1 Writing Programs with NCURSES
3 Writing Programs with NCURSES
5 by Eric S. Raymond and Zeyd M. Ben-Halim
6 updates since release 1.9.9e by Thomas Dickey
11 + A Brief History of Curses
12 + Scope of This Document
15 + An Overview of Curses
16 o Compiling Programs using Curses
18 o Standard Windows and Function Naming Conventions
24 o Using Forms Characters
25 o Character Attributes and Color
28 + Function Descriptions
29 o Initialization and Wrapup
30 o Causing Output to the Terminal
31 o Low-Level Capability Access
33 + Hints, Tips, and Tricks
34 o Some Notes of Caution
35 o Temporarily Leaving ncurses Mode
36 o Using ncurses under xterm
37 o Handling Multiple Terminal Screens
38 o Testing for Terminal Capabilities
40 o Special Features of ncurses
41 + Compatibility with Older Versions
42 o Refresh of Overlapping Windows
44 + XSI Curses Conformance
46 + Compiling With the Panels Library
48 + Panels, Input, and the Standard Screen
50 + Miscellaneous Other Facilities
52 + Compiling with the menu Library
57 + Processing Menu Input
58 + Miscellaneous Other Features
60 + Compiling with the forms Library
62 + Creating and Freeing Fields and Forms
63 + Fetching and Changing Field Attributes
64 o Fetching Size and Location Data
65 o Changing the Field Location
66 o The Justification Attribute
67 o Field Display Attributes
71 + Variable-Sized Fields
79 + Direct Field Buffer Manipulation
81 + Control of Form Display
82 + Input Processing in the Forms Driver
83 o Page Navigation Requests
84 o Inter-Field Navigation Requests
85 o Intra-Field Navigation Requests
87 o Field Editing Requests
89 o Application Commands
91 + Field Change Commands
93 + Custom Validation Types
96 o Validation Function Arguments
97 o Order Functions For Custom Types
99 _________________________________________________________________
103 This document is an introduction to programming with curses. It is not
104 an exhaustive reference for the curses Application Programming
105 Interface (API); that role is filled by the curses manual pages.
106 Rather, it is intended to help C programmers ease into using the
109 This document is aimed at C applications programmers not yet
110 specifically familiar with ncurses. If you are already an experienced
111 curses programmer, you should nevertheless read the sections on Mouse
112 Interfacing, Debugging, Compatibility with Older Versions, and Hints,
113 Tips, and Tricks. These will bring you up to speed on the special
114 features and quirks of the ncurses implementation. If you are not so
115 experienced, keep reading.
117 The curses package is a subroutine library for terminal-independent
118 screen-painting and input-event handling which presents a high level
119 screen model to the programmer, hiding differences between terminal
120 types and doing automatic optimization of output to change one screen
121 full of text into another. Curses uses terminfo, which is a database
122 format that can describe the capabilities of thousands of different
125 The curses API may seem something of an archaism on UNIX desktops
126 increasingly dominated by X, Motif, and Tcl/Tk. Nevertheless, UNIX
127 still supports tty lines and X supports xterm(1); the curses API has
128 the advantage of (a) back-portability to character-cell terminals, and
129 (b) simplicity. For an application that does not require bit-mapped
130 graphics and multiple fonts, an interface implementation using curses
131 will typically be a great deal simpler and less expensive than one
134 A Brief History of Curses
136 Historically, the first ancestor of curses was the routines written to
137 provide screen-handling for the vi editor; these used the termcap
138 database facility (both released in 3BSD) for describing terminal
139 capabilities. These routines were abstracted into a documented library
140 and first released with the early BSD UNIX versions. All of this work
141 was done by students at the University of California (Berkeley
142 campus). The curses library was first published in 4.0BSD, a year
143 after 3BSD (i.e., late 1980).
145 After graduation, one of those students went to work at AT&T Bell
146 Labs, and made an improved termcap library called terminfo (i.e.,
147 "libterm"), and adapted the curses library to use this. That was
148 subsequently released in System V Release 2 (early 1984). Thereafter,
149 other developers added to the curses and terminfo libraries. For
150 instance, a student at Cornell University wrote an improved terminfo
151 library as well as a tool (tic) to compile the terminal descriptions.
152 As a general rule, AT&T did not identify the developers in the
153 source-code or documentation; the tic and infocmp programs are the
156 System V Release 3 from Bell Labs featured a rewritten and
157 much-improved curses library, along with the tic program (late 1986).
159 To recap, terminfo is based on Berkeley's termcap database, but
160 contains a number of improvements and extensions. Parameterized
161 capabilities strings were introduced, making it possible to describe
162 multiple video attributes, and colors and to handle far more unusual
163 terminals than possible with termcap. In the later AT&T System V
164 releases, curses evolved to use more facilities and offer more
165 capabilities, going far beyond BSD curses in power and flexibility.
167 Scope of This Document
169 This document describes ncurses, a free implementation of the System V
170 curses API with some clearly marked extensions. It includes the
171 following System V curses features:
172 * Support for multiple screen highlights (BSD curses could only
173 handle one "standout" highlight, usually reverse-video).
174 * Support for line- and box-drawing using forms characters.
175 * Recognition of function keys on input.
177 * Support for pads (windows of larger than screen size on which the
178 screen or a subwindow defines a viewport).
180 Also, this package makes use of the insert and delete line and
181 character features of terminals so equipped, and determines how to
182 optimally use these features with no help from the programmer. It
183 allows arbitrary combinations of video attributes to be displayed,
184 even on terminals that leave "magic cookies" on the screen to mark
185 changes in attributes.
187 The ncurses package can also capture and use event reports from a
188 mouse in some environments (notably, xterm under the X window system).
189 This document includes tips for using the mouse.
191 The ncurses package was originated by Pavel Curtis. The original
192 maintainer of this package is Zeyd Ben-Halim <zmbenhal@netcom.com>.
193 Eric S. Raymond <esr@snark.thyrsus.com> wrote many of the new features
194 in versions after 1.8.1 and wrote most of this introduction. Juergen
195 Pfeifer wrote all of the menu and forms code as well as the Ada95
196 binding. Ongoing work is being done by Thomas Dickey (maintainer).
197 Contact the current maintainers at bug-ncurses@gnu.org.
199 This document also describes the panels extension library, similarly
200 modeled on the SVr4 panels facility. This library allows you to
201 associate backing store with each of a stack or deck of overlapping
202 windows, and provides operations for moving windows around in the
203 stack that change their visibility in the natural way (handling window
206 Finally, this document describes in detail the menus and forms
207 extension libraries, also cloned from System V, which support easy
208 construction and sequences of menus and fill-in forms.
212 In this document, the following terminology is used with reasonable
216 A data structure describing a sub-rectangle of the screen
217 (possibly the entire screen). You can write to a window as
218 though it were a miniature screen, scrolling independently of
219 other windows on the physical screen.
222 A subset of windows which are as large as the terminal screen,
223 i.e., they start at the upper left hand corner and encompass
224 the lower right hand corner. One of these, stdscr, is
225 automatically provided for the programmer.
228 The package's idea of what the terminal display currently looks
229 like, i.e., what the user sees now. This is a special screen.
233 An Overview of Curses
235 Compiling Programs using Curses
237 In order to use the library, it is necessary to have certain types and
238 variables defined. Therefore, the programmer must have a line:
241 at the top of the program source. The screen package uses the Standard
242 I/O library, so <curses.h> includes <stdio.h>. <curses.h> also
243 includes <termios.h>, <termio.h>, or <sgtty.h> depending on your
244 system. It is redundant (but harmless) for the programmer to do these
245 includes, too. In linking with curses you need to have -lncurses in
246 your LDFLAGS or on the command line. There is no need for any other
251 In order to update the screen optimally, it is necessary for the
252 routines to know what the screen currently looks like and what the
253 programmer wants it to look like next. For this purpose, a data type
254 (structure) named WINDOW is defined which describes a window image to
255 the routines, including its starting position on the screen (the (y,
256 x) coordinates of the upper left hand corner) and its size. One of
257 these (called curscr, for current screen) is a screen image of what
258 the terminal currently looks like. Another screen (called stdscr, for
259 standard screen) is provided by default to make changes on.
261 A window is a purely internal representation. It is used to build and
262 store a potential image of a portion of the terminal. It does not bear
263 any necessary relation to what is really on the terminal screen; it is
264 more like a scratchpad or write buffer.
266 To make the section of physical screen corresponding to a window
267 reflect the contents of the window structure, the routine refresh()
268 (or wrefresh() if the window is not stdscr) is called.
270 A given physical screen section may be within the scope of any number
271 of overlapping windows. Also, changes can be made to windows in any
272 order, without regard to motion efficiency. Then, at will, the
273 programmer can effectively say "make it look like this," and let the
274 package implementation determine the most efficient way to repaint the
277 Standard Windows and Function Naming Conventions
279 As hinted above, the routines can use several windows, but two are
280 automatically given: curscr, which knows what the terminal looks like,
281 and stdscr, which is what the programmer wants the terminal to look
282 like next. The user should never actually access curscr directly.
283 Changes should be made to through the API, and then the routine
284 refresh() (or wrefresh()) called.
286 Many functions are defined to use stdscr as a default screen. For
287 example, to add a character to stdscr, one calls addch() with the
288 desired character as argument. To write to a different window. use the
289 routine waddch() (for window-specific addch()) is provided. This
290 convention of prepending function names with a "w" when they are to be
291 applied to specific windows is consistent. The only routines which do
292 not follow it are those for which a window must always be specified.
294 In order to move the current (y, x) coordinates from one point to
295 another, the routines move() and wmove() are provided. However, it is
296 often desirable to first move and then perform some I/O operation. In
297 order to avoid clumsiness, most I/O routines can be preceded by the
298 prefix "mv" and the desired (y, x) coordinates prepended to the
299 arguments to the function. For example, the calls
311 mvwaddch(win, y, x, ch);
313 Note that the window description pointer (win) comes before the added
314 (y, x) coordinates. If a function requires a window pointer, it is
315 always the first parameter passed.
319 The curses library sets some variables describing the terminal
321 type name description
322 ------------------------------------------------------------------
323 int LINES number of lines on the terminal
324 int COLS number of columns on the terminal
326 The curses.h also introduces some #define constants and types of
330 boolean type, actually a "char" (e.g., bool doneit;)
333 boolean "true" flag (1).
336 boolean "false" flag (0).
339 error flag returned by routines on a failure (-1).
342 error flag returned by routines when things go right.
346 Now we describe how to actually use the screen package. In it, we
347 assume all updating, reading, etc. is applied to stdscr. These
348 instructions will work on any window, providing you change the
349 function names and parameters as mentioned above.
351 Here is a sample program to motivate the discussion:
356 static void finish(int sig);
359 main(int argc, char *argv[])
363 /* initialize your non-curses data structures here */
365 (void) signal(SIGINT, finish); /* arrange interrupts to terminate */
367 (void) initscr(); /* initialize the curses library */
368 keypad(stdscr, TRUE); /* enable keyboard mapping */
369 (void) nonl(); /* tell curses not to do NL->CR/NL on output */
370 (void) cbreak(); /* take input chars one at a time, no wait for \n */
371 (void) echo(); /* echo input - in color */
378 * Simple color assignment, often all we need. Color pair 0 cannot
379 * be redefined. This example uses the same value for the color
380 * pair as for the foreground color, though of course that is not
383 init_pair(1, COLOR_RED, COLOR_BLACK);
384 init_pair(2, COLOR_GREEN, COLOR_BLACK);
385 init_pair(3, COLOR_YELLOW, COLOR_BLACK);
386 init_pair(4, COLOR_BLUE, COLOR_BLACK);
387 init_pair(5, COLOR_CYAN, COLOR_BLACK);
388 init_pair(6, COLOR_MAGENTA, COLOR_BLACK);
389 init_pair(7, COLOR_WHITE, COLOR_BLACK);
394 int c = getch(); /* refresh, accept single keystroke of input */
395 attrset(COLOR_PAIR(num % 8));
398 /* process the command keystroke */
401 finish(0); /* we are done */
404 static void finish(int sig)
408 /* do your non-curses wrapup here */
415 In order to use the screen package, the routines must know about
416 terminal characteristics, and the space for curscr and stdscr must be
417 allocated. These function initscr() does both these things. Since it
418 must allocate space for the windows, it can overflow memory when
419 attempting to do so. On the rare occasions this happens, initscr()
420 will terminate the program with an error message. initscr() must
421 always be called before any of the routines which affect windows are
422 used. If it is not, the program will core dump as soon as either
423 curscr or stdscr are referenced. However, it is usually best to wait
424 to call it until after you are sure you will need it, like after
425 checking for startup errors. Terminal status changing routines like
426 nl() and cbreak() should be called after initscr().
428 Once the screen windows have been allocated, you can set them up for
429 your program. If you want to, say, allow a screen to scroll, use
430 scrollok(). If you want the cursor to be left in place after the last
431 change, use leaveok(). If this is not done, refresh() will move the
432 cursor to the window's current (y, x) coordinates after updating it.
434 You can create new windows of your own using the functions newwin(),
435 derwin(), and subwin(). The routine delwin() will allow you to get rid
436 of old windows. All the options described above can be applied to any
441 Now that we have set things up, we will want to actually update the
442 terminal. The basic functions used to change what will go on a window
443 are addch() and move(). addch() adds a character at the current (y, x)
444 coordinates. move() changes the current (y, x) coordinates to whatever
445 you want them to be. It returns ERR if you try to move off the window.
446 As mentioned above, you can combine the two into mvaddch() to do both
449 The other output functions, such as addstr() and printw(), all call
450 addch() to add characters to the window.
452 After you have put on the window what you want there, when you want
453 the portion of the terminal covered by the window to be made to look
454 like it, you must call refresh(). In order to optimize finding
455 changes, refresh() assumes that any part of the window not changed
456 since the last refresh() of that window has not been changed on the
457 terminal, i.e., that you have not refreshed a portion of the terminal
458 with an overlapping window. If this is not the case, the routine
459 touchwin() is provided to make it look like the entire window has been
460 changed, thus making refresh() check the whole subsection of the
461 terminal for changes.
463 If you call wrefresh() with curscr as its argument, it will make the
464 screen look like curscr thinks it looks like. This is useful for
465 implementing a command which would redraw the screen in case it get
470 The complementary function to addch() is getch() which, if echo is
471 set, will call addch() to echo the character. Since the screen package
472 needs to know what is on the terminal at all times, if characters are
473 to be echoed, the tty must be in raw or cbreak mode. Since initially
474 the terminal has echoing enabled and is in ordinary "cooked" mode, one
475 or the other has to changed before calling getch(); otherwise, the
476 program's output will be unpredictable.
478 When you need to accept line-oriented input in a window, the functions
479 wgetstr() and friends are available. There is even a wscanw() function
480 that can do scanf()(3)-style multi-field parsing on window input.
481 These pseudo-line-oriented functions turn on echoing while they
484 The example code above uses the call keypad(stdscr, TRUE) to enable
485 support for function-key mapping. With this feature, the getch() code
486 watches the input stream for character sequences that correspond to
487 arrow and function keys. These sequences are returned as
488 pseudo-character values. The #define values returned are listed in the
489 curses.h The mapping from sequences to #define values is determined by
490 key_ capabilities in the terminal's terminfo entry.
492 Using Forms Characters
494 The addch() function (and some others, including box() and border())
495 can accept some pseudo-character arguments which are specially defined
496 by ncurses. These are #define values set up in the curses.h header;
497 see there for a complete list (look for the prefix ACS_).
499 The most useful of the ACS defines are the forms-drawing characters.
500 You can use these to draw boxes and simple graphs on the screen. If
501 the terminal does not have such characters, curses.h will map them to
502 a recognizable (though ugly) set of ASCII defaults.
504 Character Attributes and Color
506 The ncurses package supports screen highlights including standout,
507 reverse-video, underline, and blink. It also supports color, which is
508 treated as another kind of highlight.
510 Highlights are encoded, internally, as high bits of the
511 pseudo-character type (chtype) that curses.h uses to represent the
512 contents of a screen cell. See the curses.h header file for a complete
513 list of highlight mask values (look for the prefix A_).
515 There are two ways to make highlights. One is to logical-or the value
516 of the highlights you want into the character argument of an addch()
517 call, or any other output call that takes a chtype argument.
519 The other is to set the current-highlight value. This is logical-ORed
520 with any highlight you specify the first way. You do this with the
521 functions attron(), attroff(), and attrset(); see the manual pages for
522 details. Color is a special kind of highlight. The package actually
523 thinks in terms of color pairs, combinations of foreground and
524 background colors. The sample code above sets up eight color pairs,
525 all of the guaranteed-available colors on black. Note that each color
526 pair is, in effect, given the name of its foreground color. Any other
527 range of eight non-conflicting values could have been used as the
528 first arguments of the init_pair() values.
530 Once you have done an init_pair() that creates color-pair N, you can
531 use COLOR_PAIR(N) as a highlight that invokes that particular color
532 combination. Note that COLOR_PAIR(N), for constant N, is itself a
533 compile-time constant and can be used in initializers.
537 The ncurses library also provides a mouse interface.
539 NOTE: this facility is specific to ncurses, it is not part of
540 either the XSI Curses standard, nor of System V Release 4, nor BSD
541 curses. System V Release 4 curses contains code with similar
542 interface definitions, however it is not documented. Other than by
543 disassembling the library, we have no way to determine exactly how
544 that mouse code works. Thus, we recommend that you wrap
545 mouse-related code in an #ifdef using the feature macro
546 NCURSES_MOUSE_VERSION so it will not be compiled and linked on
549 Presently, mouse event reporting works in the following environments:
550 * xterm and similar programs such as rxvt.
551 * Linux console, when configured with gpm(1), Alessandro Rubini's
553 * FreeBSD sysmouse (console)
556 The mouse interface is very simple. To activate it, you use the
557 function mousemask(), passing it as first argument a bit-mask that
558 specifies what kinds of events you want your program to be able to
559 see. It will return the bit-mask of events that actually become
560 visible, which may differ from the argument if the mouse device is not
561 capable of reporting some of the event types you specify.
563 Once the mouse is active, your application's command loop should watch
564 for a return value of KEY_MOUSE from wgetch(). When you see this, a
565 mouse event report has been queued. To pick it off the queue, use the
566 function getmouse() (you must do this before the next wgetch(),
567 otherwise another mouse event might come in and make the first one
570 Each call to getmouse() fills a structure (the address of which you
571 will pass it) with mouse event data. The event data includes
572 zero-origin, screen-relative character-cell coordinates of the mouse
573 pointer. It also includes an event mask. Bits in this mask will be
574 set, corresponding to the event type being reported.
576 The mouse structure contains two additional fields which may be
577 significant in the future as ncurses interfaces to new kinds of
578 pointing device. In addition to x and y coordinates, there is a slot
579 for a z coordinate; this might be useful with touch-screens that can
580 return a pressure or duration parameter. There is also a device ID
581 field, which could be used to distinguish between multiple pointing
584 The class of visible events may be changed at any time via
585 mousemask(). Events that can be reported include presses, releases,
586 single-, double- and triple-clicks (you can set the maximum
587 button-down time for clicks). If you do not make clicks visible, they
588 will be reported as press-release pairs. In some environments, the
589 event mask may include bits reporting the state of shift, alt, and
590 ctrl keys on the keyboard during the event.
592 A function to check whether a mouse event fell within a given window
593 is also supplied. You can use this to see whether a given window
594 should consider a mouse event relevant to it.
596 Because mouse event reporting will not be available in all
597 environments, it would be unwise to build ncurses applications that
598 require the use of a mouse. Rather, you should use the mouse as a
599 shortcut for point-and-shoot commands your application would normally
600 accept from the keyboard. Two of the test games in the ncurses
601 distribution (bs and knight) contain code that illustrates how this
604 See the manual page curs_mouse(3X) for full details of the
605 mouse-interface functions.
609 In order to clean up after the ncurses routines, the routine endwin()
610 is provided. It restores tty modes to what they were when initscr()
611 was first called, and moves the cursor down to the lower-left corner.
612 Thus, anytime after the call to initscr, endwin() should be called
615 Function Descriptions
617 We describe the detailed behavior of some important curses functions
618 here, as a supplement to the manual page descriptions.
620 Initialization and Wrapup
623 The first function called should almost always be initscr().
624 This will determine the terminal type and initialize curses
625 data structures. initscr() also arranges that the first call to
626 refresh() will clear the screen. If an error occurs a message
627 is written to standard error and the program exits. Otherwise
628 it returns a pointer to stdscr. A few functions may be called
629 before initscr (slk_init(), filter(), ripoffline(), use_env(),
630 and, if you are using multiple terminals, newterm().)
633 Your program should always call endwin() before exiting or
634 shelling out of the program. This function will restore tty
635 modes, move the cursor to the lower left corner of the screen,
636 reset the terminal into the proper non-visual mode. Calling
637 refresh() or doupdate() after a temporary escape from the
638 program will restore the ncurses screen from before the escape.
640 newterm(type, ofp, ifp)
641 A program which outputs to more than one terminal should use
642 newterm() instead of initscr(). newterm() should be called once
643 for each terminal. It returns a variable of type SCREEN * which
644 should be saved as a reference to that terminal. (NOTE: a
645 SCREEN variable is not a screen in the sense we are describing
646 in this introduction, but a collection of parameters used to
647 assist in optimizing the display.) The arguments are the type
648 of the terminal (a string) and FILE pointers for the output and
649 input of the terminal. If type is NULL then the environment
650 variable $TERM is used. endwin() should called once at wrapup
651 time for each terminal opened using this function.
654 This function is used to switch to a different terminal
655 previously opened by newterm(). The screen reference for the
656 new terminal is passed as the parameter. The previous terminal
657 is returned by the function. All other calls affect only the
661 The inverse of newterm(); deallocates the data structures
662 associated with a given SCREEN reference.
664 Causing Output to the Terminal
666 refresh() and wrefresh(win)
667 These functions must be called to actually get any output on
668 the terminal, as other routines merely manipulate data
669 structures. wrefresh() copies the named window to the physical
670 terminal screen, taking into account what is already there in
671 order to do optimizations. refresh() does a refresh of stdscr.
672 Unless leaveok() has been enabled, the physical cursor of the
673 terminal is left at the location of the window's cursor.
675 doupdate() and wnoutrefresh(win)
676 These two functions allow multiple updates with more efficiency
677 than wrefresh. To use them, it is important to understand how
678 curses works. In addition to all the window structures, curses
679 keeps two data structures representing the terminal screen: a
680 physical screen, describing what is actually on the screen, and
681 a virtual screen, describing what the programmer wants to have
682 on the screen. wrefresh works by first copying the named window
683 to the virtual screen (wnoutrefresh()), and then calling the
684 routine to update the screen (doupdate()). If the programmer
685 wishes to output several windows at once, a series of calls to
686 wrefresh will result in alternating calls to wnoutrefresh() and
687 doupdate(), causing several bursts of output to the screen. By
688 calling wnoutrefresh() for each window, it is then possible to
689 call doupdate() once, resulting in only one burst of output,
690 with fewer total characters transmitted (this also avoids a
691 visually annoying flicker at each update).
693 Low-Level Capability Access
695 setupterm(term, filenum, errret)
696 This routine is called to initialize a terminal's description,
697 without setting up the curses screen structures or changing the
698 tty-driver mode bits. term is the character string representing
699 the name of the terminal being used. filenum is the UNIX file
700 descriptor of the terminal to be used for output. errret is a
701 pointer to an integer, in which a success or failure indication
702 is returned. The values returned can be 1 (all is well), 0 (no
703 such terminal), or -1 (some problem locating the terminfo
706 The value of term can be given as NULL, which will cause the
707 value of TERM in the environment to be used. The errret pointer
708 can also be given as NULL, meaning no error code is wanted. If
709 errret is defaulted, and something goes wrong, setupterm() will
710 print an appropriate error message and exit, rather than
711 returning. Thus, a simple program can call setupterm(0, 1, 0)
712 and not worry about initialization errors.
714 After the call to setupterm(), the global variable cur_term is
715 set to point to the current structure of terminal capabilities.
716 By calling setupterm() for each terminal, and saving and
717 restoring cur_term, it is possible for a program to use two or
718 more terminals at once. Setupterm() also stores the names
719 section of the terminal description in the global character
720 array ttytype[]. Subsequent calls to setupterm() will overwrite
721 this array, so you will have to save it yourself if need be.
725 NOTE: These functions are not part of the standard curses API!
728 This function can be used to explicitly set a trace level. If
729 the trace level is nonzero, execution of your program will
730 generate a file called "trace" in the current working directory
731 containing a report on the library's actions. Higher trace
732 levels enable more detailed (and verbose) reporting -- see
733 comments attached to TRACE_ defines in the curses.h file for
734 details. (It is also possible to set a trace level by assigning
735 a trace level value to the environment variable NCURSES_TRACE).
738 This function can be used to output your own debugging
739 information. It is only available only if you link with
740 -lncurses_g. It can be used the same way as printf(), only it
741 outputs a newline after the end of arguments. The output goes
742 to a file called trace in the current directory.
744 Trace logs can be difficult to interpret due to the sheer volume of
745 data dumped in them. There is a script called tracemunch included with
746 the ncurses distribution that can alleviate this problem somewhat; it
747 compacts long sequences of similar operations into more succinct
748 single-line pseudo-operations. These pseudo-ops can be distinguished
749 by the fact that they are named in capital letters.
751 Hints, Tips, and Tricks
753 The ncurses manual pages are a complete reference for this library. In
754 the remainder of this document, we discuss various useful methods that
755 may not be obvious from the manual page descriptions.
757 Some Notes of Caution
759 If you find yourself thinking you need to use noraw() or nocbreak(),
760 think again and move carefully. It is probably better design to use
761 getstr() or one of its relatives to simulate cooked mode. The noraw()
762 and nocbreak() functions try to restore cooked mode, but they may end
763 up clobbering some control bits set before you started your
764 application. Also, they have always been poorly documented, and are
765 likely to hurt your application's usability with other curses
768 Bear in mind that refresh() is a synonym for wrefresh(stdscr). Do not
769 try to mix use of stdscr with use of windows declared by newwin(); a
770 refresh() call will blow them off the screen. The right way to handle
771 this is to use subwin(), or not touch stdscr at all and tile your
772 screen with declared windows which you then wnoutrefresh() somewhere
773 in your program event loop, with a single doupdate() call to trigger
776 You are much less likely to run into problems if you design your
777 screen layouts to use tiled rather than overlapping windows.
778 Historically, curses support for overlapping windows has been weak,
779 fragile, and poorly documented. The ncurses library is not yet an
780 exception to this rule.
782 There is a panels library included in the ncurses distribution that
783 does a pretty good job of strengthening the overlapping-windows
786 Try to avoid using the global variables LINES and COLS. Use getmaxyx()
787 on the stdscr context instead. Reason: your code may be ported to run
788 in an environment with window resizes, in which case several screens
789 could be open with different sizes.
791 Temporarily Leaving NCURSES Mode
793 Sometimes you will want to write a program that spends most of its
794 time in screen mode, but occasionally returns to ordinary "cooked"
795 mode. A common reason for this is to support shell-out. This behavior
796 is simple to arrange in ncurses.
798 To leave ncurses mode, call endwin() as you would if you were
799 intending to terminate the program. This will take the screen back to
800 cooked mode; you can do your shell-out. When you want to return to
801 ncurses mode, simply call refresh() or doupdate(). This will repaint
804 There is a boolean function, isendwin(), which code can use to test
805 whether ncurses screen mode is active. It returns TRUE in the interval
806 between an endwin() call and the following refresh(), FALSE otherwise.
808 Here is some sample code for shellout:
809 addstr("Shelling out...");
810 def_prog_mode(); /* save current tty modes */
811 endwin(); /* restore original tty modes */
812 system("sh"); /* run shell */
813 addstr("returned.\n"); /* prepare return message */
814 refresh(); /* restore save modes, repaint screen */
816 Using NCURSES under XTERM
818 A resize operation in X sends SIGWINCH to the application running
819 under xterm. The easiest way to handle SIGWINCH is to do an endwin,
820 followed by an refresh and a screen repaint you code yourself. The
821 refresh will pick up the new screen size from the xterm's environment.
823 That is the standard way, of course (it even works with some vendor's
824 curses implementations). Its drawback is that it clears the screen to
825 reinitialize the display, and does not resize subwindows which must be
826 shrunk. Ncurses provides an extension which works better, the
827 resizeterm function. That function ensures that all windows are
828 limited to the new screen dimensions, and pads stdscr with blanks if
829 the screen is larger.
831 The ncurses library provides a SIGWINCH signal handler, which pushes a
832 KEY_RESIZE via the wgetch() calls. When ncurses returns that code, it
833 calls resizeterm to update the size of the standard screen's window,
834 repainting that (filling with blanks or truncating as needed). It also
835 resizes other windows, but its effect may be less satisfactory because
836 it cannot know how you want the screen re-painted. You will usually
837 have to write special-purpose code to handle KEY_RESIZE yourself.
839 Handling Multiple Terminal Screens
841 The initscr() function actually calls a function named newterm() to do
842 most of its work. If you are writing a program that opens multiple
843 terminals, use newterm() directly.
845 For each call, you will have to specify a terminal type and a pair of
846 file pointers; each call will return a screen reference, and stdscr
847 will be set to the last one allocated. You will switch between screens
848 with the set_term call. Note that you will also have to call
849 def_shell_mode and def_prog_mode on each tty yourself.
851 Testing for Terminal Capabilities
853 Sometimes you may want to write programs that test for the presence of
854 various capabilities before deciding whether to go into ncurses mode.
855 An easy way to do this is to call setupterm(), then use the functions
856 tigetflag(), tigetnum(), and tigetstr() to do your testing.
858 A particularly useful case of this often comes up when you want to
859 test whether a given terminal type should be treated as "smart"
860 (cursor-addressable) or "stupid". The right way to test this is to see
861 if the return value of tigetstr("cup") is non-NULL. Alternatively, you
862 can include the term.h file and test the value of the macro
867 Use the addchstr() family of functions for fast screen-painting of
868 text when you know the text does not contain any control characters.
869 Try to make attribute changes infrequent on your screens. Do not use
870 the immedok() option!
872 Special Features of NCURSES
874 The wresize() function allows you to resize a window in place. The
875 associated resizeterm() function simplifies the construction of
876 SIGWINCH handlers, for resizing all windows.
878 The define_key() function allows you to define at runtime function-key
879 control sequences which are not in the terminal description. The
880 keyok() function allows you to temporarily enable or disable
881 interpretation of any function-key control sequence.
883 The use_default_colors() function allows you to construct applications
884 which can use the terminal's default foreground and background colors
885 as an additional "default" color. Several terminal emulators support
886 this feature, which is based on ISO 6429.
888 Ncurses supports up 16 colors, unlike SVr4 curses which defines only
889 8. While most terminals which provide color allow only 8 colors, about
890 a quarter (including XFree86 xterm) support 16 colors.
892 Compatibility with Older Versions
894 Despite our best efforts, there are some differences between ncurses
895 and the (undocumented!) behavior of older curses implementations.
896 These arise from ambiguities or omissions in the documentation of the
899 Refresh of Overlapping Windows
901 If you define two windows A and B that overlap, and then alternately
902 scribble on and refresh them, the changes made to the overlapping
903 region under historic curses versions were often not documented
906 To understand why this is a problem, remember that screen updates are
907 calculated between two representations of the entire display. The
908 documentation says that when you refresh a window, it is first copied
909 to the virtual screen, and then changes are calculated to update the
910 physical screen (and applied to the terminal). But "copied to" is not
911 very specific, and subtle differences in how copying works can produce
912 different behaviors in the case where two overlapping windows are each
913 being refreshed at unpredictable intervals.
915 What happens to the overlapping region depends on what wnoutrefresh()
916 does with its argument -- what portions of the argument window it
917 copies to the virtual screen. Some implementations do "change copy",
918 copying down only locations in the window that have changed (or been
919 marked changed with wtouchln() and friends). Some implementations do
920 "entire copy", copying all window locations to the virtual screen
921 whether or not they have changed.
923 The ncurses library itself has not always been consistent on this
924 score. Due to a bug, versions 1.8.7 to 1.9.8a did entire copy.
925 Versions 1.8.6 and older, and versions 1.9.9 and newer, do change
928 For most commercial curses implementations, it is not documented and
929 not known for sure (at least not to the ncurses maintainers) whether
930 they do change copy or entire copy. We know that System V release 3
931 curses has logic in it that looks like an attempt to do change copy,
932 but the surrounding logic and data representations are sufficiently
933 complex, and our knowledge sufficiently indirect, that it is hard to
934 know whether this is reliable. It is not clear what the SVr4
935 documentation and XSI standard intend. The XSI Curses standard barely
936 mentions wnoutrefresh(); the SVr4 documents seem to be describing
937 entire-copy, but it is possible with some effort and straining to read
940 It might therefore be unwise to rely on either behavior in programs
941 that might have to be linked with other curses implementations.
942 Instead, you can do an explicit touchwin() before the wnoutrefresh()
943 call to guarantee an entire-contents copy anywhere.
945 The really clean way to handle this is to use the panels library. If,
946 when you want a screen update, you do update_panels(), it will do all
947 the necessary wnoutrefresh() calls for whatever panel stacking order
948 you have defined. Then you can do one doupdate() and there will be a
949 single burst of physical I/O that will do all your updates.
953 If you have been using a very old versions of ncurses (1.8.7 or older)
954 you may be surprised by the behavior of the erase functions. In older
955 versions, erased areas of a window were filled with a blank modified
956 by the window's current attribute (as set by wattrset(), wattron(),
957 wattroff() and friends).
959 In newer versions, this is not so. Instead, the attribute of erased
960 blanks is normal unless and until it is modified by the functions
961 bkgdset() or wbkgdset().
963 This change in behavior conforms ncurses to System V Release 4 and the
966 XSI Curses Conformance
968 The ncurses library is intended to be base-level conformant with the
969 XSI Curses standard from X/Open. Many extended-level features (in
970 fact, almost all features not directly concerned with wide characters
971 and internationalization) are also supported.
973 One effect of XSI conformance is the change in behavior described
974 under "Background Erase -- Compatibility with Old Versions".
976 Also, ncurses meets the XSI requirement that every macro entry point
977 have a corresponding function which may be linked (and will be
978 prototype-checked) if the macro definition is disabled with #undef.
982 The ncurses library by itself provides good support for screen
983 displays in which the windows are tiled (non-overlapping). In the more
984 general case that windows may overlap, you have to use a series of
985 wnoutrefresh() calls followed by a doupdate(), and be careful about
986 the order you do the window refreshes in. It has to be bottom-upwards,
987 otherwise parts of windows that should be obscured will show through.
989 When your interface design is such that windows may dive deeper into
990 the visibility stack or pop to the top at runtime, the resulting
991 book-keeping can be tedious and difficult to get right. Hence the
994 The panel library first appeared in AT&T System V. The version
995 documented here is the panel code distributed with ncurses.
997 Compiling With the Panels Library
999 Your panels-using modules must import the panels library declarations
1003 and must be linked explicitly with the panels library using an -lpanel
1004 argument. Note that they must also link the ncurses library with
1005 -lncurses. Many linkers are two-pass and will accept either order, but
1006 it is still good practice to put -lpanel first and -lncurses second.
1010 A panel object is a window that is implicitly treated as part of a
1011 deck including all other panel objects. The deck has an implicit
1012 bottom-to-top visibility order. The panels library includes an update
1013 function (analogous to refresh()) that displays all panels in the deck
1014 in the proper order to resolve overlaps. The standard window, stdscr,
1015 is considered below all panels.
1017 Details on the panels functions are available in the man pages. We
1018 will just hit the highlights here.
1020 You create a panel from a window by calling new_panel() on a window
1021 pointer. It then becomes the top of the deck. The panel's window is
1022 available as the value of panel_window() called with the panel pointer
1025 You can delete a panel (removing it from the deck) with del_panel.
1026 This will not deallocate the associated window; you have to do that
1027 yourself. You can replace a panel's window with a different window by
1028 calling replace_window. The new window may be of different size; the
1029 panel code will re-compute all overlaps. This operation does not
1030 change the panel's position in the deck.
1032 To move a panel's window, use move_panel(). The mvwin() function on
1033 the panel's window is not sufficient because it does not update the
1034 panels library's representation of where the windows are. This
1035 operation leaves the panel's depth, contents, and size unchanged.
1037 Two functions (top_panel(), bottom_panel()) are provided for
1038 rearranging the deck. The first pops its argument window to the top of
1039 the deck; the second sends it to the bottom. Either operation leaves
1040 the panel's screen location, contents, and size unchanged.
1042 The function update_panels() does all the wnoutrefresh() calls needed
1043 to prepare for doupdate() (which you must call yourself, afterwards).
1045 Typically, you will want to call update_panels() and doupdate() just
1046 before accepting command input, once in each cycle of interaction with
1047 the user. If you call update_panels() after each and every panel
1048 write, you will generate a lot of unnecessary refresh activity and
1051 Panels, Input, and the Standard Screen
1053 You should not mix wnoutrefresh() or wrefresh() operations with panels
1054 code; this will work only if the argument window is either in the top
1055 panel or unobscured by any other panels.
1057 The stsdcr window is a special case. It is considered below all
1058 panels. Because changes to panels may obscure parts of stdscr, though,
1059 you should call update_panels() before doupdate() even when you only
1062 Note that wgetch automatically calls wrefresh. Therefore, before
1063 requesting input from a panel window, you need to be sure that the
1064 panel is totally unobscured.
1066 There is presently no way to display changes to one obscured panel
1067 without repainting all panels.
1071 It is possible to remove a panel from the deck temporarily; use
1072 hide_panel for this. Use show_panel() to render it visible again. The
1073 predicate function panel_hidden tests whether or not a panel is
1076 The panel_update code ignores hidden panels. You cannot do top_panel()
1077 or bottom_panel on a hidden panel(). Other panels operations are
1080 Miscellaneous Other Facilities
1082 It is possible to navigate the deck using the functions panel_above()
1083 and panel_below. Handed a panel pointer, they return the panel above
1084 or below that panel. Handed NULL, they return the bottom-most or
1087 Every panel has an associated user pointer, not used by the panel
1088 code, to which you can attach application data. See the man page
1089 documentation of set_panel_userptr() and panel_userptr for details.
1093 A menu is a screen display that assists the user to choose some subset
1094 of a given set of items. The menu library is a curses extension that
1095 supports easy programming of menu hierarchies with a uniform but
1098 The menu library first appeared in AT&T System V. The version
1099 documented here is the menu code distributed with ncurses.
1101 Compiling With the menu Library
1103 Your menu-using modules must import the menu library declarations with
1106 and must be linked explicitly with the menus library using an -lmenu
1107 argument. Note that they must also link the ncurses library with
1108 -lncurses. Many linkers are two-pass and will accept either order, but
1109 it is still good practice to put -lmenu first and -lncurses second.
1113 The menus created by this library consist of collections of items
1114 including a name string part and a description string part. To make
1115 menus, you create groups of these items and connect them with menu
1118 The menu can then by posted, that is written to an associated window.
1119 Actually, each menu has two associated windows; a containing window in
1120 which the programmer can scribble titles or borders, and a subwindow
1121 in which the menu items proper are displayed. If this subwindow is too
1122 small to display all the items, it will be a scrollable viewport on
1123 the collection of items.
1125 A menu may also be unposted (that is, undisplayed), and finally freed
1126 to make the storage associated with it and its items available for
1129 The general flow of control of a menu program looks like this:
1130 1. Initialize curses.
1131 2. Create the menu items, using new_item().
1132 3. Create the menu using new_menu().
1133 4. Post the menu using post_menu().
1134 5. Refresh the screen.
1135 6. Process user requests via an input loop.
1136 7. Unpost the menu using unpost_menu().
1137 8. Free the menu, using free_menu().
1138 9. Free the items using free_item().
1139 10. Terminate curses.
1143 Menus may be multi-valued or (the default) single-valued (see the
1144 manual page menu_opts(3x) to see how to change the default). Both
1145 types always have a current item.
1147 From a single-valued menu you can read the selected value simply by
1148 looking at the current item. From a multi-valued menu, you get the
1149 selected set by looping through the items applying the item_value()
1150 predicate function. Your menu-processing code can use the function
1151 set_item_value() to flag the items in the select set.
1153 Menu items can be made unselectable using set_item_opts() or
1154 item_opts_off() with the O_SELECTABLE argument. This is the only
1155 option so far defined for menus, but it is good practice to code as
1156 though other option bits might be on.
1160 The menu library calculates a minimum display size for your window,
1161 based on the following variables:
1162 * The number and maximum length of the menu items
1163 * Whether the O_ROWMAJOR option is enabled
1164 * Whether display of descriptions is enabled
1165 * Whatever menu format may have been set by the programmer
1166 * The length of the menu mark string used for highlighting selected
1169 The function set_menu_format() allows you to set the maximum size of
1170 the viewport or menu page that will be used to display menu items. You
1171 can retrieve any format associated with a menu with menu_format(). The
1172 default format is rows=16, columns=1.
1174 The actual menu page may be smaller than the format size. This depends
1175 on the item number and size and whether O_ROWMAJOR is on. This option
1176 (on by default) causes menu items to be displayed in a "raster-scan"
1177 pattern, so that if more than one item will fit horizontally the first
1178 couple of items are side-by-side in the top row. The alternative is
1179 column-major display, which tries to put the first several items in
1182 As mentioned above, a menu format not large enough to allow all items
1183 to fit on-screen will result in a menu display that is vertically
1186 You can scroll it with requests to the menu driver, which will be
1187 described in the section on menu input handling.
1189 Each menu has a mark string used to visually tag selected items; see
1190 the menu_mark(3x) manual page for details. The mark string length also
1191 influences the menu page size.
1193 The function scale_menu() returns the minimum display size that the
1194 menu code computes from all these factors. There are other menu
1195 display attributes including a select attribute, an attribute for
1196 selectable items, an attribute for unselectable items, and a pad
1197 character used to separate item name text from description text. These
1198 have reasonable defaults which the library allows you to change (see
1199 the menu_attribs(3x) manual page.
1203 Each menu has, as mentioned previously, a pair of associated windows.
1204 Both these windows are painted when the menu is posted and erased when
1205 the menu is unposted.
1207 The outer or frame window is not otherwise touched by the menu
1208 routines. It exists so the programmer can associate a title, a border,
1209 or perhaps help text with the menu and have it properly refreshed or
1210 erased at post/unpost time. The inner window or subwindow is where the
1211 current menu page is displayed.
1213 By default, both windows are stdscr. You can set them with the
1214 functions in menu_win(3x).
1216 When you call post_menu(), you write the menu to its subwindow. When
1217 you call unpost_menu(), you erase the subwindow, However, neither of
1218 these actually modifies the screen. To do that, call wrefresh() or
1221 Processing Menu Input
1223 The main loop of your menu-processing code should call menu_driver()
1224 repeatedly. The first argument of this routine is a menu pointer; the
1225 second is a menu command code. You should write an input-fetching
1226 routine that maps input characters to menu command codes, and pass its
1227 output to menu_driver(). The menu command codes are fully documented
1230 The simplest group of command codes is REQ_NEXT_ITEM, REQ_PREV_ITEM,
1231 REQ_FIRST_ITEM, REQ_LAST_ITEM, REQ_UP_ITEM, REQ_DOWN_ITEM,
1232 REQ_LEFT_ITEM, REQ_RIGHT_ITEM. These change the currently selected
1233 item. These requests may cause scrolling of the menu page if it only
1234 partially displayed.
1236 There are explicit requests for scrolling which also change the
1237 current item (because the select location does not change, but the
1238 item there does). These are REQ_SCR_DLINE, REQ_SCR_ULINE,
1239 REQ_SCR_DPAGE, and REQ_SCR_UPAGE.
1241 The REQ_TOGGLE_ITEM selects or deselects the current item. It is for
1242 use in multi-valued menus; if you use it with O_ONEVALUE on, you will
1243 get an error return (E_REQUEST_DENIED).
1245 Each menu has an associated pattern buffer. The menu_driver() logic
1246 tries to accumulate printable ASCII characters passed in in that
1247 buffer; when it matches a prefix of an item name, that item (or the
1248 next matching item) is selected. If appending a character yields no
1249 new match, that character is deleted from the pattern buffer, and
1250 menu_driver() returns E_NO_MATCH.
1252 Some requests change the pattern buffer directly: REQ_CLEAR_PATTERN,
1253 REQ_BACK_PATTERN, REQ_NEXT_MATCH, REQ_PREV_MATCH. The latter two are
1254 useful when pattern buffer input matches more than one item in a
1257 Each successful scroll or item navigation request clears the pattern
1258 buffer. It is also possible to set the pattern buffer explicitly with
1261 Finally, menu driver requests above the constant MAX_COMMAND are
1262 considered application-specific commands. The menu_driver() code
1263 ignores them and returns E_UNKNOWN_COMMAND.
1265 Miscellaneous Other Features
1267 Various menu options can affect the processing and visual appearance
1268 and input processing of menus. See menu_opts(3x) for details.
1270 It is possible to change the current item from application code; this
1271 is useful if you want to write your own navigation requests. It is
1272 also possible to explicitly set the top row of the menu display. See
1273 mitem_current(3x). If your application needs to change the menu
1274 subwindow cursor for any reason, pos_menu_cursor() will restore it to
1275 the correct location for continuing menu driver processing.
1277 It is possible to set hooks to be called at menu initialization and
1278 wrapup time, and whenever the selected item changes. See
1281 Each item, and each menu, has an associated user pointer on which you
1282 can hang application data. See mitem_userptr(3x) and menu_userptr(3x).
1286 The form library is a curses extension that supports easy programming
1287 of on-screen forms for data entry and program control.
1289 The form library first appeared in AT&T System V. The version
1290 documented here is the form code distributed with ncurses.
1292 Compiling With the form Library
1294 Your form-using modules must import the form library declarations with
1297 and must be linked explicitly with the forms library using an -lform
1298 argument. Note that they must also link the ncurses library with
1299 -lncurses. Many linkers are two-pass and will accept either order, but
1300 it is still good practice to put -lform first and -lncurses second.
1304 A form is a collection of fields; each field may be either a label
1305 (explanatory text) or a data-entry location. Long forms may be
1306 segmented into pages; each entry to a new page clears the screen.
1308 To make forms, you create groups of fields and connect them with form
1309 frame objects; the form library makes this relatively simple.
1311 Once defined, a form can be posted, that is written to an associated
1312 window. Actually, each form has two associated windows; a containing
1313 window in which the programmer can scribble titles or borders, and a
1314 subwindow in which the form fields proper are displayed.
1316 As the form user fills out the posted form, navigation and editing
1317 keys support movement between fields, editing keys support modifying
1318 field, and plain text adds to or changes data in a current field. The
1319 form library allows you (the forms designer) to bind each navigation
1320 and editing key to any keystroke accepted by curses Fields may have
1321 validation conditions on them, so that they check input data for type
1322 and value. The form library supplies a rich set of pre-defined field
1323 types, and makes it relatively easy to define new ones.
1325 Once its transaction is completed (or aborted), a form may be unposted
1326 (that is, undisplayed), and finally freed to make the storage
1327 associated with it and its items available for re-use.
1329 The general flow of control of a form program looks like this:
1330 1. Initialize curses.
1331 2. Create the form fields, using new_field().
1332 3. Create the form using new_form().
1333 4. Post the form using post_form().
1334 5. Refresh the screen.
1335 6. Process user requests via an input loop.
1336 7. Unpost the form using unpost_form().
1337 8. Free the form, using free_form().
1338 9. Free the fields using free_field().
1339 10. Terminate curses.
1341 Note that this looks much like a menu program; the form library
1342 handles tasks which are in many ways similar, and its interface was
1343 obviously designed to resemble that of the menu library wherever
1346 In forms programs, however, the "process user requests" is somewhat
1347 more complicated than for menus. Besides menu-like navigation
1348 operations, the menu driver loop has to support field editing and data
1351 Creating and Freeing Fields and Forms
1353 The basic function for creating fields is new_field():
1354 FIELD *new_field(int height, int width, /* new field size */
1355 int top, int left, /* upper left corner */
1356 int offscreen, /* number of offscreen rows */
1357 int nbuf); /* number of working buffers */
1359 Menu items always occupy a single row, but forms fields may have
1360 multiple rows. So new_field() requires you to specify a width and
1361 height (the first two arguments, which mist both be greater than
1364 You must also specify the location of the field's upper left corner on
1365 the screen (the third and fourth arguments, which must be zero or
1366 greater). Note that these coordinates are relative to the form
1367 subwindow, which will coincide with stdscr by default but need not be
1368 stdscr if you have done an explicit set_form_win() call.
1370 The fifth argument allows you to specify a number of off-screen rows.
1371 If this is zero, the entire field will always be displayed. If it is
1372 nonzero, the form will be scrollable, with only one screen-full
1373 (initially the top part) displayed at any given time. If you make a
1374 field dynamic and grow it so it will no longer fit on the screen, the
1375 form will become scrollable even if the offscreen argument was
1378 The forms library allocates one working buffer per field; the size of
1379 each buffer is ((height + offscreen)*width + 1, one character for each
1380 position in the field plus a NUL terminator. The sixth argument is the
1381 number of additional data buffers to allocate for the field; your
1382 application can use them for its own purposes.
1383 FIELD *dup_field(FIELD *field, /* field to copy */
1384 int top, int left); /* location of new copy */
1386 The function dup_field() duplicates an existing field at a new
1387 location. Size and buffering information are copied; some attribute
1388 flags and status bits are not (see the form_field_new(3X) for
1390 FIELD *link_field(FIELD *field, /* field to copy */
1391 int top, int left); /* location of new copy */
1393 The function link_field() also duplicates an existing field at a new
1394 location. The difference from dup_field() is that it arranges for the
1395 new field's buffer to be shared with the old one.
1397 Besides the obvious use in making a field editable from two different
1398 form pages, linked fields give you a way to hack in dynamic labels. If
1399 you declare several fields linked to an original, and then make them
1400 inactive, changes from the original will still be propagated to the
1403 As with duplicated fields, linked fields have attribute bits separate
1406 As you might guess, all these field-allocations return NULL if the
1407 field allocation is not possible due to an out-of-memory error or
1408 out-of-bounds arguments.
1410 To connect fields to a form, use
1411 FORM *new_form(FIELD **fields);
1413 This function expects to see a NULL-terminated array of field
1414 pointers. Said fields are connected to a newly-allocated form object;
1415 its address is returned (or else NULL if the allocation fails).
1417 Note that new_field() does not copy the pointer array into private
1418 storage; if you modify the contents of the pointer array during forms
1419 processing, all manner of bizarre things might happen. Also note that
1420 any given field may only be connected to one form.
1422 The functions free_field() and free_form are available to free field
1423 and form objects. It is an error to attempt to free a field connected
1424 to a form, but not vice-versa; thus, you will generally free your form
1427 Fetching and Changing Field Attributes
1429 Each form field has a number of location and size attributes
1430 associated with it. There are other field attributes used to control
1431 display and editing of the field. Some (for example, the O_STATIC bit)
1432 involve sufficient complications to be covered in sections of their
1433 own later on. We cover the functions used to get and set several basic
1436 When a field is created, the attributes not specified by the new_field
1437 function are copied from an invisible system default field. In
1438 attribute-setting and -fetching functions, the argument NULL is taken
1439 to mean this field. Changes to it persist as defaults until your forms
1440 application terminates.
1442 Fetching Size and Location Data
1444 You can retrieve field sizes and locations through:
1445 int field_info(FIELD *field, /* field from which to fetch */
1446 int *height, *int width, /* field size */
1447 int *top, int *left, /* upper left corner */
1448 int *offscreen, /* number of offscreen rows */
1449 int *nbuf); /* number of working buffers */
1451 This function is a sort of inverse of new_field(); instead of setting
1452 size and location attributes of a new field, it fetches them from an
1455 Changing the Field Location
1457 It is possible to move a field's location on the screen:
1458 int move_field(FIELD *field, /* field to alter */
1459 int top, int left); /* new upper-left corner */
1461 You can, of course. query the current location through field_info().
1463 The Justification Attribute
1465 One-line fields may be unjustified, justified right, justified left,
1466 or centered. Here is how you manipulate this attribute:
1467 int set_field_just(FIELD *field, /* field to alter */
1468 int justmode); /* mode to set */
1470 int field_just(FIELD *field); /* fetch mode of field */
1472 The mode values accepted and returned by this functions are
1473 preprocessor macros NO_JUSTIFICATION, JUSTIFY_RIGHT, JUSTIFY_LEFT, or
1476 Field Display Attributes
1478 For each field, you can set a foreground attribute for entered
1479 characters, a background attribute for the entire field, and a pad
1480 character for the unfilled portion of the field. You can also control
1481 pagination of the form.
1483 This group of four field attributes controls the visual appearance of
1484 the field on the screen, without affecting in any way the data in the
1486 int set_field_fore(FIELD *field, /* field to alter */
1487 chtype attr); /* attribute to set */
1489 chtype field_fore(FIELD *field); /* field to query */
1491 int set_field_back(FIELD *field, /* field to alter */
1492 chtype attr); /* attribute to set */
1494 chtype field_back(FIELD *field); /* field to query */
1496 int set_field_pad(FIELD *field, /* field to alter */
1497 int pad); /* pad character to set */
1499 chtype field_pad(FIELD *field);
1501 int set_new_page(FIELD *field, /* field to alter */
1502 int flag); /* TRUE to force new page */
1504 chtype new_page(FIELD *field); /* field to query */
1506 The attributes set and returned by the first four functions are normal
1507 curses(3x) display attribute values (A_STANDOUT, A_BOLD, A_REVERSE
1508 etc). The page bit of a field controls whether it is displayed at the
1509 start of a new form screen.
1513 There is also a large collection of field option bits you can set to
1514 control various aspects of forms processing. You can manipulate them
1515 with these functions:
1516 int set_field_opts(FIELD *field, /* field to alter */
1517 int attr); /* attribute to set */
1519 int field_opts_on(FIELD *field, /* field to alter */
1520 int attr); /* attributes to turn on */
1522 int field_opts_off(FIELD *field, /* field to alter */
1523 int attr); /* attributes to turn off */
1525 int field_opts(FIELD *field); /* field to query */
1527 By default, all options are on. Here are the available option bits:
1530 Controls whether the field is visible on the screen. Can be
1531 used during form processing to hide or pop up fields depending
1532 on the value of parent fields.
1535 Controls whether the field is active during forms processing
1536 (i.e. visited by form navigation keys). Can be used to make
1537 labels or derived fields with buffer values alterable by the
1538 forms application, not the user.
1541 Controls whether data is displayed during field entry. If this
1542 option is turned off on a field, the library will accept and
1543 edit data in that field, but it will not be displayed and the
1544 visible field cursor will not move. You can turn off the
1545 O_PUBLIC bit to define password fields.
1548 Controls whether the field's data can be modified. When this
1549 option is off, all editing requests except REQ_PREV_CHOICE and
1550 REQ_NEXT_CHOICE will fail. Such read-only fields may be useful
1554 Controls word-wrapping in multi-line fields. Normally, when any
1555 character of a (blank-separated) word reaches the end of the
1556 current line, the entire word is wrapped to the next line
1557 (assuming there is one). When this option is off, the word will
1558 be split across the line break.
1561 Controls field blanking. When this option is on, entering a
1562 character at the first field position erases the entire field
1563 (except for the just-entered character).
1566 Controls automatic skip to next field when this one fills.
1567 Normally, when the forms user tries to type more data into a
1568 field than will fit, the editing location jumps to next field.
1569 When this option is off, the user's cursor will hang at the end
1570 of the field. This option is ignored in dynamic fields that
1571 have not reached their size limit.
1574 Controls whether validation is applied to blank fields.
1575 Normally, it is not; the user can leave a field blank without
1576 invoking the usual validation check on exit. If this option is
1577 off on a field, exit from it will invoke a validation check.
1580 Controls whether validation occurs on every exit, or only after
1581 the field is modified. Normally the latter is true. Setting
1582 O_PASSOK may be useful if your field's validation function may
1583 change during forms processing.
1586 Controls whether the field is fixed to its initial dimensions.
1587 If you turn this off, the field becomes dynamic and will
1588 stretch to fit entered data.
1590 A field's options cannot be changed while the field is currently
1591 selected. However, options may be changed on posted fields that are
1594 The option values are bit-masks and can be composed with logical-or in
1599 Every field has a status flag, which is set to FALSE when the field is
1600 created and TRUE when the value in field buffer 0 changes. This flag
1601 can be queried and set directly:
1602 int set_field_status(FIELD *field, /* field to alter */
1603 int status); /* mode to set */
1605 int field_status(FIELD *field); /* fetch mode of field */
1607 Setting this flag under program control can be useful if you use the
1608 same form repeatedly, looking for modified fields each time.
1610 Calling field_status() on a field not currently selected for input
1611 will return a correct value. Calling field_status() on a field that is
1612 currently selected for input may not necessarily give a correct field
1613 status value, because entered data is not necessarily copied to buffer
1614 zero before the exit validation check. To guarantee that the returned
1615 status value reflects reality, call field_status() either (1) in the
1616 field's exit validation check routine, (2) from the field's or form's
1617 initialization or termination hooks, or (3) just after a
1618 REQ_VALIDATION request has been processed by the forms driver.
1622 Each field structure contains one character pointer slot that is not
1623 used by the forms library. It is intended to be used by applications
1624 to store private per-field data. You can manipulate it with:
1625 int set_field_userptr(FIELD *field, /* field to alter */
1626 char *userptr); /* mode to set */
1628 char *field_userptr(FIELD *field); /* fetch mode of field */
1630 (Properly, this user pointer field ought to have (void *) type. The
1631 (char *) type is retained for System V compatibility.)
1633 It is valid to set the user pointer of the default field (with a
1634 set_field_userptr() call passed a NULL field pointer.) When a new
1635 field is created, the default-field user pointer is copied to
1636 initialize the new field's user pointer.
1638 Variable-Sized Fields
1640 Normally, a field is fixed at the size specified for it at creation
1641 time. If, however, you turn off its O_STATIC bit, it becomes dynamic
1642 and will automatically resize itself to accommodate data as it is
1643 entered. If the field has extra buffers associated with it, they will
1644 grow right along with the main input buffer.
1646 A one-line dynamic field will have a fixed height (1) but variable
1647 width, scrolling horizontally to display data within the field area as
1648 originally dimensioned and located. A multi-line dynamic field will
1649 have a fixed width, but variable height (number of rows), scrolling
1650 vertically to display data within the field area as originally
1651 dimensioned and located.
1653 Normally, a dynamic field is allowed to grow without limit. But it is
1654 possible to set an upper limit on the size of a dynamic field. You do
1655 it with this function:
1656 int set_max_field(FIELD *field, /* field to alter (may not be NULL) */
1657 int max_size); /* upper limit on field size */
1659 If the field is one-line, max_size is taken to be a column size limit;
1660 if it is multi-line, it is taken to be a line size limit. To disable
1661 any limit, use an argument of zero. The growth limit can be changed
1662 whether or not the O_STATIC bit is on, but has no effect until it is.
1664 The following properties of a field change when it becomes dynamic:
1665 * If there is no growth limit, there is no final position of the
1666 field; therefore O_AUTOSKIP and O_NL_OVERLOAD are ignored.
1667 * Field justification will be ignored (though whatever justification
1668 is set up will be retained internally and can be queried).
1669 * The dup_field() and link_field() calls copy dynamic-buffer sizes.
1670 If the O_STATIC option is set on one of a collection of links,
1671 buffer resizing will occur only when the field is edited through
1673 * The call field_info() will retrieve the original static size of
1674 the field; use dynamic_field_info() to get the actual dynamic
1679 By default, a field will accept any data that will fit in its input
1680 buffer. However, it is possible to attach a validation type to a
1681 field. If you do this, any attempt to leave the field while it
1682 contains data that does not match the validation type will fail. Some
1683 validation types also have a character-validity check for each time a
1684 character is entered in the field.
1686 A field's validation check (if any) is not called when
1687 set_field_buffer() modifies the input buffer, nor when that buffer is
1688 changed through a linked field.
1690 The form library provides a rich set of pre-defined validation types,
1691 and gives you the capability to define custom ones of your own. You
1692 can examine and change field validation attributes with the following
1694 int set_field_type(FIELD *field, /* field to alter */
1695 FIELDTYPE *ftype, /* type to associate */
1696 ...); /* additional arguments*/
1698 FIELDTYPE *field_type(FIELD *field); /* field to query */
1700 The validation type of a field is considered an attribute of the
1701 field. As with other field attributes, Also, doing set_field_type()
1702 with a NULL field default will change the system default for
1703 validation of newly-created fields.
1705 Here are the pre-defined validation types:
1709 This field type accepts alphabetic data; no blanks, no digits, no
1710 special characters (this is checked at character-entry time). It is
1712 int set_field_type(FIELD *field, /* field to alter */
1713 TYPE_ALPHA, /* type to associate */
1714 int width); /* maximum width of field */
1716 The width argument sets a minimum width of data. Typically you will
1717 want to set this to the field width; if it is greater than the field
1718 width, the validation check will always fail. A minimum width of zero
1719 makes field completion optional.
1723 This field type accepts alphabetic data and digits; no blanks, no
1724 special characters (this is checked at character-entry time). It is
1726 int set_field_type(FIELD *field, /* field to alter */
1727 TYPE_ALNUM, /* type to associate */
1728 int width); /* maximum width of field */
1730 The width argument sets a minimum width of data. As with TYPE_ALPHA,
1731 typically you will want to set this to the field width; if it is
1732 greater than the field width, the validation check will always fail. A
1733 minimum width of zero makes field completion optional.
1737 This type allows you to restrict a field's values to be among a
1738 specified set of string values (for example, the two-letter postal
1739 codes for U.S. states). It is set up with:
1740 int set_field_type(FIELD *field, /* field to alter */
1741 TYPE_ENUM, /* type to associate */
1742 char **valuelist; /* list of possible values */
1743 int checkcase; /* case-sensitive? */
1744 int checkunique); /* must specify uniquely? */
1746 The valuelist parameter must point at a NULL-terminated list of valid
1747 strings. The checkcase argument, if true, makes comparison with the
1748 string case-sensitive.
1750 When the user exits a TYPE_ENUM field, the validation procedure tries
1751 to complete the data in the buffer to a valid entry. If a complete
1752 choice string has been entered, it is of course valid. But it is also
1753 possible to enter a prefix of a valid string and have it completed for
1756 By default, if you enter such a prefix and it matches more than one
1757 value in the string list, the prefix will be completed to the first
1758 matching value. But the checkunique argument, if true, requires prefix
1759 matches to be unique in order to be valid.
1761 The REQ_NEXT_CHOICE and REQ_PREV_CHOICE input requests can be
1762 particularly useful with these fields.
1766 This field type accepts an integer. It is set up as follows:
1767 int set_field_type(FIELD *field, /* field to alter */
1768 TYPE_INTEGER, /* type to associate */
1769 int padding, /* # places to zero-pad to */
1770 int vmin, int vmax); /* valid range */
1772 Valid characters consist of an optional leading minus and digits. The
1773 range check is performed on exit. If the range maximum is less than or
1774 equal to the minimum, the range is ignored.
1776 If the value passes its range check, it is padded with as many leading
1777 zero digits as necessary to meet the padding argument.
1779 A TYPE_INTEGER value buffer can conveniently be interpreted with the C
1780 library function atoi(3).
1784 This field type accepts a decimal number. It is set up as follows:
1785 int set_field_type(FIELD *field, /* field to alter */
1786 TYPE_NUMERIC, /* type to associate */
1787 int padding, /* # places of precision */
1788 double vmin, double vmax); /* valid range */
1790 Valid characters consist of an optional leading minus and digits.
1791 possibly including a decimal point. If your system supports locale's,
1792 the decimal point character used must be the one defined by your
1793 locale. The range check is performed on exit. If the range maximum is
1794 less than or equal to the minimum, the range is ignored.
1796 If the value passes its range check, it is padded with as many
1797 trailing zero digits as necessary to meet the padding argument.
1799 A TYPE_NUMERIC value buffer can conveniently be interpreted with the C
1800 library function atof(3).
1804 This field type accepts data matching a regular expression. It is set
1806 int set_field_type(FIELD *field, /* field to alter */
1807 TYPE_REGEXP, /* type to associate */
1808 char *regexp); /* expression to match */
1810 The syntax for regular expressions is that of regcomp(3). The check
1811 for regular-expression match is performed on exit.
1813 Direct Field Buffer Manipulation
1815 The chief attribute of a field is its buffer contents. When a form has
1816 been completed, your application usually needs to know the state of
1817 each field buffer. You can find this out with:
1818 char *field_buffer(FIELD *field, /* field to query */
1819 int bufindex); /* number of buffer to query */
1821 Normally, the state of the zero-numbered buffer for each field is set
1822 by the user's editing actions on that field. It is sometimes useful to
1823 be able to set the value of the zero-numbered (or some other) buffer
1824 from your application:
1825 int set_field_buffer(FIELD *field, /* field to alter */
1826 int bufindex, /* number of buffer to alter */
1827 char *value); /* string value to set */
1829 If the field is not large enough and cannot be resized to a
1830 sufficiently large size to contain the specified value, the value will
1831 be truncated to fit.
1833 Calling field_buffer() with a null field pointer will raise an error.
1834 Calling field_buffer() on a field not currently selected for input
1835 will return a correct value. Calling field_buffer() on a field that is
1836 currently selected for input may not necessarily give a correct field
1837 buffer value, because entered data is not necessarily copied to buffer
1838 zero before the exit validation check. To guarantee that the returned
1839 buffer value reflects on-screen reality, call field_buffer() either
1840 (1) in the field's exit validation check routine, (2) from the field's
1841 or form's initialization or termination hooks, or (3) just after a
1842 REQ_VALIDATION request has been processed by the forms driver.
1846 As with field attributes, form attributes inherit a default from a
1847 system default form structure. These defaults can be queried or set by
1848 of these functions using a form-pointer argument of NULL.
1850 The principal attribute of a form is its field list. You can query and
1851 change this list with:
1852 int set_form_fields(FORM *form, /* form to alter */
1853 FIELD **fields); /* fields to connect */
1855 char *form_fields(FORM *form); /* fetch fields of form */
1857 int field_count(FORM *form); /* count connect fields */
1859 The second argument of set_form_fields() may be a NULL-terminated
1860 field pointer array like the one required by new_form(). In that case,
1861 the old fields of the form are disconnected but not freed (and
1862 eligible to be connected to other forms), then the new fields are
1865 It may also be null, in which case the old fields are disconnected
1866 (and not freed) but no new ones are connected.
1868 The field_count() function simply counts the number of fields
1869 connected to a given from. It returns -1 if the form-pointer argument
1872 Control of Form Display
1874 In the overview section, you saw that to display a form you normally
1875 start by defining its size (and fields), posting it, and refreshing
1876 the screen. There is an hidden step before posting, which is the
1877 association of the form with a frame window (actually, a pair of
1878 windows) within which it will be displayed. By default, the forms
1879 library associates every form with the full-screen window stdscr.
1881 By making this step explicit, you can associate a form with a declared
1882 frame window on your screen display. This can be useful if you want to
1883 adapt the form display to different screen sizes, dynamically tile
1884 forms on the screen, or use a form as part of an interface layout
1887 The two windows associated with each form have the same functions as
1888 their analogues in the menu library. Both these windows are painted
1889 when the form is posted and erased when the form is unposted.
1891 The outer or frame window is not otherwise touched by the form
1892 routines. It exists so the programmer can associate a title, a border,
1893 or perhaps help text with the form and have it properly refreshed or
1894 erased at post/unpost time. The inner window or subwindow is where the
1895 current form page is actually displayed.
1897 In order to declare your own frame window for a form, you will need to
1898 know the size of the form's bounding rectangle. You can get this
1900 int scale_form(FORM *form, /* form to query */
1901 int *rows, /* form rows */
1902 int *cols); /* form cols */
1904 The form dimensions are passed back in the locations pointed to by the
1905 arguments. Once you have this information, you can use it to declare
1906 of windows, then use one of these functions:
1907 int set_form_win(FORM *form, /* form to alter */
1908 WINDOW *win); /* frame window to connect */
1910 WINDOW *form_win(FORM *form); /* fetch frame window of form */
1912 int set_form_sub(FORM *form, /* form to alter */
1913 WINDOW *win); /* form subwindow to connect */
1915 WINDOW *form_sub(FORM *form); /* fetch form subwindow of form */
1917 Note that curses operations, including refresh(), on the form, should
1918 be done on the frame window, not the form subwindow.
1920 It is possible to check from your application whether all of a
1921 scrollable field is actually displayed within the menu subwindow. Use
1923 int data_ahead(FORM *form); /* form to be queried */
1925 int data_behind(FORM *form); /* form to be queried */
1927 The function data_ahead() returns TRUE if (a) the current field is
1928 one-line and has undisplayed data off to the right, (b) the current
1929 field is multi-line and there is data off-screen below it.
1931 The function data_behind() returns TRUE if the first (upper left hand)
1932 character position is off-screen (not being displayed).
1934 Finally, there is a function to restore the form window's cursor to
1935 the value expected by the forms driver:
1936 int pos_form_cursor(FORM *) /* form to be queried */
1938 If your application changes the form window cursor, call this function
1939 before handing control back to the forms driver in order to
1942 Input Processing in the Forms Driver
1944 The function form_driver() handles virtualized input requests for form
1945 navigation, editing, and validation requests, just as menu_driver does
1946 for menus (see the section on menu input handling).
1947 int form_driver(FORM *form, /* form to pass input to */
1948 int request); /* form request code */
1950 Your input virtualization function needs to take input and then
1951 convert it to either an alphanumeric character (which is treated as
1952 data to be entered in the currently-selected field), or a forms
1955 The forms driver provides hooks (through input-validation and
1956 field-termination functions) with which your application code can
1957 check that the input taken by the driver matched what was expected.
1959 Page Navigation Requests
1961 These requests cause page-level moves through the form, triggering
1962 display of a new form screen.
1965 Move to the next form page.
1968 Move to the previous form page.
1971 Move to the first form page.
1974 Move to the last form page.
1976 These requests treat the list as cyclic; that is, REQ_NEXT_PAGE from
1977 the last page goes to the first, and REQ_PREV_PAGE from the first page
1980 Inter-Field Navigation Requests
1982 These requests handle navigation between fields on the same page.
1988 Move to previous field.
1991 Move to the first field.
1994 Move to the last field.
1997 Move to sorted next field.
2000 Move to sorted previous field.
2003 Move to the sorted first field.
2006 Move to the sorted last field.
2012 Move right to field.
2020 These requests treat the list of fields on a page as cyclic; that is,
2021 REQ_NEXT_FIELD from the last field goes to the first, and
2022 REQ_PREV_FIELD from the first field goes to the last. The order of the
2023 fields for these (and the REQ_FIRST_FIELD and REQ_LAST_FIELD requests)
2024 is simply the order of the field pointers in the form array (as set up
2025 by new_form() or set_form_fields()
2027 It is also possible to traverse the fields as if they had been sorted
2028 in screen-position order, so the sequence goes left-to-right and
2029 top-to-bottom. To do this, use the second group of four
2030 sorted-movement requests.
2032 Finally, it is possible to move between fields using visual directions
2033 up, down, right, and left. To accomplish this, use the third group of
2034 four requests. Note, however, that the position of a form for purposes
2035 of these requests is its upper-left corner.
2037 For example, suppose you have a multi-line field B, and two
2038 single-line fields A and C on the same line with B, with A to the left
2039 of B and C to the right of B. A REQ_MOVE_RIGHT from A will go to B
2040 only if A, B, and C all share the same first line; otherwise it will
2043 Intra-Field Navigation Requests
2045 These requests drive movement of the edit cursor within the currently
2049 Move to next character.
2052 Move to previous character.
2058 Move to previous line.
2064 Move to previous word.
2067 Move to beginning of field.
2070 Move to end of field.
2073 Move to beginning of line.
2076 Move to end of line.
2082 Move right in field.
2090 Each word is separated from the previous and next characters by
2091 whitespace. The commands to move to beginning and end of line or field
2092 look for the first or last non-pad character in their ranges.
2096 Fields that are dynamic and have grown and fields explicitly created
2097 with offscreen rows are scrollable. One-line fields scroll
2098 horizontally; multi-line fields scroll vertically. Most scrolling is
2099 triggered by editing and intra-field movement (the library scrolls the
2100 field to keep the cursor visible). It is possible to explicitly
2101 request scrolling with the following requests:
2104 Scroll vertically forward a line.
2107 Scroll vertically backward a line.
2110 Scroll vertically forward a page.
2113 Scroll vertically backward a page.
2116 Scroll vertically forward half a page.
2119 Scroll vertically backward half a page.
2122 Scroll horizontally forward a character.
2125 Scroll horizontally backward a character.
2128 Scroll horizontally one field width forward.
2131 Scroll horizontally one field width backward.
2134 Scroll horizontally one half field width forward.
2137 Scroll horizontally one half field width backward.
2139 For scrolling purposes, a page of a field is the height of its visible
2144 When you pass the forms driver an ASCII character, it is treated as a
2145 request to add the character to the field's data buffer. Whether this
2146 is an insertion or a replacement depends on the field's edit mode
2147 (insertion is the default.
2149 The following requests support editing the field and changing the edit
2159 New line request (see below for explanation).
2162 Insert space at character location.
2165 Insert blank line at character location.
2168 Delete character at cursor.
2171 Delete previous word at cursor.
2174 Delete line at cursor.
2177 Delete word at cursor.
2180 Clear to end of line.
2183 Clear to end of field.
2188 The behavior of the REQ_NEW_LINE and REQ_DEL_PREV requests is
2189 complicated and partly controlled by a pair of forms options. The
2190 special cases are triggered when the cursor is at the beginning of a
2191 field, or on the last line of the field.
2193 First, we consider REQ_NEW_LINE:
2195 The normal behavior of REQ_NEW_LINE in insert mode is to break the
2196 current line at the position of the edit cursor, inserting the portion
2197 of the current line after the cursor as a new line following the
2198 current and moving the cursor to the beginning of that new line (you
2199 may think of this as inserting a newline in the field buffer).
2201 The normal behavior of REQ_NEW_LINE in overlay mode is to clear the
2202 current line from the position of the edit cursor to end of line. The
2203 cursor is then moved to the beginning of the next line.
2205 However, REQ_NEW_LINE at the beginning of a field, or on the last line
2206 of a field, instead does a REQ_NEXT_FIELD. O_NL_OVERLOAD option is
2207 off, this special action is disabled.
2209 Now, let us consider REQ_DEL_PREV:
2211 The normal behavior of REQ_DEL_PREV is to delete the previous
2212 character. If insert mode is on, and the cursor is at the start of a
2213 line, and the text on that line will fit on the previous one, it
2214 instead appends the contents of the current line to the previous one
2215 and deletes the current line (you may think of this as deleting a
2216 newline from the field buffer).
2218 However, REQ_DEL_PREV at the beginning of a field is instead treated
2219 as a REQ_PREV_FIELD.
2221 If the O_BS_OVERLOAD option is off, this special action is disabled
2222 and the forms driver just returns E_REQUEST_DENIED.
2224 See Form Options for discussion of how to set and clear the overload
2229 If the type of your field is ordered, and has associated functions for
2230 getting the next and previous values of the type from a given value,
2231 there are requests that can fetch that value into the field buffer:
2234 Place the successor value of the current value in the buffer.
2237 Place the predecessor value of the current value in the buffer.
2239 Of the built-in field types, only TYPE_ENUM has built-in successor and
2240 predecessor functions. When you define a field type of your own (see
2241 Custom Validation Types), you can associate our own ordering
2244 Application Commands
2246 Form requests are represented as integers above the curses value
2247 greater than KEY_MAX and less than or equal to the constant
2248 MAX_COMMAND. If your input-virtualization routine returns a value
2249 above MAX_COMMAND, the forms driver will ignore it.
2253 It is possible to set function hooks to be executed whenever the
2254 current field or form changes. Here are the functions that support
2256 typedef void (*HOOK)(); /* pointer to function returning void */
2258 int set_form_init(FORM *form, /* form to alter */
2259 HOOK hook); /* initialization hook */
2261 HOOK form_init(FORM *form); /* form to query */
2263 int set_form_term(FORM *form, /* form to alter */
2264 HOOK hook); /* termination hook */
2266 HOOK form_term(FORM *form); /* form to query */
2268 int set_field_init(FORM *form, /* form to alter */
2269 HOOK hook); /* initialization hook */
2271 HOOK field_init(FORM *form); /* form to query */
2273 int set_field_term(FORM *form, /* form to alter */
2274 HOOK hook); /* termination hook */
2276 HOOK field_term(FORM *form); /* form to query */
2278 These functions allow you to either set or query four different hooks.
2279 In each of the set functions, the second argument should be the
2280 address of a hook function. These functions differ only in the timing
2284 This hook is called when the form is posted; also, just after
2285 each page change operation.
2288 This hook is called when the form is posted; also, just after
2292 This hook is called just after field validation; that is, just
2293 before the field is altered. It is also called when the form is
2297 This hook is called when the form is unposted; also, just
2298 before each page change operation.
2300 Calls to these hooks may be triggered
2301 1. When user editing requests are processed by the forms driver
2302 2. When the current page is changed by set_current_field() call
2303 3. When the current field is changed by a set_form_page() call
2305 See Field Change Commands for discussion of the latter two cases.
2307 You can set a default hook for all fields by passing one of the set
2308 functions a NULL first argument.
2310 You can disable any of these hooks by (re)setting them to NULL, the
2313 Field Change Commands
2315 Normally, navigation through the form will be driven by the user's
2316 input requests. But sometimes it is useful to be able to move the
2317 focus for editing and viewing under control of your application, or
2318 ask which field it currently is in. The following functions help you
2320 int set_current_field(FORM *form, /* form to alter */
2321 FIELD *field); /* field to shift to */
2323 FIELD *current_field(FORM *form); /* form to query */
2325 int field_index(FORM *form, /* form to query */
2326 FIELD *field); /* field to get index of */
2328 The function field_index() returns the index of the given field in the
2329 given form's field array (the array passed to new_form() or
2332 The initial current field of a form is the first active field on the
2333 first page. The function set_form_fields() resets this.
2335 It is also possible to move around by pages.
2336 int set_form_page(FORM *form, /* form to alter */
2337 int page); /* page to go to (0-origin) */
2339 int form_page(FORM *form); /* return form's current page */
2341 The initial page of a newly-created form is 0. The function
2342 set_form_fields() resets this.
2346 Like fields, forms may have control option bits. They can be changed
2347 or queried with these functions:
2348 int set_form_opts(FORM *form, /* form to alter */
2349 int attr); /* attribute to set */
2351 int form_opts_on(FORM *form, /* form to alter */
2352 int attr); /* attributes to turn on */
2354 int form_opts_off(FORM *form, /* form to alter */
2355 int attr); /* attributes to turn off */
2357 int form_opts(FORM *form); /* form to query */
2359 By default, all options are on. Here are the available option bits:
2362 Enable overloading of REQ_NEW_LINE as described in Editing
2363 Requests. The value of this option is ignored on dynamic fields
2364 that have not reached their size limit; these have no last
2365 line, so the circumstances for triggering a REQ_NEXT_FIELD
2369 Enable overloading of REQ_DEL_PREV as described in Editing
2372 The option values are bit-masks and can be composed with logical-or in
2375 Custom Validation Types
2377 The form library gives you the capability to define custom validation
2378 types of your own. Further, the optional additional arguments of
2379 set_field_type effectively allow you to parameterize validation types.
2380 Most of the complications in the validation-type interface have to do
2381 with the handling of the additional arguments within custom validation
2386 The simplest way to create a custom data type is to compose it from
2387 two preexisting ones:
2388 FIELD *link_fieldtype(FIELDTYPE *type1,
2391 This function creates a field type that will accept any of the values
2392 legal for either of its argument field types (which may be either
2393 predefined or programmer-defined). If a set_field_type() call later
2394 requires arguments, the new composite type expects all arguments for
2395 the first type, than all arguments for the second. Order functions
2396 (see Order Requests) associated with the component types will work on
2397 the composite; what it does is check the validation function for the
2398 first type, then for the second, to figure what type the buffer
2399 contents should be treated as.
2403 To create a field type from scratch, you need to specify one or both
2404 of the following things:
2405 * A character-validation function, to check each character as it is
2407 * A field-validation function to be applied on exit from the field.
2409 Here is how you do that:
2410 typedef int (*HOOK)(); /* pointer to function returning int */
2412 FIELDTYPE *new_fieldtype(HOOK f_validate, /* field validator */
2413 HOOK c_validate) /* character validator */
2415 int free_fieldtype(FIELDTYPE *ftype); /* type to free */
2417 At least one of the arguments of new_fieldtype() must be non-NULL. The
2418 forms driver will automatically call the new type's validation
2419 functions at appropriate points in processing a field of the new type.
2421 The function free_fieldtype() deallocates the argument fieldtype,
2422 freeing all storage associated with it.
2424 Normally, a field validator is called when the user attempts to leave
2425 the field. Its first argument is a field pointer, from which it can
2426 get to field buffer 0 and test it. If the function returns TRUE, the
2427 operation succeeds; if it returns FALSE, the edit cursor stays in the
2430 A character validator gets the character passed in as a first
2431 argument. It too should return TRUE if the character is valid, FALSE
2434 Validation Function Arguments
2436 Your field- and character- validation functions will be passed a
2437 second argument as well. This second argument is the address of a
2438 structure (which we will call a pile) built from any of the
2439 field-type-specific arguments passed to set_field_type(). If no such
2440 arguments are defined for the field type, this pile pointer argument
2443 In order to arrange for such arguments to be passed to your validation
2444 functions, you must associate a small set of storage-management
2445 functions with the type. The forms driver will use these to synthesize
2446 a pile from the trailing arguments of each set_field_type() argument,
2447 and a pointer to the pile will be passed to the validation functions.
2449 Here is how you make the association:
2450 typedef char *(*PTRHOOK)(); /* pointer to function returning (char *) */
2451 typedef void (*VOIDHOOK)(); /* pointer to function returning void */
2453 int set_fieldtype_arg(FIELDTYPE *type, /* type to alter */
2454 PTRHOOK make_str, /* make structure from args */
2455 PTRHOOK copy_str, /* make copy of structure */
2456 VOIDHOOK free_str); /* free structure storage */
2458 Here is how the storage-management hooks are used:
2461 This function is called by set_field_type(). It gets one
2462 argument, a va_list of the type-specific arguments passed to
2463 set_field_type(). It is expected to return a pile pointer to a
2464 data structure that encapsulates those arguments.
2467 This function is called by form library functions that allocate
2468 new field instances. It is expected to take a pile pointer,
2469 copy the pile to allocated storage, and return the address of
2473 This function is called by field- and type-deallocation
2474 routines in the library. It takes a pile pointer argument, and
2475 is expected to free the storage of that pile.
2477 The make_str and copy_str functions may return NULL to signal
2478 allocation failure. The library routines will that call them will
2479 return error indication when this happens. Thus, your validation
2480 functions should never see a NULL file pointer and need not check
2483 Order Functions For Custom Types
2485 Some custom field types are simply ordered in the same well-defined
2486 way that TYPE_ENUM is. For such types, it is possible to define
2487 successor and predecessor functions to support the REQ_NEXT_CHOICE and
2488 REQ_PREV_CHOICE requests. Here is how:
2489 typedef int (*INTHOOK)(); /* pointer to function returning int */
2491 int set_fieldtype_arg(FIELDTYPE *type, /* type to alter */
2492 INTHOOK succ, /* get successor value */
2493 INTHOOK pred); /* get predecessor value */
2495 The successor and predecessor arguments will each be passed two
2496 arguments; a field pointer, and a pile pointer (as for the validation
2497 functions). They are expected to use the function field_buffer() to
2498 read the current value, and set_field_buffer() on buffer 0 to set the
2499 next or previous value. Either hook may return TRUE to indicate
2500 success (a legal next or previous value was set) or FALSE to indicate
2505 The interface for defining custom types is complicated and tricky.
2506 Rather than attempting to create a custom type entirely from scratch,
2507 you should start by studying the library source code for whichever of
2508 the pre-defined types seems to be closest to what you want.
2510 Use that code as a model, and evolve it towards what you really want.
2511 You will avoid many problems and annoyances that way. The code in the
2512 ncurses library has been specifically exempted from the package
2513 copyright to support this.
2515 If your custom type defines order functions, have do something
2516 intuitive with a blank field. A useful convention is to make the
2517 successor of a blank field the types minimum value, and its
2518 predecessor the maximum.