windows around in the
186 stack that change their visibility in the natural way (handling window
189 Finally, this document describes in detail the menus and forms
190 extension libraries, also cloned from System V, which support easy
191 construction and sequences of menus and fill-in forms.
195 In this document, the following terminology is used with reasonable
199 A data structure describing a sub-rectangle of the screen
200 (possibly the entire screen). You can write to a window as
201 though it were a miniature screen, scrolling independently of
202 other windows on the physical screen.
205 A subset of windows which are as large as the terminal screen,
206 i.e., they start at the upper left hand corner and encompass
207 the lower right hand corner. One of these, stdscr, is
208 automatically provided for the programmer.
211 The package's idea of what the terminal display currently looks
212 like, i.e., what the user sees now. This is a special screen.
216 An Overview of Curses
218 Compiling Programs using Curses
220 In order to use the library, it is necessary to have certain types and
221 variables defined. Therefore, the programmer must have a line:
224 at the top of the program source. The screen package uses the Standard
225 I/O library, so <curses.h> includes <stdio.h>. <curses.h> also
226 includes <termios.h>, <termio.h>, or <sgtty.h> depending on your
227 system. It is redundant (but harmless) for the programmer to do these
228 includes, too. In linking with curses you need to have -lncurses in
229 your LDFLAGS or on the command line. There is no need for any other
234 In order to update the screen optimally, it is necessary for the
235 routines to know what the screen currently looks like and what the
236 programmer wants it to look like next. For this purpose, a data type
237 (structure) named WINDOW is defined which describes a window image to
238 the routines, including its starting position on the screen (the (y,
239 x) coordinates of the upper left hand corner) and its size. One of
240 these (called curscr, for current screen) is a screen image of what
241 the terminal currently looks like. Another screen (called stdscr, for
242 standard screen) is provided by default to make changes on.
244 A window is a purely internal representation. It is used to build and
245 store a potential image of a portion of the terminal. It doesn't bear
246 any necessary relation to what is really on the terminal screen; it's
247 more like a scratchpad or write buffer.
249 To make the section of physical screen corresponding to a window
250 reflect the contents of the window structure, the routine refresh()
251 (or wrefresh() if the window is not stdscr) is called.
253 A given physical screen section may be within the scope of any number
254 of overlapping windows. Also, changes can be made to windows in any
255 order, without regard to motion efficiency. Then, at will, the
256 programmer can effectively say ``make it look like this,'' and let the
257 package implementation determine the most efficient way to repaint the
260 Standard Windows and Function Naming Conventions
262 As hinted above, the routines can use several windows, but two are
263 automatically given: curscr, which knows what the terminal looks like,
264 and stdscr, which is what the programmer wants the terminal to look
265 like next. The user should never actually access curscr directly.
266 Changes should be made to through the API, and then the routine
267 refresh() (or wrefresh()) called.
269 Many functions are defined to use stdscr as a default screen. For
270 example, to add a character to stdscr, one calls addch() with the
271 desired character as argument. To write to a different window. use the
272 routine waddch() (for `w'indow-specific addch()) is provided. This
273 convention of prepending function names with a `w' when they are to be
274 applied to specific windows is consistent. The only routines which do
275 not follow it are those for which a window must always be specified.
277 In order to move the current (y, x) coordinates from one point to
278 another, the routines move() and wmove() are provided. However, it is
279 often desirable to first move and then perform some I/O operation. In
280 order to avoid clumsiness, most I/O routines can be preceded by the
281 prefix 'mv' and the desired (y, x) coordinates prepended to the
282 arguments to the function. For example, the calls
294 mvwaddch(win, y, x, ch);
296 Note that the window description pointer (win) comes before the added
297 (y, x) coordinates. If a function requires a window pointer, it is
298 always the first parameter passed.
302 The curses library sets some variables describing the terminal
304 type name description
305 ------------------------------------------------------------------
306 int LINES number of lines on the terminal
307 int COLS number of columns on the terminal
309 The curses.h also introduces some #define constants and types of
313 boolean type, actually a `char' (e.g., bool doneit;)
316 boolean `true' flag (1).
319 boolean `false' flag (0).
322 error flag returned by routines on a failure (-1).
325 error flag returned by routines when things go right.
329 Now we describe how to actually use the screen package. In it, we
330 assume all updating, reading, etc. is applied to stdscr. These
331 instructions will work on any window, providing you change the
332 function names and parameters as mentioned above.
334 Here is a sample program to motivate the discussion:
338 static void finish(int sig);
341 main(int argc, char *argv[])
345 /* initialize your non-curses data structures here */
347 (void) signal(SIGINT, finish); /* arrange interrupts to terminate */
349 (void) initscr(); /* initialize the curses library */
350 keypad(stdscr, TRUE); /* enable keyboard mapping */
351 (void) nonl(); /* tell curses not to do NL->CR/NL on output */
352 (void) cbreak(); /* take input chars one at a time, no wait for \n */
353 (void) echo(); /* echo input - in color */
360 * Simple color assignment, often all we need. Color pair 0 cannot
361 * be redefined. This example uses the same value for the color
362 * pair as for the foreground color, though of course that is not
365 init_pair(1, COLOR_RED, COLOR_BLACK);
366 init_pair(2, COLOR_GREEN, COLOR_BLACK);
367 init_pair(3, COLOR_YELLOW, COLOR_BLACK);
368 init_pair(4, COLOR_BLUE, COLOR_BLACK);
369 init_pair(5, COLOR_CYAN, COLOR_BLACK);
370 init_pair(6, COLOR_MAGENTA, COLOR_BLACK);
371 init_pair(7, COLOR_WHITE, COLOR_BLACK);
376 int c = getch(); /* refresh, accept single keystroke of input */
377 attrset(COLOR_PAIR(num % 8));
380 /* process the command keystroke */
383 finish(0); /* we're done */
386 static void finish(int sig)
390 /* do your non-curses wrapup here */
397 In order to use the screen package, the routines must know about
398 terminal characteristics, and the space for curscr and stdscr must be
399 allocated. These function initscr() does both these things. Since it
400 must allocate space for the windows, it can overflow memory when
401 attempting to do so. On the rare occasions this happens, initscr()
402 will terminate the program with an error message. initscr() must
403 always be called before any of the routines which affect windows are
404 used. If it is not, the program will core dump as soon as either
405 curscr or stdscr are referenced. However, it is usually best to wait
406 to call it until after you are sure you will need it, like after
407 checking for startup errors. Terminal status changing routines like
408 nl() and cbreak() should be called after initscr().
410 Once the screen windows have been allocated, you can set them up for
411 your program. If you want to, say, allow a screen to scroll, use
412 scrollok(). If you want the cursor to be left in place after the last
413 change, use leaveok(). If this isn't done, refresh() will move the
414 cursor to the window's current (y, x) coordinates after updating it.
416 You can create new windows of your own using the functions newwin(),
417 derwin(), and subwin(). The routine delwin() will allow you to get rid
418 of old windows. All the options described above can be applied to any
423 Now that we have set things up, we will want to actually update the
424 terminal. The basic functions used to change what will go on a window
425 are addch() and move(). addch() adds a character at the current (y, x)
426 coordinates. move() changes the current (y, x) coordinates to whatever
427 you want them to be. It returns ERR if you try to move off the window.
428 As mentioned above, you can combine the two into mvaddch() to do both
431 The other output functions, such as addstr() and printw(), all call
432 addch() to add characters to the window.
434 After you have put on the window what you want there, when you want
435 the portion of the terminal covered by the window to be made to look
436 like it, you must call refresh(). In order to optimize finding
437 changes, refresh() assumes that any part of the window not changed
438 since the last refresh() of that window has not been changed on the
439 terminal, i.e., that you have not refreshed a portion of the terminal
440 with an overlapping window. If this is not the case, the routine
441 touchwin() is provided to make it look like the entire window has been
442 changed, thus making refresh() check the whole subsection of the
443 terminal for changes.
445 If you call wrefresh() with curscr as its argument, it will make the
446 screen look like curscr thinks it looks like. This is useful for
447 implementing a command which would redraw the screen in case it get
452 The complementary function to addch() is getch() which, if echo is
453 set, will call addch() to echo the character. Since the screen package
454 needs to know what is on the terminal at all times, if characters are
455 to be echoed, the tty must be in raw or cbreak mode. Since initially
456 the terminal has echoing enabled and is in ordinary ``cooked'' mode,
457 one or the other has to changed before calling getch(); otherwise, the
458 program's output will be unpredictable.
460 When you need to accept line-oriented input in a window, the functions
461 wgetstr() and friends are available. There is even a wscanw() function
462 that can do scanf()(3)-style multi-field parsing on window input.
463 These pseudo-line-oriented functions turn on echoing while they
466 The example code above uses the call keypad(stdscr, TRUE) to enable
467 support for function-key mapping. With this feature, the getch() code
468 watches the input stream for character sequences that correspond to
469 arrow and function keys. These sequences are returned as
470 pseudo-character values. The #define values returned are listed in the
471 curses.h The mapping from sequences to #define values is determined by
472 key_ capabilities in the terminal's terminfo entry.
474 Using Forms Characters
476 The addch() function (and some others, including box() and border())
477 can accept some pseudo-character arguments which are specially defined
478 by ncurses. These are #define values set up in the curses.h header;
479 see there for a complete list (look for the prefix ACS_).
481 The most useful of the ACS defines are the forms-drawing characters.
482 You can use these to draw boxes and simple graphs on the screen. If
483 the terminal does not have such characters, curses.h will map them to
484 a recognizable (though ugly) set of ASCII defaults.
486 Character Attributes and Color
488 The ncurses package supports screen highlights including standout,
489 reverse-video, underline, and blink. It also supports color, which is
490 treated as another kind of highlight.
492 Highlights are encoded, internally, as high bits of the
493 pseudo-character type (chtype) that curses.h uses to represent the
494 contents of a screen cell. See the curses.h header file for a complete
495 list of highlight mask values (look for the prefix A_).
497 There are two ways to make highlights. One is to logical-or the value
498 of the highlights you want into the character argument of an addch()
499 call, or any other output call that takes a chtype argument.
501 The other is to set the current-highlight value. This is logical-or'ed
502 with any highlight you specify the first way. You do this with the
503 functions attron(), attroff(), and attrset(); see the manual pages for
504 details. Color is a special kind of highlight. The package actually
505 thinks in terms of color pairs, combinations of foreground and
506 background colors. The sample code above sets up eight color pairs,
507 all of the guaranteed-available colors on black. Note that each color
508 pair is, in effect, given the name of its foreground color. Any other
509 range of eight non-conflicting values could have been used as the
510 first arguments of the init_pair() values.
512 Once you've done an init_pair() that creates color-pair N, you can use
513 COLOR_PAIR(N) as a highlight that invokes that particular color
514 combination. Note that COLOR_PAIR(N), for constant N, is itself a
515 compile-time constant and can be used in initializers.
519 The ncurses library also provides a mouse interface.
521 NOTE: this facility is specific to ncurses, it is not part of
522 either the XSI Curses standard, nor of System V Release 4, nor BSD
523 curses. System V Release 4 curses contains code with similar
524 interface definitions, however it is not documented. Other than by
525 disassembling the library, we have no way to determine exactly how
526 that mouse code works. Thus, we recommend that you wrap
527 mouse-related code in an #ifdef using the feature macro
528 NCURSES_MOUSE_VERSION so it will not be compiled and linked on
531 Presently, mouse event reporting works in the following environments:
532 * xterm and similar programs such as rxvt.
533 * Linux console, when configured with gpm(1), Alessandro Rubini's
535 * FreeBSD sysmouse (console)
538 The mouse interface is very simple. To activate it, you use the
539 function mousemask(), passing it as first argument a bit-mask that
540 specifies what kinds of events you want your program to be able to
541 see. It will return the bit-mask of events that actually become
542 visible, which may differ from the argument if the mouse device is not
543 capable of reporting some of the event types you specify.
545 Once the mouse is active, your application's command loop should watch
546 for a return value of KEY_MOUSE from wgetch(). When you see this, a
547 mouse event report has been queued. To pick it off the queue, use the
548 function getmouse() (you must do this before the next wgetch(),
549 otherwise another mouse event might come in and make the first one
552 Each call to getmouse() fills a structure (the address of which you'll
553 pass it) with mouse event data. The event data includes zero-origin,
554 screen-relative character-cell coordinates of the mouse pointer. It
555 also includes an event mask. Bits in this mask will be set,
556 corresponding to the event type being reported.
558 The mouse structure contains two additional fields which may be
559 significant in the future as ncurses interfaces to new kinds of
560 pointing device. In addition to x and y coordinates, there is a slot
561 for a z coordinate; this might be useful with touch-screens that can
562 return a pressure or duration parameter. There is also a device ID
563 field, which could be used to distinguish between multiple pointing
566 The class of visible events may be changed at any time via
567 mousemask(). Events that can be reported include presses, releases,
568 single-, double- and triple-clicks (you can set the maximum
569 button-down time for clicks). If you don't make clicks visible, they
570 will be reported as press-release pairs. In some environments, the
571 event mask may include bits reporting the state of shift, alt, and
572 ctrl keys on the keyboard during the event.
574 A function to check whether a mouse event fell within a given window
575 is also supplied. You can use this to see whether a given window
576 should consider a mouse event relevant to it.
578 Because mouse event reporting will not be available in all
579 environments, it would be unwise to build ncurses applications that
580 require the use of a mouse. Rather, you should use the mouse as a
581 shortcut for point-and-shoot commands your application would normally
582 accept from the keyboard. Two of the test games in the ncurses
583 distribution (bs and knight) contain code that illustrates how this
586 See the manual page curs_mouse(3X) for full details of the
587 mouse-interface functions.
591 In order to clean up after the ncurses routines, the routine endwin()
592 is provided. It restores tty modes to what they were when initscr()
593 was first called, and moves the cursor down to the lower-left corner.
594 Thus, anytime after the call to initscr, endwin() should be called
597 Function Descriptions
599 We describe the detailed behavior of some important curses functions
600 here, as a supplement to the manual page descriptions.
602 Initialization and Wrapup
605 The first function called should almost always be initscr().
606 This will determine the terminal type and initialize curses
607 data structures. initscr() also arranges that the first call to
608 refresh() will clear the screen. If an error occurs a message
609 is written to standard error and the program exits. Otherwise
610 it returns a pointer to stdscr. A few functions may be called
611 before initscr (slk_init(), filter(), ripoffline(), use_env(),
612 and, if you are using multiple terminals, newterm().)
615 Your program should always call endwin() before exiting or
616 shelling out of the program. This function will restore tty
617 modes, move the cursor to the lower left corner of the screen,
618 reset the terminal into the proper non-visual mode. Calling
619 refresh() or doupdate() after a temporary escape from the
620 program will restore the ncurses screen from before the escape.
622 newterm(type, ofp, ifp)
623 A program which outputs to more than one terminal should use
624 newterm() instead of initscr(). newterm() should be called once
625 for each terminal. It returns a variable of type SCREEN * which
626 should be saved as a reference to that terminal. (NOTE: a
627 SCREEN variable is not a screen in the sense we are describing
628 in this introduction, but a collection of parameters used to
629 assist in optimizing the display.) The arguments are the type
630 of the terminal (a string) and FILE pointers for the output and
631 input of the terminal. If type is NULL then the environment
632 variable $TERM is used. endwin() should called once at wrapup
633 time for each terminal opened using this function.
636 This function is used to switch to a different terminal
637 previously opened by newterm(). The screen reference for the
638 new terminal is passed as the parameter. The previous terminal
639 is returned by the function. All other calls affect only the
643 The inverse of newterm(); deallocates the data structures
644 associated with a given SCREEN reference.
646 Causing Output to the Terminal
648 refresh() and wrefresh(win)
649 These functions must be called to actually get any output on
650 the terminal, as other routines merely manipulate data
651 structures. wrefresh() copies the named window to the physical
652 terminal screen, taking into account what is already there in
653 order to do optimizations. refresh() does a refresh of stdscr.
654 Unless leaveok() has been enabled, the physical cursor of the
655 terminal is left at the location of the window's cursor.
657 doupdate() and wnoutrefresh(win)
658 These two functions allow multiple updates with more efficiency
659 than wrefresh. To use them, it is important to understand how
660 curses works. In addition to all the window structures, curses
661 keeps two data structures representing the terminal screen: a
662 physical screen, describing what is actually on the screen, and
663 a virtual screen, describing what the programmer wants to have
664 on the screen. wrefresh works by first copying the named window
665 to the virtual screen (wnoutrefresh()), and then calling the
666 routine to update the screen (doupdate()). If the programmer
667 wishes to output several windows at once, a series of calls to
668 wrefresh will result in alternating calls to wnoutrefresh() and
669 doupdate(), causing several bursts of output to the screen. By
670 calling wnoutrefresh() for each window, it is then possible to
671 call doupdate() once, resulting in only one burst of output,
672 with fewer total characters transmitted (this also avoids a
673 visually annoying flicker at each update).
675 Low-Level Capability Access
677 setupterm(term, filenum, errret)
678 This routine is called to initialize a terminal's description,
679 without setting up the curses screen structures or changing the
680 tty-driver mode bits. term is the character string representing
681 the name of the terminal being used. filenum is the UNIX file
682 descriptor of the terminal to be used for output. errret is a
683 pointer to an integer, in which a success or failure indication
684 is returned. The values returned can be 1 (all is well), 0 (no
685 such terminal), or -1 (some problem locating the terminfo
688 The value of term can be given as NULL, which will cause the
689 value of TERM in the environment to be used. The errret pointer
690 can also be given as NULL, meaning no error code is wanted. If
691 errret is defaulted, and something goes wrong, setupterm() will
692 print an appropriate error message and exit, rather than
693 returning. Thus, a simple program can call setupterm(0, 1, 0)
694 and not worry about initialization errors.
696 After the call to setupterm(), the global variable cur_term is
697 set to point to the current structure of terminal capabilities.
698 By calling setupterm() for each terminal, and saving and
699 restoring cur_term, it is possible for a program to use two or
700 more terminals at once. Setupterm() also stores the names
701 section of the terminal description in the global character
702 array ttytype[]. Subsequent calls to setupterm() will overwrite
703 this array, so you'll have to save it yourself if need be.
707 NOTE: These functions are not part of the standard curses API!
710 This function can be used to explicitly set a trace level. If
711 the trace level is nonzero, execution of your program will
712 generate a file called `trace' in the current working directory
713 containing a report on the library's actions. Higher trace
714 levels enable more detailed (and verbose) reporting -- see
715 comments attached to TRACE_ defines in the curses.h file for
716 details. (It is also possible to set a trace level by assigning
717 a trace level value to the environment variable NCURSES_TRACE).
720 This function can be used to output your own debugging
721 information. It is only available only if you link with
722 -lncurses_g. It can be used the same way as printf(), only it
723 outputs a newline after the end of arguments. The output goes
724 to a file called trace in the current directory.
726 Trace logs can be difficult to interpret due to the sheer volume of
727 data dumped in them. There is a script called tracemunch included with
728 the ncurses distribution that can alleviate this problem somewhat; it
729 compacts long sequences of similar operations into more succinct
730 single-line pseudo-operations. These pseudo-ops can be distinguished
731 by the fact that they are named in capital letters.
733 Hints, Tips, and Tricks
735 The ncurses manual pages are a complete reference for this library. In
736 the remainder of this document, we discuss various useful methods that
737 may not be obvious from the manual page descriptions.
739 Some Notes of Caution
741 If you find yourself thinking you need to use noraw() or nocbreak(),
742 think again and move carefully. It's probably better design to use
743 getstr() or one of its relatives to simulate cooked mode. The noraw()
744 and nocbreak() functions try to restore cooked mode, but they may end
745 up clobbering some control bits set before you started your
746 application. Also, they have always been poorly documented, and are
747 likely to hurt your application's usability with other curses
750 Bear in mind that refresh() is a synonym for wrefresh(stdscr). Don't
751 try to mix use of stdscr with use of windows declared by newwin(); a
752 refresh() call will blow them off the screen. The right way to handle
753 this is to use subwin(), or not touch stdscr at all and tile your
754 screen with declared windows which you then wnoutrefresh() somewhere
755 in your program event loop, with a single doupdate() call to trigger
758 You are much less likely to run into problems if you design your
759 screen layouts to use tiled rather than overlapping windows.
760 Historically, curses support for overlapping windows has been weak,
761 fragile, and poorly documented. The ncurses library is not yet an
762 exception to this rule.
764 There is a panels library included in the ncurses distribution that
765 does a pretty good job of strengthening the overlapping-windows
768 Try to avoid using the global variables LINES and COLS. Use getmaxyx()
769 on the stdscr context instead. Reason: your code may be ported to run
770 in an environment with window resizes, in which case several screens
771 could be open with different sizes.
773 Temporarily Leaving NCURSES Mode
775 Sometimes you will want to write a program that spends most of its
776 time in screen mode, but occasionally returns to ordinary `cooked'
777 mode. A common reason for this is to support shell-out. This behavior
778 is simple to arrange in ncurses.
780 To leave ncurses mode, call endwin() as you would if you were
781 intending to terminate the program. This will take the screen back to
782 cooked mode; you can do your shell-out. When you want to return to
783 ncurses mode, simply call refresh() or doupdate(). This will repaint
786 There is a boolean function, isendwin(), which code can use to test
787 whether ncurses screen mode is active. It returns TRUE in the interval
788 between an endwin() call and the following refresh(), FALSE otherwise.
790 Here is some sample code for shellout:
791 addstr("Shelling out...");
792 def_prog_mode(); /* save current tty modes */
793 endwin(); /* restore original tty modes */
794 system("sh"); /* run shell */
795 addstr("returned.\n"); /* prepare return message */
796 refresh(); /* restore save modes, repaint screen */
798 Using NCURSES under XTERM
800 A resize operation in X sends SIGWINCH to the application running
801 under xterm. The easiest way to handle SIGWINCH is to do an endwin,
802 followed by an refresh and a screen repaint you code yourself. The
803 refresh will pick up the new screen size from the xterm's environment.
805 That is the standard way, of course (it even works with some vendor's
806 curses implementations). Its drawback is that it clears the screen to
807 reinitialize the display, and does not resize subwindows which must be
808 shrunk. Ncurses provides an extension which works better, the
809 resizeterm function. That function ensures that all windows are
810 limited to the new screen dimensions, and pads stdscr with blanks if
811 the screen is larger.
813 The ncurses library provides a SIGWINCH signal handler, which pushes a
814 KEY_RESIZE via the wgetch() calls. When ncurses returns that code, it
815 calls resizeterm to update the size of the standard screen's window,
816 repainting that (filling with blanks or truncating as needed). It also
817 resizes other windows, but its effect may be less satisfactory because
818 it cannot know how you want the screen re-painted. You will usually
819 have to write special-purpose code to handle KEY_RESIZE yourself.
821 Handling Multiple Terminal Screens
823 The initscr() function actually calls a function named newterm() to do
824 most of its work. If you are writing a program that opens multiple
825 terminals, use newterm() directly.
827 For each call, you will have to specify a terminal type and a pair of
828 file pointers; each call will return a screen reference, and stdscr
829 will be set to the last one allocated. You will switch between screens
830 with the set_term call. Note that you will also have to call
831 def_shell_mode and def_prog_mode on each tty yourself.
833 Testing for Terminal Capabilities
835 Sometimes you may want to write programs that test for the presence of
836 various capabilities before deciding whether to go into ncurses mode.
837 An easy way to do this is to call setupterm(), then use the functions
838 tigetflag(), tigetnum(), and tigetstr() to do your testing.
840 A particularly useful case of this often comes up when you want to
841 test whether a given terminal type should be treated as `smart'
842 (cursor-addressable) or `stupid'. The right way to test this is to see
843 if the return value of tigetstr("cup") is non-NULL. Alternatively, you
844 can include the term.h file and test the value of the macro
849 Use the addchstr() family of functions for fast screen-painting of
850 text when you know the text doesn't contain any control characters.
851 Try to make attribute changes infrequent on your screens. Don't use
852 the immedok() option!
854 Special Features of NCURSES
856 The wresize() function allows you to resize a window in place. The
857 associated resizeterm() function simplifies the construction of
858 SIGWINCH handlers, for resizing all windows.
860 The define_key() function allows you to define at runtime function-key
861 control sequences which are not in the terminal description. The
862 keyok() function allows you to temporarily enable or disable
863 interpretation of any function-key control sequence.
865 The use_default_colors() function allows you to construct applications
866 which can use the terminal's default foreground and background colors
867 as an additional "default" color. Several terminal emulators support
868 this feature, which is based on ISO 6429.
870 Ncurses supports up 16 colors, unlike SVr4 curses which defines only
871 8. While most terminals which provide color allow only 8 colors, about
872 a quarter (including XFree86 xterm) support 16 colors.
874 Compatibility with Older Versions
876 Despite our best efforts, there are some differences between ncurses
877 and the (undocumented!) behavior of older curses implementations.
878 These arise from ambiguities or omissions in the documentation of the
881 Refresh of Overlapping Windows
883 If you define two windows A and B that overlap, and then alternately
884 scribble on and refresh them, the changes made to the overlapping
885 region under historic curses versions were often not documented
888 To understand why this is a problem, remember that screen updates are
889 calculated between two representations of the entire display. The
890 documentation says that when you refresh a window, it is first copied
891 to the virtual screen, and then changes are calculated to update the
892 physical screen (and applied to the terminal). But "copied to" is not
893 very specific, and subtle differences in how copying works can produce
894 different behaviors in the case where two overlapping windows are each
895 being refreshed at unpredictable intervals.
897 What happens to the overlapping region depends on what wnoutrefresh()
898 does with its argument -- what portions of the argument window it
899 copies to the virtual screen. Some implementations do "change copy",
900 copying down only locations in the window that have changed (or been
901 marked changed with wtouchln() and friends). Some implementations do
902 "entire copy", copying all window locations to the virtual screen
903 whether or not they have changed.
905 The ncurses library itself has not always been consistent on this
906 score. Due to a bug, versions 1.8.7 to 1.9.8a did entire copy.
907 Versions 1.8.6 and older, and versions 1.9.9 and newer, do change
910 For most commercial curses implementations, it is not documented and
911 not known for sure (at least not to the ncurses maintainers) whether
912 they do change copy or entire copy. We know that System V release 3
913 curses has logic in it that looks like an attempt to do change copy,
914 but the surrounding logic and data representations are sufficiently
915 complex, and our knowledge sufficiently indirect, that it's hard to
916 know whether this is reliable. It is not clear what the SVr4
917 documentation and XSI standard intend. The XSI Curses standard barely
918 mentions wnoutrefresh(); the SVr4 documents seem to be describing
919 entire-copy, but it is possible with some effort and straining to read
922 It might therefore be unwise to rely on either behavior in programs
923 that might have to be linked with other curses implementations.
924 Instead, you can do an explicit touchwin() before the wnoutrefresh()
925 call to guarantee an entire-contents copy anywhere.
927 The really clean way to handle this is to use the panels library. If,
928 when you want a screen update, you do update_panels(), it will do all
929 the necessary wnoutrefresh() calls for whatever panel stacking order
930 you have defined. Then you can do one doupdate() and there will be a
931 single burst of physical I/O that will do all your updates.
935 If you have been using a very old versions of ncurses (1.8.7 or older)
936 you may be surprised by the behavior of the erase functions. In older
937 versions, erased areas of a window were filled with a blank modified
938 by the window's current attribute (as set by wattrset(), wattron(),
939 wattroff() and friends).
941 In newer versions, this is not so. Instead, the attribute of erased
942 blanks is normal unless and until it is modified by the functions
943 bkgdset() or wbkgdset().
945 This change in behavior conforms ncurses to System V Release 4 and the
948 XSI Curses Conformance
950 The ncurses library is intended to be base-level conformant with the
951 XSI Curses standard from X/Open. Many extended-level features (in
952 fact, almost all features not directly concerned with wide characters
953 and internationalization) are also supported.
955 One effect of XSI conformance is the change in behavior described
956 under "Background Erase -- Compatibility with Old Versions".
958 Also, ncurses meets the XSI requirement that every macro entry point
959 have a corresponding function which may be linked (and will be
960 prototype-checked) if the macro definition is disabled with #undef.
964 The ncurses library by itself provides good support for screen
965 displays in which the windows are tiled (non-overlapping). In the more
966 general case that windows may overlap, you have to use a series of
967 wnoutrefresh() calls followed by a doupdate(), and be careful about
968 the order you do the window refreshes in. It has to be bottom-upwards,
969 otherwise parts of windows that should be obscured will show through.
971 When your interface design is such that windows may dive deeper into
972 the visibility stack or pop to the top at runtime, the resulting
973 book-keeping can be tedious and difficult to get right. Hence the
976 The panel library first appeared in AT&T System V. The version
977 documented here is the panel code distributed with ncurses.
979 Compiling With the Panels Library
981 Your panels-using modules must import the panels library declarations
985 and must be linked explicitly with the panels library using an -lpanel
986 argument. Note that they must also link the ncurses library with
987 -lncurses. Many linkers are two-pass and will accept either order, but
988 it is still good practice to put -lpanel first and -lncurses second.
992 A panel object is a window that is implicitly treated as part of a
993 deck including all other panel objects. The deck has an implicit
994 bottom-to-top visibility order. The panels library includes an update
995 function (analogous to refresh()) that displays all panels in the deck
996 in the proper order to resolve overlaps. The standard window, stdscr,
997 is considered below all panels.
999 Details on the panels functions are available in the man pages. We'll
1000 just hit the highlights here.
1002 You create a panel from a window by calling new_panel() on a window
1003 pointer. It then becomes the top of the deck. The panel's window is
1004 available as the value of panel_window() called with the panel pointer
1007 You can delete a panel (removing it from the deck) with del_panel.
1008 This will not deallocate the associated window; you have to do that
1009 yourself. You can replace a panel's window with a different window by
1010 calling replace_window. The new window may be of different size; the
1011 panel code will re-compute all overlaps. This operation doesn't change
1012 the panel's position in the deck.
1014 To move a panel's window, use move_panel(). The mvwin() function on
1015 the panel's window isn't sufficient because it doesn't update the
1016 panels library's representation of where the windows are. This
1017 operation leaves the panel's depth, contents, and size unchanged.
1019 Two functions (top_panel(), bottom_panel()) are provided for
1020 rearranging the deck. The first pops its argument window to the top of
1021 the deck; the second sends it to the bottom. Either operation leaves
1022 the panel's screen location, contents, and size unchanged.
1024 The function update_panels() does all the wnoutrefresh() calls needed
1025 to prepare for doupdate() (which you must call yourself, afterwards).
1027 Typically, you will want to call update_panels() and doupdate() just
1028 before accepting command input, once in each cycle of interaction with
1029 the user. If you call update_panels() after each and every panel
1030 write, you'll generate a lot of unnecessary refresh activity and
1033 Panels, Input, and the Standard Screen
1035 You shouldn't mix wnoutrefresh() or wrefresh() operations with panels
1036 code; this will work only if the argument window is either in the top
1037 panel or unobscured by any other panels.
1039 The stsdcr window is a special case. It is considered below all
1040 panels. Because changes to panels may obscure parts of stdscr, though,
1041 you should call update_panels() before doupdate() even when you only
1044 Note that wgetch automatically calls wrefresh. Therefore, before
1045 requesting input from a panel window, you need to be sure that the
1046 panel is totally unobscured.
1048 There is presently no way to display changes to one obscured panel
1049 without repainting all panels.
1053 It's possible to remove a panel from the deck temporarily; use
1054 hide_panel for this. Use show_panel() to render it visible again. The
1055 predicate function panel_hidden tests whether or not a panel is
1058 The panel_update code ignores hidden panels. You cannot do top_panel()
1059 or bottom_panel on a hidden panel(). Other panels operations are
1062 Miscellaneous Other Facilities
1064 It's possible to navigate the deck using the functions panel_above()
1065 and panel_below. Handed a panel pointer, they return the panel above
1066 or below that panel. Handed NULL, they return the bottom-most or
1069 Every panel has an associated user pointer, not used by the panel
1070 code, to which you can attach application data. See the man page
1071 documentation of set_panel_userptr() and panel_userptr for details.
1075 A menu is a screen display that assists the user to choose some subset
1076 of a given set of items. The menu library is a curses extension that
1077 supports easy programming of menu hierarchies with a uniform but
1080 The menu library first appeared in AT&T System V. The version
1081 documented here is the menu code distributed with ncurses.
1083 Compiling With the menu Library
1085 Your menu-using modules must import the menu library declarations with
1088 and must be linked explicitly with the menus library using an -lmenu
1089 argument. Note that they must also link the ncurses library with
1090 -lncurses. Many linkers are two-pass and will accept either order, but
1091 it is still good practice to put -lmenu first and -lncurses second.
1095 The menus created by this library consist of collections of items
1096 including a name string part and a description string part. To make
1097 menus, you create groups of these items and connect them with menu
1100 The menu can then by posted, that is written to an associated window.
1101 Actually, each menu has two associated windows; a containing window in
1102 which the programmer can scribble titles or borders, and a subwindow
1103 in which the menu items proper are displayed. If this subwindow is too
1104 small to display all the items, it will be a scrollable viewport on
1105 the collection of items.
1107 A menu may also be unposted (that is, undisplayed), and finally freed
1108 to make the storage associated with it and its items available for
1111 The general flow of control of a menu program looks like this:
1112 1. Initialize curses.
1113 2. Create the menu items, using new_item().
1114 3. Create the menu using new_menu().
1115 4. Post the menu using post_menu().
1116 5. Refresh the screen.
1117 6. Process user requests via an input loop.
1118 7. Unpost the menu using unpost_menu().
1119 8. Free the menu, using free_menu().
1120 9. Free the items using free_item().
1121 10. Terminate curses.
1125 Menus may be multi-valued or (the default) single-valued (see the
1126 manual page menu_opts(3x) to see how to change the default). Both
1127 types always have a current item.
1129 From a single-valued menu you can read the selected value simply by
1130 looking at the current item. From a multi-valued menu, you get the
1131 selected set by looping through the items applying the item_value()
1132 predicate function. Your menu-processing code can use the function
1133 set_item_value() to flag the items in the select set.
1135 Menu items can be made unselectable using set_item_opts() or
1136 item_opts_off() with the O_SELECTABLE argument. This is the only
1137 option so far defined for menus, but it is good practice to code as
1138 though other option bits might be on.
1142 The menu library calculates a minimum display size for your window,
1143 based on the following variables:
1144 * The number and maximum length of the menu items
1145 * Whether the O_ROWMAJOR option is enabled
1146 * Whether display of descriptions is enabled
1147 * Whatever menu format may have been set by the programmer
1148 * The length of the menu mark string used for highlighting selected
1151 The function set_menu_format() allows you to set the maximum size of
1152 the viewport or menu page that will be used to display menu items. You
1153 can retrieve any format associated with a menu with menu_format(). The
1154 default format is rows=16, columns=1.
1156 The actual menu page may be smaller than the format size. This depends
1157 on the item number and size and whether O_ROWMAJOR is on. This option
1158 (on by default) causes menu items to be displayed in a `raster-scan'
1159 pattern, so that if more than one item will fit horizontally the first
1160 couple of items are side-by-side in the top row. The alternative is
1161 column-major display, which tries to put the first several items in
1164 As mentioned above, a menu format not large enough to allow all items
1165 to fit on-screen will result in a menu display that is vertically
1168 You can scroll it with requests to the menu driver, which will be
1169 described in the section on menu input handling.
1171 Each menu has a mark string used to visually tag selected items; see
1172 the menu_mark(3x) manual page for details. The mark string length also
1173 influences the menu page size.
1175 The function scale_menu() returns the minimum display size that the
1176 menu code computes from all these factors. There are other menu
1177 display attributes including a select attribute, an attribute for
1178 selectable items, an attribute for unselectable items, and a pad
1179 character used to separate item name text from description text. These
1180 have reasonable defaults which the library allows you to change (see
1181 the menu_attribs(3x) manual page.
1185 Each menu has, as mentioned previously, a pair of associated windows.
1186 Both these windows are painted when the menu is posted and erased when
1187 the menu is unposted.
1189 The outer or frame window is not otherwise touched by the menu
1190 routines. It exists so the programmer can associate a title, a border,
1191 or perhaps help text with the menu and have it properly refreshed or
1192 erased at post/unpost time. The inner window or subwindow is where the
1193 current menu page is displayed.
1195 By default, both windows are stdscr. You can set them with the
1196 functions in menu_win(3x).
1198 When you call post_menu(), you write the menu to its subwindow. When
1199 you call unpost_menu(), you erase the subwindow, However, neither of
1200 these actually modifies the screen. To do that, call wrefresh() or
1203 Processing Menu Input
1205 The main loop of your menu-processing code should call menu_driver()
1206 repeatedly. The first argument of this routine is a menu pointer; the
1207 second is a menu command code. You should write an input-fetching
1208 routine that maps input characters to menu command codes, and pass its
1209 output to menu_driver(). The menu command codes are fully documented
1212 The simplest group of command codes is REQ_NEXT_ITEM, REQ_PREV_ITEM,
1213 REQ_FIRST_ITEM, REQ_LAST_ITEM, REQ_UP_ITEM, REQ_DOWN_ITEM,
1214 REQ_LEFT_ITEM, REQ_RIGHT_ITEM. These change the currently selected
1215 item. These requests may cause scrolling of the menu page if it only
1216 partially displayed.
1218 There are explicit requests for scrolling which also change the
1219 current item (because the select location does not change, but the
1220 item there does). These are REQ_SCR_DLINE, REQ_SCR_ULINE,
1221 REQ_SCR_DPAGE, and REQ_SCR_UPAGE.
1223 The REQ_TOGGLE_ITEM selects or deselects the current item. It is for
1224 use in multi-valued menus; if you use it with O_ONEVALUE on, you'll
1225 get an error return (E_REQUEST_DENIED).
1227 Each menu has an associated pattern buffer. The menu_driver() logic
1228 tries to accumulate printable ASCII characters passed in in that
1229 buffer; when it matches a prefix of an item name, that item (or the
1230 next matching item) is selected. If appending a character yields no
1231 new match, that character is deleted from the pattern buffer, and
1232 menu_driver() returns E_NO_MATCH.
1234 Some requests change the pattern buffer directly: REQ_CLEAR_PATTERN,
1235 REQ_BACK_PATTERN, REQ_NEXT_MATCH, REQ_PREV_MATCH. The latter two are
1236 useful when pattern buffer input matches more than one item in a
1239 Each successful scroll or item navigation request clears the pattern
1240 buffer. It is also possible to set the pattern buffer explicitly with
1243 Finally, menu driver requests above the constant MAX_COMMAND are
1244 considered application-specific commands. The menu_driver() code
1245 ignores them and returns E_UNKNOWN_COMMAND.
1247 Miscellaneous Other Features
1249 Various menu options can affect the processing and visual appearance
1250 and input processing of menus. See menu_opts(3x) for details.
1252 It is possible to change the current item from application code; this
1253 is useful if you want to write your own navigation requests. It is
1254 also possible to explicitly set the top row of the menu display. See
1255 mitem_current(3x). If your application needs to change the menu
1256 subwindow cursor for any reason, pos_menu_cursor() will restore it to
1257 the correct location for continuing menu driver processing.
1259 It is possible to set hooks to be called at menu initialization and
1260 wrapup time, and whenever the selected item changes. See
1263 Each item, and each menu, has an associated user pointer on which you
1264 can hang application data. See mitem_userptr(3x) and menu_userptr(3x).
1268 The form library is a curses extension that supports easy programming
1269 of on-screen forms for data entry and program control.
1271 The form library first appeared in AT&T System V. The version
1272 documented here is the form code distributed with ncurses.
1274 Compiling With the form Library
1276 Your form-using modules must import the form library declarations with
1279 and must be linked explicitly with the forms library using an -lform
1280 argument. Note that they must also link the ncurses library with
1281 -lncurses. Many linkers are two-pass and will accept either order, but
1282 it is still good practice to put -lform first and -lncurses second.
1286 A form is a collection of fields; each field may be either a label
1287 (explanatory text) or a data-entry location. Long forms may be
1288 segmented into pages; each entry to a new page clears the screen.
1290 To make forms, you create groups of fields and connect them with form
1291 frame objects; the form library makes this relatively simple.
1293 Once defined, a form can be posted, that is written to an associated
1294 window. Actually, each form has two associated windows; a containing
1295 window in which the programmer can scribble titles or borders, and a
1296 subwindow in which the form fields proper are displayed.
1298 As the form user fills out the posted form, navigation and editing
1299 keys support movement between fields, editing keys support modifying
1300 field, and plain text adds to or changes data in a current field. The
1301 form library allows you (the forms designer) to bind each navigation
1302 and editing key to any keystroke accepted by curses Fields may have
1303 validation conditions on them, so that they check input data for type
1304 and value. The form library supplies a rich set of pre-defined field
1305 types, and makes it relatively easy to define new ones.
1307 Once its transaction is completed (or aborted), a form may be unposted
1308 (that is, undisplayed), and finally freed to make the storage
1309 associated with it and its items available for re-use.
1311 The general flow of control of a form program looks like this:
1312 1. Initialize curses.
1313 2. Create the form fields, using new_field().
1314 3. Create the form using new_form().
1315 4. Post the form using post_form().
1316 5. Refresh the screen.
1317 6. Process user requests via an input loop.
1318 7. Unpost the form using unpost_form().
1319 8. Free the form, using free_form().
1320 9. Free the fields using free_field().
1321 10. Terminate curses.
1323 Note that this looks much like a menu program; the form library
1324 handles tasks which are in many ways similar, and its interface was
1325 obviously designed to resemble that of the menu library wherever
1328 In forms programs, however, the `process user requests' is somewhat
1329 more complicated than for menus. Besides menu-like navigation
1330 operations, the menu driver loop has to support field editing and data
1333 Creating and Freeing Fields and Forms
1335 The basic function for creating fields is new_field():
1336 FIELD *new_field(int height, int width, /* new field size */
1337 int top, int left, /* upper left corner */
1338 int offscreen, /* number of offscreen rows */
1339 int nbuf); /* number of working buffers */
1341 Menu items always occupy a single row, but forms fields may have
1342 multiple rows. So new_field() requires you to specify a width and
1343 height (the first two arguments, which mist both be greater than
1346 You must also specify the location of the field's upper left corner on
1347 the screen (the third and fourth arguments, which must be zero or
1348 greater). Note that these coordinates are relative to the form
1349 subwindow, which will coincide with stdscr by default but need not be
1350 stdscr if you've done an explicit set_form_win() call.
1352 The fifth argument allows you to specify a number of off-screen rows.
1353 If this is zero, the entire field will always be displayed. If it is
1354 nonzero, the form will be scrollable, with only one screen-full
1355 (initially the top part) displayed at any given time. If you make a
1356 field dynamic and grow it so it will no longer fit on the screen, the
1357 form will become scrollable even if the offscreen argument was
1360 The forms library allocates one working buffer per field; the size of
1361 each buffer is ((height + offscreen)*width + 1, one character for each
1362 position in the field plus a NUL terminator. The sixth argument is the
1363 number of additional data buffers to allocate for the field; your
1364 application can use them for its own purposes.
1365 FIELD *dup_field(FIELD *field, /* field to copy */
1366 int top, int left); /* location of new copy */
1368 The function dup_field() duplicates an existing field at a new
1369 location. Size and buffering information are copied; some attribute
1370 flags and status bits are not (see the form_field_new(3X) for
1372 FIELD *link_field(FIELD *field, /* field to copy */
1373 int top, int left); /* location of new copy */
1375 The function link_field() also duplicates an existing field at a new
1376 location. The difference from dup_field() is that it arranges for the
1377 new field's buffer to be shared with the old one.
1379 Besides the obvious use in making a field editable from two different
1380 form pages, linked fields give you a way to hack in dynamic labels. If
1381 you declare several fields linked to an original, and then make them
1382 inactive, changes from the original will still be propagated to the
1385 As with duplicated fields, linked fields have attribute bits separate
1388 As you might guess, all these field-allocations return NULL if the
1389 field allocation is not possible due to an out-of-memory error or
1390 out-of-bounds arguments.
1392 To connect fields to a form, use
1393 FORM *new_form(FIELD **fields);
1395 This function expects to see a NULL-terminated array of field
1396 pointers. Said fields are connected to a newly-allocated form object;
1397 its address is returned (or else NULL if the allocation fails).
1399 Note that new_field() does not copy the pointer array into private
1400 storage; if you modify the contents of the pointer array during forms
1401 processing, all manner of bizarre things might happen. Also note that
1402 any given field may only be connected to one form.
1404 The functions free_field() and free_form are available to free field
1405 and form objects. It is an error to attempt to free a field connected
1406 to a form, but not vice-versa; thus, you will generally free your form
1409 Fetching and Changing Field Attributes
1411 Each form field has a number of location and size attributes
1412 associated with it. There are other field attributes used to control
1413 display and editing of the field. Some (for example, the O_STATIC bit)
1414 involve sufficient complications to be covered in sections of their
1415 own later on. We cover the functions used to get and set several basic
1418 When a field is created, the attributes not specified by the new_field
1419 function are copied from an invisible system default field. In
1420 attribute-setting and -fetching functions, the argument NULL is taken
1421 to mean this field. Changes to it persist as defaults until your forms
1422 application terminates.
1424 Fetching Size and Location Data
1426 You can retrieve field sizes and locations through:
1427 int field_info(FIELD *field, /* field from which to fetch */
1428 int *height, *int width, /* field size */
1429 int *top, int *left, /* upper left corner */
1430 int *offscreen, /* number of offscreen rows */
1431 int *nbuf); /* number of working buffers */
1433 This function is a sort of inverse of new_field(); instead of setting
1434 size and location attributes of a new field, it fetches them from an
1437 Changing the Field Location
1439 It is possible to move a field's location on the screen:
1440 int move_field(FIELD *field, /* field to alter */
1441 int top, int left); /* new upper-left corner */
1443 You can, of course. query the current location through field_info().
1445 The Justification Attribute
1447 One-line fields may be unjustified, justified right, justified left,
1448 or centered. Here is how you manipulate this attribute:
1449 int set_field_just(FIELD *field, /* field to alter */
1450 int justmode); /* mode to set */
1452 int field_just(FIELD *field); /* fetch mode of field */
1454 The mode values accepted and returned by this functions are
1455 preprocessor macros NO_JUSTIFICATION, JUSTIFY_RIGHT, JUSTIFY_LEFT, or
1458 Field Display Attributes
1460 For each field, you can set a foreground attribute for entered
1461 characters, a background attribute for the entire field, and a pad
1462 character for the unfilled portion of the field. You can also control
1463 pagination of the form.
1465 This group of four field attributes controls the visual appearance of
1466 the field on the screen, without affecting in any way the data in the
1468 int set_field_fore(FIELD *field, /* field to alter */
1469 chtype attr); /* attribute to set */
1471 chtype field_fore(FIELD *field); /* field to query */
1473 int set_field_back(FIELD *field, /* field to alter */
1474 chtype attr); /* attribute to set */
1476 chtype field_back(FIELD *field); /* field to query */
1478 int set_field_pad(FIELD *field, /* field to alter */
1479 int pad); /* pad character to set */
1481 chtype field_pad(FIELD *field);
1483 int set_new_page(FIELD *field, /* field to alter */
1484 int flag); /* TRUE to force new page */
1486 chtype new_page(FIELD *field); /* field to query */
1488 The attributes set and returned by the first four functions are normal
1489 curses(3x) display attribute values (A_STANDOUT, A_BOLD, A_REVERSE
1490 etc). The page bit of a field controls whether it is displayed at the
1491 start of a new form screen.
1495 There is also a large collection of field option bits you can set to
1496 control various aspects of forms processing. You can manipulate them
1497 with these functions:
1498 int set_field_opts(FIELD *field, /* field to alter */
1499 int attr); /* attribute to set */
1501 int field_opts_on(FIELD *field, /* field to alter */
1502 int attr); /* attributes to turn on */
1504 int field_opts_off(FIELD *field, /* field to alter */
1505 int attr); /* attributes to turn off */
1507 int field_opts(FIELD *field); /* field to query */
1509 By default, all options are on. Here are the available option bits:
1512 Controls whether the field is visible on the screen. Can be
1513 used during form processing to hide or pop up fields depending
1514 on the value of parent fields.
1517 Controls whether the field is active during forms processing
1518 (i.e. visited by form navigation keys). Can be used to make
1519 labels or derived fields with buffer values alterable by the
1520 forms application, not the user.
1523 Controls whether data is displayed during field entry. If this
1524 option is turned off on a field, the library will accept and
1525 edit data in that field, but it will not be displayed and the
1526 visible field cursor will not move. You can turn off the
1527 O_PUBLIC bit to define password fields.
1530 Controls whether the field's data can be modified. When this
1531 option is off, all editing requests except REQ_PREV_CHOICE and
1532 REQ_NEXT_CHOICE will fail. Such read-only fields may be useful
1536 Controls word-wrapping in multi-line fields. Normally, when any
1537 character of a (blank-separated) word reaches the end of the
1538 current line, the entire word is wrapped to the next line
1539 (assuming there is one). When this option is off, the word will
1540 be split across the line break.
1543 Controls field blanking. When this option is on, entering a
1544 character at the first field position erases the entire field
1545 (except for the just-entered character).
1548 Controls automatic skip to next field when this one fills.
1549 Normally, when the forms user tries to type more data into a
1550 field than will fit, the editing location jumps to next field.
1551 When this option is off, the user's cursor will hang at the end
1552 of the field. This option is ignored in dynamic fields that
1553 have not reached their size limit.
1556 Controls whether validation is applied to blank fields.
1557 Normally, it is not; the user can leave a field blank without
1558 invoking the usual validation check on exit. If this option is
1559 off on a field, exit from it will invoke a validation check.
1562 Controls whether validation occurs on every exit, or only after
1563 the field is modified. Normally the latter is true. Setting
1564 O_PASSOK may be useful if your field's validation function may
1565 change during forms processing.
1568 Controls whether the field is fixed to its initial dimensions.
1569 If you turn this off, the field becomes dynamic and will
1570 stretch to fit entered data.
1572 A field's options cannot be changed while the field is currently
1573 selected. However, options may be changed on posted fields that are
1576 The option values are bit-masks and can be composed with logical-or in
1581 Every field has a status flag, which is set to FALSE when the field is
1582 created and TRUE when the value in field buffer 0 changes. This flag
1583 can be queried and set directly:
1584 int set_field_status(FIELD *field, /* field to alter */
1585 int status); /* mode to set */
1587 int field_status(FIELD *field); /* fetch mode of field */
1589 Setting this flag under program control can be useful if you use the
1590 same form repeatedly, looking for modified fields each time.
1592 Calling field_status() on a field not currently selected for input
1593 will return a correct value. Calling field_status() on a field that is
1594 currently selected for input may not necessarily give a correct field
1595 status value, because entered data isn't necessarily copied to buffer
1596 zero before the exit validation check. To guarantee that the returned
1597 status value reflects reality, call field_status() either (1) in the
1598 field's exit validation check routine, (2) from the field's or form's
1599 initialization or termination hooks, or (3) just after a
1600 REQ_VALIDATION request has been processed by the forms driver.
1604 Each field structure contains one character pointer slot that is not
1605 used by the forms library. It is intended to be used by applications
1606 to store private per-field data. You can manipulate it with:
1607 int set_field_userptr(FIELD *field, /* field to alter */
1608 char *userptr); /* mode to set */
1610 char *field_userptr(FIELD *field); /* fetch mode of field */
1612 (Properly, this user pointer field ought to have (void *) type. The
1613 (char *) type is retained for System V compatibility.)
1615 It is valid to set the user pointer of the default field (with a
1616 set_field_userptr() call passed a NULL field pointer.) When a new
1617 field is created, the default-field user pointer is copied to
1618 initialize the new field's user pointer.
1620 Variable-Sized Fields
1622 Normally, a field is fixed at the size specified for it at creation
1623 time. If, however, you turn off its O_STATIC bit, it becomes dynamic
1624 and will automatically resize itself to accommodate data as it is
1625 entered. If the field has extra buffers associated with it, they will
1626 grow right along with the main input buffer.
1628 A one-line dynamic field will have a fixed height (1) but variable
1629 width, scrolling horizontally to display data within the field area as
1630 originally dimensioned and located. A multi-line dynamic field will
1631 have a fixed width, but variable height (number of rows), scrolling
1632 vertically to display data within the field area as originally
1633 dimensioned and located.
1635 Normally, a dynamic field is allowed to grow without limit. But it is
1636 possible to set an upper limit on the size of a dynamic field. You do
1637 it with this function:
1638 int set_max_field(FIELD *field, /* field to alter (may not be NULL) */
1639 int max_size); /* upper limit on field size */
1641 If the field is one-line, max_size is taken to be a column size limit;
1642 if it is multi-line, it is taken to be a line size limit. To disable
1643 any limit, use an argument of zero. The growth limit can be changed
1644 whether or not the O_STATIC bit is on, but has no effect until it is.
1646 The following properties of a field change when it becomes dynamic:
1647 * If there is no growth limit, there is no final position of the
1648 field; therefore O_AUTOSKIP and O_NL_OVERLOAD are ignored.
1649 * Field justification will be ignored (though whatever justification
1650 is set up will be retained internally and can be queried).
1651 * The dup_field() and link_field() calls copy dynamic-buffer sizes.
1652 If the O_STATIC option is set on one of a collection of links,
1653 buffer resizing will occur only when the field is edited through
1655 * The call field_info() will retrieve the original static size of
1656 the field; use dynamic_field_info() to get the actual dynamic
1661 By default, a field will accept any data that will fit in its input
1662 buffer. However, it is possible to attach a validation type to a
1663 field. If you do this, any attempt to leave the field while it
1664 contains data that doesn't match the validation type will fail. Some
1665 validation types also have a character-validity check for each time a
1666 character is entered in the field.
1668 A field's validation check (if any) is not called when
1669 set_field_buffer() modifies the input buffer, nor when that buffer is
1670 changed through a linked field.
1672 The form library provides a rich set of pre-defined validation types,
1673 and gives you the capability to define custom ones of your own. You
1674 can examine and change field validation attributes with the following
1676 int set_field_type(FIELD *field, /* field to alter */
1677 FIELDTYPE *ftype, /* type to associate */
1678 ...); /* additional arguments*/
1680 FIELDTYPE *field_type(FIELD *field); /* field to query */
1682 The validation type of a field is considered an attribute of the
1683 field. As with other field attributes, Also, doing set_field_type()
1684 with a NULL field default will change the system default for
1685 validation of newly-created fields.
1687 Here are the pre-defined validation types:
1691 This field type accepts alphabetic data; no blanks, no digits, no
1692 special characters (this is checked at character-entry time). It is
1694 int set_field_type(FIELD *field, /* field to alter */
1695 TYPE_ALPHA, /* type to associate */
1696 int width); /* maximum width of field */
1698 The width argument sets a minimum width of data. Typically you'll want
1699 to set this to the field width; if it's greater than the field width,
1700 the validation check will always fail. A minimum width of zero makes
1701 field completion optional.
1705 This field type accepts alphabetic data and digits; no blanks, no
1706 special characters (this is checked at character-entry time). It is
1708 int set_field_type(FIELD *field, /* field to alter */
1709 TYPE_ALNUM, /* type to associate */
1710 int width); /* maximum width of field */
1712 The width argument sets a minimum width of data. As with TYPE_ALPHA,
1713 typically you'll want to set this to the field width; if it's greater
1714 than the field width, the validation check will always fail. A minimum
1715 width of zero makes field completion optional.
1719 This type allows you to restrict a field's values to be among a
1720 specified set of string values (for example, the two-letter postal
1721 codes for U.S. states). It is set up with:
1722 int set_field_type(FIELD *field, /* field to alter */
1723 TYPE_ENUM, /* type to associate */
1724 char **valuelist; /* list of possible values */
1725 int checkcase; /* case-sensitive? */
1726 int checkunique); /* must specify uniquely? */
1728 The valuelist parameter must point at a NULL-terminated list of valid
1729 strings. The checkcase argument, if true, makes comparison with the
1730 string case-sensitive.
1732 When the user exits a TYPE_ENUM field, the validation procedure tries
1733 to complete the data in the buffer to a valid entry. If a complete
1734 choice string has been entered, it is of course valid. But it is also
1735 possible to enter a prefix of a valid string and have it completed for
1738 By default, if you enter such a prefix and it matches more than one
1739 value in the string list, the prefix will be completed to the first
1740 matching value. But the checkunique argument, if true, requires prefix
1741 matches to be unique in order to be valid.
1743 The REQ_NEXT_CHOICE and REQ_PREV_CHOICE input requests can be
1744 particularly useful with these fields.
1748 This field type accepts an integer. It is set up as follows:
1749 int set_field_type(FIELD *field, /* field to alter */
1750 TYPE_INTEGER, /* type to associate */
1751 int padding, /* # places to zero-pad to */
1752 int vmin, int vmax); /* valid range */
1754 Valid characters consist of an optional leading minus and digits. The
1755 range check is performed on exit. If the range maximum is less than or
1756 equal to the minimum, the range is ignored.
1758 If the value passes its range check, it is padded with as many leading
1759 zero digits as necessary to meet the padding argument.
1761 A TYPE_INTEGER value buffer can conveniently be interpreted with the C
1762 library function atoi(3).
1766 This field type accepts a decimal number. It is set up as follows:
1767 int set_field_type(FIELD *field, /* field to alter */
1768 TYPE_NUMERIC, /* type to associate */
1769 int padding, /* # places of precision */
1770 double vmin, double vmax); /* valid range */
1772 Valid characters consist of an optional leading minus and digits.
1773 possibly including a decimal point. If your system supports locale's,
1774 the decimal point character used must be the one defined by your
1775 locale. The range check is performed on exit. If the range maximum is
1776 less than or equal to the minimum, the range is ignored.
1778 If the value passes its range check, it is padded with as many
1779 trailing zero digits as necessary to meet the padding argument.
1781 A TYPE_NUMERIC value buffer can conveniently be interpreted with the C
1782 library function atof(3).
1786 This field type accepts data matching a regular expression. It is set
1788 int set_field_type(FIELD *field, /* field to alter */
1789 TYPE_REGEXP, /* type to associate */
1790 char *regexp); /* expression to match */
1792 The syntax for regular expressions is that of regcomp(3). The check
1793 for regular-expression match is performed on exit.
1795 Direct Field Buffer Manipulation
1797 The chief attribute of a field is its buffer contents. When a form has
1798 been completed, your application usually needs to know the state of
1799 each field buffer. You can find this out with:
1800 char *field_buffer(FIELD *field, /* field to query */
1801 int bufindex); /* number of buffer to query */
1803 Normally, the state of the zero-numbered buffer for each field is set
1804 by the user's editing actions on that field. It's sometimes useful to
1805 be able to set the value of the zero-numbered (or some other) buffer
1806 from your application:
1807 int set_field_buffer(FIELD *field, /* field to alter */
1808 int bufindex, /* number of buffer to alter */
1809 char *value); /* string value to set */
1811 If the field is not large enough and cannot be resized to a
1812 sufficiently large size to contain the specified value, the value will
1813 be truncated to fit.
1815 Calling field_buffer() with a null field pointer will raise an error.
1816 Calling field_buffer() on a field not currently selected for input
1817 will return a correct value. Calling field_buffer() on a field that is
1818 currently selected for input may not necessarily give a correct field
1819 buffer value, because entered data isn't necessarily copied to buffer
1820 zero before the exit validation check. To guarantee that the returned
1821 buffer value reflects on-screen reality, call field_buffer() either
1822 (1) in the field's exit validation check routine, (2) from the field's
1823 or form's initialization or termination hooks, or (3) just after a
1824 REQ_VALIDATION request has been processed by the forms driver.
1828 As with field attributes, form attributes inherit a default from a
1829 system default form structure. These defaults can be queried or set by
1830 of these functions using a form-pointer argument of NULL.
1832 The principal attribute of a form is its field list. You can query and
1833 change this list with:
1834 int set_form_fields(FORM *form, /* form to alter */
1835 FIELD **fields); /* fields to connect */
1837 char *form_fields(FORM *form); /* fetch fields of form */
1839 int field_count(FORM *form); /* count connect fields */
1841 The second argument of set_form_fields() may be a NULL-terminated
1842 field pointer array like the one required by new_form(). In that case,
1843 the old fields of the form are disconnected but not freed (and
1844 eligible to be connected to other forms), then the new fields are
1847 It may also be null, in which case the old fields are disconnected
1848 (and not freed) but no new ones are connected.
1850 The field_count() function simply counts the number of fields
1851 connected to a given from. It returns -1 if the form-pointer argument
1854 Control of Form Display
1856 In the overview section, you saw that to display a form you normally
1857 start by defining its size (and fields), posting it, and refreshing
1858 the screen. There is an hidden step before posting, which is the
1859 association of the form with a frame window (actually, a pair of
1860 windows) within which it will be displayed. By default, the forms
1861 library associates every form with the full-screen window stdscr.
1863 By making this step explicit, you can associate a form with a declared
1864 frame window on your screen display. This can be useful if you want to
1865 adapt the form display to different screen sizes, dynamically tile
1866 forms on the screen, or use a form as part of an interface layout
1869 The two windows associated with each form have the same functions as
1870 their analogues in the menu library. Both these windows are painted
1871 when the form is posted and erased when the form is unposted.
1873 The outer or frame window is not otherwise touched by the form
1874 routines. It exists so the programmer can associate a title, a border,
1875 or perhaps help text with the form and have it properly refreshed or
1876 erased at post/unpost time. The inner window or subwindow is where the
1877 current form page is actually displayed.
1879 In order to declare your own frame window for a form, you'll need to
1880 know the size of the form's bounding rectangle. You can get this
1882 int scale_form(FORM *form, /* form to query */
1883 int *rows, /* form rows */
1884 int *cols); /* form cols */
1886 The form dimensions are passed back in the locations pointed to by the
1887 arguments. Once you have this information, you can use it to declare
1888 of windows, then use one of these functions:
1889 int set_form_win(FORM *form, /* form to alter */
1890 WINDOW *win); /* frame window to connect */
1892 WINDOW *form_win(FORM *form); /* fetch frame window of form */
1894 int set_form_sub(FORM *form, /* form to alter */
1895 WINDOW *win); /* form subwindow to connect */
1897 WINDOW *form_sub(FORM *form); /* fetch form subwindow of form */
1899 Note that curses operations, including refresh(), on the form, should
1900 be done on the frame window, not the form subwindow.
1902 It is possible to check from your application whether all of a
1903 scrollable field is actually displayed within the menu subwindow. Use
1905 int data_ahead(FORM *form); /* form to be queried */
1907 int data_behind(FORM *form); /* form to be queried */
1909 The function data_ahead() returns TRUE if (a) the current field is
1910 one-line and has undisplayed data off to the right, (b) the current
1911 field is multi-line and there is data off-screen below it.
1913 The function data_behind() returns TRUE if the first (upper left hand)
1914 character position is off-screen (not being displayed).
1916 Finally, there is a function to restore the form window's cursor to
1917 the value expected by the forms driver:
1918 int pos_form_cursor(FORM *) /* form to be queried */
1920 If your application changes the form window cursor, call this function
1921 before handing control back to the forms driver in order to
1924 Input Processing in the Forms Driver
1926 The function form_driver() handles virtualized input requests for form
1927 navigation, editing, and validation requests, just as menu_driver does
1928 for menus (see the section on menu input handling).
1929 int form_driver(FORM *form, /* form to pass input to */
1930 int request); /* form request code */
1932 Your input virtualization function needs to take input and then
1933 convert it to either an alphanumeric character (which is treated as
1934 data to be entered in the currently-selected field), or a forms
1937 The forms driver provides hooks (through input-validation and
1938 field-termination functions) with which your application code can
1939 check that the input taken by the driver matched what was expected.
1941 Page Navigation Requests
1943 These requests cause page-level moves through the form, triggering
1944 display of a new form screen.
1947 Move to the next form page.
1950 Move to the previous form page.
1953 Move to the first form page.
1956 Move to the last form page.
1958 These requests treat the list as cyclic; that is, REQ_NEXT_PAGE from
1959 the last page goes to the first, and REQ_PREV_PAGE from the first page
1962 Inter-Field Navigation Requests
1964 These requests handle navigation between fields on the same page.
1970 Move to previous field.
1973 Move to the first field.
1976 Move to the last field.
1979 Move to sorted next field.
1982 Move to sorted previous field.
1985 Move to the sorted first field.
1988 Move to the sorted last field.
1994 Move right to field.
2002 These requests treat the list of fields on a page as cyclic; that is,
2003 REQ_NEXT_FIELD from the last field goes to the first, and
2004 REQ_PREV_FIELD from the first field goes to the last. The order of the
2005 fields for these (and the REQ_FIRST_FIELD and REQ_LAST_FIELD requests)
2006 is simply the order of the field pointers in the form array (as set up
2007 by new_form() or set_form_fields()
2009 It is also possible to traverse the fields as if they had been sorted
2010 in screen-position order, so the sequence goes left-to-right and
2011 top-to-bottom. To do this, use the second group of four
2012 sorted-movement requests.
2014 Finally, it is possible to move between fields using visual directions
2015 up, down, right, and left. To accomplish this, use the third group of
2016 four requests. Note, however, that the position of a form for purposes
2017 of these requests is its upper-left corner.
2019 For example, suppose you have a multi-line field B, and two
2020 single-line fields A and C on the same line with B, with A to the left
2021 of B and C to the right of B. A REQ_MOVE_RIGHT from A will go to B
2022 only if A, B, and C all share the same first line; otherwise it will
2025 Intra-Field Navigation Requests
2027 These requests drive movement of the edit cursor within the currently
2031 Move to next character.
2034 Move to previous character.
2040 Move to previous line.
2046 Move to previous word.
2049 Move to beginning of field.
2052 Move to end of field.
2055 Move to beginning of line.
2058 Move to end of line.
2064 Move right in field.
2072 Each word is separated from the previous and next characters by
2073 whitespace. The commands to move to beginning and end of line or field
2074 look for the first or last non-pad character in their ranges.
2078 Fields that are dynamic and have grown and fields explicitly created
2079 with offscreen rows are scrollable. One-line fields scroll
2080 horizontally; multi-line fields scroll vertically. Most scrolling is
2081 triggered by editing and intra-field movement (the library scrolls the
2082 field to keep the cursor visible). It is possible to explicitly
2083 request scrolling with the following requests:
2086 Scroll vertically forward a line.
2089 Scroll vertically backward a line.
2092 Scroll vertically forward a page.
2095 Scroll vertically backward a page.
2098 Scroll vertically forward half a page.
2101 Scroll vertically backward half a page.
2104 Scroll horizontally forward a character.
2107 Scroll horizontally backward a character.
2110 Scroll horizontally one field width forward.
2113 Scroll horizontally one field width backward.
2116 Scroll horizontally one half field width forward.
2119 Scroll horizontally one half field width backward.
2121 For scrolling purposes, a page of a field is the height of its visible
2126 When you pass the forms driver an ASCII character, it is treated as a
2127 request to add the character to the field's data buffer. Whether this
2128 is an insertion or a replacement depends on the field's edit mode
2129 (insertion is the default.
2131 The following requests support editing the field and changing the edit
2141 New line request (see below for explanation).
2144 Insert space at character location.
2147 Insert blank line at character location.
2150 Delete character at cursor.
2153 Delete previous word at cursor.
2156 Delete line at cursor.
2159 Delete word at cursor.
2162 Clear to end of line.
2165 Clear to end of field.
2170 The behavior of the REQ_NEW_LINE and REQ_DEL_PREV requests is
2171 complicated and partly controlled by a pair of forms options. The
2172 special cases are triggered when the cursor is at the beginning of a
2173 field, or on the last line of the field.
2175 First, we consider REQ_NEW_LINE:
2177 The normal behavior of REQ_NEW_LINE in insert mode is to break the
2178 current line at the position of the edit cursor, inserting the portion
2179 of the current line after the cursor as a new line following the
2180 current and moving the cursor to the beginning of that new line (you
2181 may think of this as inserting a newline in the field buffer).
2183 The normal behavior of REQ_NEW_LINE in overlay mode is to clear the
2184 current line from the position of the edit cursor to end of line. The
2185 cursor is then moved to the beginning of the next line.
2187 However, REQ_NEW_LINE at the beginning of a field, or on the last line
2188 of a field, instead does a REQ_NEXT_FIELD. O_NL_OVERLOAD option is
2189 off, this special action is disabled.
2191 Now, let us consider REQ_DEL_PREV:
2193 The normal behavior of REQ_DEL_PREV is to delete the previous
2194 character. If insert mode is on, and the cursor is at the start of a
2195 line, and the text on that line will fit on the previous one, it
2196 instead appends the contents of the current line to the previous one
2197 and deletes the current line (you may think of this as deleting a
2198 newline from the field buffer).
2200 However, REQ_DEL_PREV at the beginning of a field is instead treated
2201 as a REQ_PREV_FIELD.
2203 If the O_BS_OVERLOAD option is off, this special action is disabled
2204 and the forms driver just returns E_REQUEST_DENIED.
2206 See Form Options for discussion of how to set and clear the overload
2211 If the type of your field is ordered, and has associated functions for
2212 getting the next and previous values of the type from a given value,
2213 there are requests that can fetch that value into the field buffer:
2216 Place the successor value of the current value in the buffer.
2219 Place the predecessor value of the current value in the buffer.
2221 Of the built-in field types, only TYPE_ENUM has built-in successor and
2222 predecessor functions. When you define a field type of your own (see
2223 Custom Validation Types), you can associate our own ordering
2226 Application Commands
2228 Form requests are represented as integers above the curses value
2229 greater than KEY_MAX and less than or equal to the constant
2230 MAX_COMMAND. If your input-virtualization routine returns a value
2231 above MAX_COMMAND, the forms driver will ignore it.
2235 It is possible to set function hooks to be executed whenever the
2236 current field or form changes. Here are the functions that support
2238 typedef void (*HOOK)(); /* pointer to function returning void */
2240 int set_form_init(FORM *form, /* form to alter */
2241 HOOK hook); /* initialization hook */
2243 HOOK form_init(FORM *form); /* form to query */
2245 int set_form_term(FORM *form, /* form to alter */
2246 HOOK hook); /* termination hook */
2248 HOOK form_term(FORM *form); /* form to query */
2250 int set_field_init(FORM *form, /* form to alter */
2251 HOOK hook); /* initialization hook */
2253 HOOK field_init(FORM *form); /* form to query */
2255 int set_field_term(FORM *form, /* form to alter */
2256 HOOK hook); /* termination hook */
2258 HOOK field_term(FORM *form); /* form to query */
2260 These functions allow you to either set or query four different hooks.
2261 In each of the set functions, the second argument should be the
2262 address of a hook function. These functions differ only in the timing
2266 This hook is called when the form is posted; also, just after
2267 each page change operation.
2270 This hook is called when the form is posted; also, just after
2274 This hook is called just after field validation; that is, just
2275 before the field is altered. It is also called when the form is
2279 This hook is called when the form is unposted; also, just
2280 before each page change operation.
2282 Calls to these hooks may be triggered
2283 1. When user editing requests are processed by the forms driver
2284 2. When the current page is changed by set_current_field() call
2285 3. When the current field is changed by a set_form_page() call
2287 See Field Change Commands for discussion of the latter two cases.
2289 You can set a default hook for all fields by passing one of the set
2290 functions a NULL first argument.
2292 You can disable any of these hooks by (re)setting them to NULL, the
2295 Field Change Commands
2297 Normally, navigation through the form will be driven by the user's
2298 input requests. But sometimes it is useful to be able to move the
2299 focus for editing and viewing under control of your application, or
2300 ask which field it currently is in. The following functions help you
2302 int set_current_field(FORM *form, /* form to alter */
2303 FIELD *field); /* field to shift to */
2305 FIELD *current_field(FORM *form); /* form to query */
2307 int field_index(FORM *form, /* form to query */
2308 FIELD *field); /* field to get index of */
2310 The function field_index() returns the index of the given field in the
2311 given form's field array (the array passed to new_form() or
2314 The initial current field of a form is the first active field on the
2315 first page. The function set_form_fields() resets this.
2317 It is also possible to move around by pages.
2318 int set_form_page(FORM *form, /* form to alter */
2319 int page); /* page to go to (0-origin) */
2321 int form_page(FORM *form); /* return form's current page */
2323 The initial page of a newly-created form is 0. The function
2324 set_form_fields() resets this.
2328 Like fields, forms may have control option bits. They can be changed
2329 or queried with these functions:
2330 int set_form_opts(FORM *form, /* form to alter */
2331 int attr); /* attribute to set */
2333 int form_opts_on(FORM *form, /* form to alter */
2334 int attr); /* attributes to turn on */
2336 int form_opts_off(FORM *form, /* form to alter */
2337 int attr); /* attributes to turn off */
2339 int form_opts(FORM *form); /* form to query */
2341 By default, all options are on. Here are the available option bits:
2344 Enable overloading of REQ_NEW_LINE as described in Editing
2345 Requests. The value of this option is ignored on dynamic fields
2346 that have not reached their size limit; these have no last
2347 line, so the circumstances for triggering a REQ_NEXT_FIELD
2351 Enable overloading of REQ_DEL_PREV as described in Editing
2354 The option values are bit-masks and can be composed with logical-or in
2357 Custom Validation Types
2359 The form library gives you the capability to define custom validation
2360 types of your own. Further, the optional additional arguments of
2361 set_field_type effectively allow you to parameterize validation types.
2362 Most of the complications in the validation-type interface have to do
2363 with the handling of the additional arguments within custom validation
2368 The simplest way to create a custom data type is to compose it from
2369 two preexisting ones:
2370 FIELD *link_fieldtype(FIELDTYPE *type1,
2373 This function creates a field type that will accept any of the values
2374 legal for either of its argument field types (which may be either
2375 predefined or programmer-defined). If a set_field_type() call later
2376 requires arguments, the new composite type expects all arguments for
2377 the first type, than all arguments for the second. Order functions
2378 (see Order Requests) associated with the component types will work on
2379 the composite; what it does is check the validation function for the
2380 first type, then for the second, to figure what type the buffer
2381 contents should be treated as.
2385 To create a field type from scratch, you need to specify one or both
2386 of the following things:
2387 * A character-validation function, to check each character as it is
2389 * A field-validation function to be applied on exit from the field.
2391 Here's how you do that:
2392 typedef int (*HOOK)(); /* pointer to function returning int */
2394 FIELDTYPE *new_fieldtype(HOOK f_validate, /* field validator */
2395 HOOK c_validate) /* character validator */
2398 int free_fieldtype(FIELDTYPE *ftype); /* type to free */
2400 At least one of the arguments of new_fieldtype() must be non-NULL. The
2401 forms driver will automatically call the new type's validation
2402 functions at appropriate points in processing a field of the new type.
2404 The function free_fieldtype() deallocates the argument fieldtype,
2405 freeing all storage associated with it.
2407 Normally, a field validator is called when the user attempts to leave
2408 the field. Its first argument is a field pointer, from which it can
2409 get to field buffer 0 and test it. If the function returns TRUE, the
2410 operation succeeds; if it returns FALSE, the edit cursor stays in the
2413 A character validator gets the character passed in as a first
2414 argument. It too should return TRUE if the character is valid, FALSE
2417 Validation Function Arguments
2419 Your field- and character- validation functions will be passed a
2420 second argument as well. This second argument is the address of a
2421 structure (which we'll call a pile) built from any of the
2422 field-type-specific arguments passed to set_field_type(). If no such
2423 arguments are defined for the field type, this pile pointer argument
2426 In order to arrange for such arguments to be passed to your validation
2427 functions, you must associate a small set of storage-management
2428 functions with the type. The forms driver will use these to synthesize
2429 a pile from the trailing arguments of each set_field_type() argument,
2430 and a pointer to the pile will be passed to the validation functions.
2432 Here is how you make the association:
2433 typedef char *(*PTRHOOK)(); /* pointer to function returning (char *) */
2434 typedef void (*VOIDHOOK)(); /* pointer to function returning void */
2436 int set_fieldtype_arg(FIELDTYPE *type, /* type to alter */
2437 PTRHOOK make_str, /* make structure from args */
2438 PTRHOOK copy_str, /* make copy of structure */
2439 VOIDHOOK free_str); /* free structure storage */
2441 Here is how the storage-management hooks are used:
2444 This function is called by set_field_type(). It gets one
2445 argument, a va_list of the type-specific arguments passed to
2446 set_field_type(). It is expected to return a pile pointer to a
2447 data structure that encapsulates those arguments.
2450 This function is called by form library functions that allocate
2451 new field instances. It is expected to take a pile pointer,
2452 copy the pile to allocated storage, and return the address of
2456 This function is called by field- and type-deallocation
2457 routines in the library. It takes a pile pointer argument, and
2458 is expected to free the storage of that pile.
2460 The make_str and copy_str functions may return NULL to signal
2461 allocation failure. The library routines will that call them will
2462 return error indication when this happens. Thus, your validation
2463 functions should never see a NULL file pointer and need not check
2466 Order Functions For Custom Types
2468 Some custom field types are simply ordered in the same well-defined
2469 way that TYPE_ENUM is. For such types, it is possible to define
2470 successor and predecessor functions to support the REQ_NEXT_CHOICE and
2471 REQ_PREV_CHOICE requests. Here's how:
2472 typedef int (*INTHOOK)(); /* pointer to function returning int */
2474 int set_fieldtype_arg(FIELDTYPE *type, /* type to alter */
2475 INTHOOK succ, /* get successor value */
2476 INTHOOK pred); /* get predecessor value */
2478 The successor and predecessor arguments will each be passed two
2479 arguments; a field pointer, and a pile pointer (as for the validation
2480 functions). They are expected to use the function field_buffer() to
2481 read the current value, and set_field_buffer() on buffer 0 to set the
2482 next or previous value. Either hook may return TRUE to indicate
2483 success (a legal next or previous value was set) or FALSE to indicate
2488 The interface for defining custom types is complicated and tricky.
2489 Rather than attempting to create a custom type entirely from scratch,
2490 you should start by studying the library source code for whichever of
2491 the pre-defined types seems to be closest to what you want.
2493 Use that code as a model, and evolve it towards what you really want.
2494 You will avoid many problems and annoyances that way. The code in the
2495 ncurses library has been specifically exempted from the package
2496 copyright to support this.
2498 If your custom type defines order functions, have do something
2499 intuitive with a blank field. A useful convention is to make the
2500 successor of a blank field the types minimum value, and its
2501 predecessor the maximum.