Gitiles
Code Review Sign In
gerrit.omnirom.org / android_external_libncurses / refs/heads/android-16 / . / doc / html / ncurses-intro.html
blob: 194850246fa63ffed5b4dab54aeb3b79d79638b7 [file] [log] [blame]
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1<!--
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2 $Id: ncurses-intro.html,v 1.57 2022/11/26 19:33:46 tom Exp $
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3 ****************************************************************************
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]4 * Copyright 2019-2020,2022 Thomas E. Dickey *
5 * Copyright 2000-2013,2017 Free Software Foundation, Inc. *
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]6 * *
7 * Permission is hereby granted, free of charge, to any person obtaining a *
8 * copy of this software and associated documentation files (the *
9 * "Software"), to deal in the Software without restriction, including *
10 * without limitation the rights to use, copy, modify, merge, publish, *
11 * distribute, distribute with modifications, sublicense, and/or sell *
12 * copies of the Software, and to permit persons to whom the Software is *
13 * furnished to do so, subject to the following conditions: *
14 * *
15 * The above copyright notice and this permission notice shall be included *
16 * in all copies or substantial portions of the Software. *
17 * *
18 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS *
19 * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF *
20 * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. *
21 * IN NO EVENT SHALL THE ABOVE COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, *
22 * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR *
23 * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR *
24 * THE USE OR OTHER DEALINGS IN THE SOFTWARE. *
25 * *
26 * Except as contained in this notice, the name(s) of the above copyright *
27 * holders shall not be used in advertising or otherwise to promote the *
28 * sale, use or other dealings in this Software without prior written *
29 * authorization. *
30 ****************************************************************************
31-->
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]32<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01//EN">
33<html>
34<head>
35 <meta name="generator" content=
36 "HTML Tidy for HTML5 for Linux version 5.6.0">
37 <title>Writing Programs with NCURSES</title>
38 <link rel="author" href="mailto:bugs-ncurses@gnu.org">
39 <meta http-equiv="Content-Type" content=
40 "text/html; charset=us-ascii">
41</head>
42<body>
43 <h1 class="no-header">Writing Programs with NCURSES</h1>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]44
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]45 <h2>Writing Programs with NCURSES</h2>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]46
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]47 <blockquote>
48 by Eric S. Raymond and Zeyd M. Ben-Halim<br>
49 updates since release 1.9.9e by Thomas Dickey
50 </blockquote>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]51
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]52 <div class="nav">
53 <h2>Contents</h2>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]54
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]55 <ul>
56 <li>
57 <a href="#introduction">Introduction</a>
58 <ul>
59 <li><a href="#history">A Brief History of Curses</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]60
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]61 <li><a href="#scope">Scope of This Document</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]62
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]63 <li><a href="#terminology">Terminology</a></li>
64 </ul>
65 </li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]66
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]67 <li>
68 <a href="#curses">The Curses Library</a>
69 <ul>
70 <li>
71 <a href="#overview">An Overview of Curses</a>
72 <ul>
73 <li><a href="#compiling">Compiling Programs using
74 Curses</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]75
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]76 <li><a href="#updating">Updating the Screen</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]77
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]78 <li><a href="#stdscr">Standard Windows and Function
79 Naming Conventions</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]80
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]81 <li><a href="#variables">Variables</a></li>
82 </ul>
83 </li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]84
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]85 <li>
86 <a href="#using">Using the Library</a>
87 <ul>
88 <li><a href="#starting">Starting up</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]89
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]90 <li><a href="#output">Output</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]91
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]92 <li><a href="#input">Input</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]93
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]94 <li><a href="#formschars">Using Forms Characters</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]95
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]96 <li><a href="#attributes">Character Attributes and
97 Color</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]98
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]99 <li><a href="#mouse">Mouse Interfacing</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]100
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]101 <li><a href="#finishing">Finishing Up</a></li>
102 </ul>
103 </li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]104
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]105 <li>
106 <a href="#functions">Function Descriptions</a>
107 <ul>
108 <li><a href="#init">Initialization and Wrapup</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]109
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]110 <li><a href="#flush">Causing Output to the
111 Terminal</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]112
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]113 <li><a href="#lowlevel">Low-Level Capability
114 Access</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]115
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]116 <li><a href="#debugging">Debugging</a></li>
117 </ul>
118 </li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]119
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]120 <li>
121 <a href="#hints">Hints, Tips, and Tricks</a>
122 <ul>
123 <li><a href="#caution">Some Notes of Caution</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]124
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]125 <li><a href="#leaving">Temporarily Leaving ncurses
126 Mode</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]127
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]128 <li><a href="#xterm">Using <code>ncurses</code> under
129 <code>xterm</code></a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]130
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]131 <li><a href="#screens">Handling Multiple Terminal
132 Screens</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]133
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]134 <li><a href="#testing">Testing for Terminal
135 Capabilities</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]136
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]137 <li><a href="#tuning">Tuning for Speed</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]138
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]139 <li><a href="#special">Special Features of
140 <code>ncurses</code></a></li>
141 </ul>
142 </li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]143
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]144 <li>
145 <a href="#compat">Compatibility with Older Versions</a>
146 <ul>
147 <li><a href="#refbug">Refresh of Overlapping
148 Windows</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]149
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]150 <li><a href="#backbug">Background Erase</a></li>
151 </ul>
152 </li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]153
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]154 <li><a href="#xsifuncs">XSI Curses Conformance</a></li>
155 </ul>
156 </li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]157
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]158 <li>
159 <a href="#panels">The Panels Library</a>
160 <ul>
161 <li><a href="#pcompile">Compiling With the Panels
162 Library</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]163
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]164 <li><a href="#poverview">Overview of Panels</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]165
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]166 <li><a href="#pstdscr">Panels, Input, and the Standard
167 Screen</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]168
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]169 <li><a href="#hiding">Hiding Panels</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]170
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]171 <li><a href="#pmisc">Miscellaneous Other Facilities</a></li>
172 </ul>
173 </li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]174
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]175 <li>
176 <a href="#menu">The Menu Library</a>
177 <ul>
178 <li><a href="#mcompile">Compiling with the menu
179 Library</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]180
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]181 <li><a href="#moverview">Overview of Menus</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]182
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]183 <li><a href="#mselect">Selecting items</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]184
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]185 <li><a href="#mdisplay">Menu Display</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]186
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]187 <li><a href="#mwindows">Menu Windows</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]188
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]189 <li><a href="#minput">Processing Menu Input</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]190
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]191 <li><a href="#mmisc">Miscellaneous Other Features</a></li>
192 </ul>
193 </li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]194
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]195 <li>
196 <a href="#form">The Forms Library</a>
197 <ul>
198 <li><a href="#fcompile">Compiling with the forms
199 Library</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]200
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]201 <li><a href="#foverview">Overview of Forms</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]202
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]203 <li><a href="#fcreate">Creating and Freeing Fields and
204 Forms</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]205
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]206 <li>
207 <a href="#fattributes">Fetching and Changing Field
208 Attributes</a>
209 <ul>
210 <li><a href="#fsizes">Fetching Size and Location
211 Data</a></li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]212
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]213 <li><a href="#flocation">Changing the Field
214 Location</a></li>
215
216 <li><a href="#fjust">The Justification Attribute</a></li>
217
218 <li><a href="#fdispatts">Field Display Attributes</a></li>
219
220 <li><a href="#foptions">Field Option Bits</a></li>
221
222 <li><a href="#fstatus">Field Status</a></li>
223
224 <li><a href="#fuser">Field User Pointer</a></li>
225 </ul>
226 </li>
227
228 <li><a href="#fdynamic">Variable-Sized Fields</a></li>
229
230 <li>
231 <a href="#fvalidation">Field Validation</a>
232 <ul>
233 <li><a href="#ftype_alpha">TYPE_ALPHA</a></li>
234
235 <li><a href="#ftype_alnum">TYPE_ALNUM</a></li>
236
237 <li><a href="#ftype_enum">TYPE_ENUM</a></li>
238
239 <li><a href="#ftype_integer">TYPE_INTEGER</a></li>
240
241 <li><a href="#ftype_numeric">TYPE_NUMERIC</a></li>
242
243 <li><a href="#ftype_regexp">TYPE_REGEXP</a></li>
244 </ul>
245 </li>
246
247 <li><a href="#fbuffer">Direct Field Buffer
248 Manipulation</a></li>
249
250 <li><a href="#formattrs">Attributes of Forms</a></li>
251
252 <li><a href="#fdisplay">Control of Form Display</a></li>
253
254 <li>
255 <a href="#fdriver">Input Processing in the Forms
256 Driver</a>
257 <ul>
258 <li><a href="#fpage">Page Navigation Requests</a></li>
259
260 <li><a href="#ffield">Inter-Field Navigation
261 Requests</a></li>
262
263 <li><a href="#fifield">Intra-Field Navigation
264 Requests</a></li>
265
266 <li><a href="#fscroll">Scrolling Requests</a></li>
267
268 <li><a href="#fedit">Field Editing Requests</a></li>
269
270 <li><a href="#forder">Order Requests</a></li>
271
272 <li><a href="#fappcmds">Application Commands</a></li>
273 </ul>
274 </li>
275
276 <li><a href="#fhooks">Field Change Hooks</a></li>
277
278 <li><a href="#ffocus">Field Change Commands</a></li>
279
280 <li><a href="#frmoptions">Form Options</a></li>
281
282 <li>
283 <a href="#fcustom">Custom Validation Types</a>
284 <ul>
285 <li><a href="#flinktypes">Union Types</a></li>
286
287 <li><a href="#fnewtypes">New Field Types</a></li>
288
289 <li><a href="#fcheckargs">Validation Function
290 Arguments</a></li>
291
292 <li><a href="#fcustorder">Order Functions For Custom
293 Types</a></li>
294
295 <li><a href="#fcustprobs">Avoiding Problems</a></li>
296 </ul>
297 </li>
298 </ul>
299 </li>
300 </ul>
301 </div>
302
303 <hr>
304
305 <h2><a name="introduction" id="introduction">Introduction</a></h2>
306
307 <p>This document is an introduction to programming with
308 <code>curses</code>. It is not an exhaustive reference for the
309 curses Application Programming Interface (API); that role is
310 filled by the <code>curses</code> manual pages. Rather, it is
311 intended to help C programmers ease into using the package.</p>
312
313 <p>This document is aimed at C applications programmers not yet
314 specifically familiar with ncurses. If you are already an
315 experienced <code>curses</code> programmer, you should
316 nevertheless read the sections on <a href="#mouse">Mouse
317 Interfacing</a>, <a href="#debugging">Debugging</a>, <a href=
318 "#compat">Compatibility with Older Versions</a>, and <a href=
319 "#hints">Hints, Tips, and Tricks</a>. These will bring you up to
320 speed on the special features and quirks of the
321 <code>ncurses</code> implementation. If you are not so
322 experienced, keep reading.</p>
323
324 <p>The <code>curses</code> package is a subroutine library for
325 terminal-independent screen-painting and input-event handling
326 which presents a high level screen model to the programmer,
327 hiding differences between terminal types and doing automatic
328 optimization of output to change one screen full of text into
329 another. <code>Curses</code> uses terminfo, which is a database
330 format that can describe the capabilities of thousands of
331 different terminals.</p>
332
333 <p>The <code>curses</code> API may seem something of an archaism
334 on UNIX desktops increasingly dominated by X, Motif, and Tcl/Tk.
335 Nevertheless, UNIX still supports tty lines and X supports
336 <em>xterm(1)</em>; the <code>curses</code> API has the advantage
337 of (a) back-portability to character-cell terminals, and (b)
338 simplicity. For an application that does not require bit-mapped
339 graphics and multiple fonts, an interface implementation using
340 <code>curses</code> will typically be a great deal simpler and
341 less expensive than one using an X toolkit.</p>
342
343 <h3><a name="history" id="history">A Brief History of Curses</a></h3>
344
345 <p>Historically, the first ancestor of <code>curses</code> was
346 the routines written to provide screen-handling for the
347 <code>vi</code> editor; these used the <code>termcap</code>
348 database facility (both released in 3BSD) for describing terminal
349 capabilities. These routines were abstracted into a documented
350 library and first released with the early BSD UNIX versions. All
351 of this work was done by students at the University of California
352 (Berkeley campus). The curses library was first published in
353 4.0BSD, a year after 3BSD (i.e., late 1980).</p>
354
355 <p>After graduation, one of those students went to work at
356 AT&amp;T Bell Labs, and made an improved <code>termcap</code>
357 library called <code>terminfo</code> (i.e.,
358 &ldquo;libterm&rdquo;), and adapted the curses library to use
359 this. That was subsequently released in System V Release 2 (early
360 1984). Thereafter, other developers added to the curses and
361 terminfo libraries. For instance, a student at Cornell University
362 wrote an improved terminfo library as well as a tool
363 (<code>tic</code>) to compile the terminal descriptions. As a
364 general rule, AT&amp;T did not identify the developers in the
365 source-code or documentation; the <code>tic</code> and
366 <code>infocmp</code> programs are the exceptions.</p>
367
368 <p>System V Release 3 (System III UNIX) from Bell Labs featured a
369 rewritten and much-improved <code>curses</code> library, along
370 with the <code>tic</code> program (late 1986).</p>
371
372 <p>To recap, terminfo is based on Berkeley's termcap database,
373 but contains a number of improvements and extensions.
374 Parameterized capabilities strings were introduced, making it
375 possible to describe multiple video attributes, and colors and to
376 handle far more unusual terminals than possible with termcap. In
377 the later AT&amp;T System V releases, <code>curses</code> evolved
378 to use more facilities and offer more capabilities, going far
379 beyond BSD curses in power and flexibility.</p>
380
381 <h3><a name="scope" id="scope">Scope of This Document</a></h3>
382
383 <p>This document describes <code>ncurses</code>, a free
384 implementation of the System V <code>curses</code> API with some
385 clearly marked extensions. It includes the following System V
386 curses features:</p>
387
388 <ul>
389 <li>Support for multiple screen highlights (BSD curses could
390 only handle one &ldquo;standout&rdquo; highlight, usually
391 reverse-video).</li>
392
393 <li>Support for line- and box-drawing using forms
394 characters.</li>
395
396 <li>Recognition of function keys on input.</li>
397
398 <li>Color support.</li>
399
400 <li>Support for pads (windows of larger than screen size on
401 which the screen or a subwindow defines a viewport).</li>
402 </ul>
403
404 <p>Also, this package makes use of the insert and delete line and
405 character features of terminals so equipped, and determines how
406 to optimally use these features with no help from the programmer.
407 It allows arbitrary combinations of video attributes to be
408 displayed, even on terminals that leave &ldquo;magic
409 cookies&rdquo; on the screen to mark changes in attributes.</p>
410
411 <p>The <code>ncurses</code> package can also capture and use
412 event reports from a mouse in some environments (notably, xterm
413 under the X window system). This document includes tips for using
414 the mouse.</p>
415
416 <p>The <code>ncurses</code> package was originated by Pavel
417 Curtis. The original maintainer of this package is <a href=
418 "mailto:zmbenhal@netcom.com">Zeyd Ben-Halim</a>
419 &lt;zmbenhal@netcom.com&gt;. <a href=
420 "mailto:esr@snark.thyrsus.com">Eric S. Raymond</a>
421 &lt;esr@snark.thyrsus.com&gt; wrote many of the new features in
422 versions after 1.8.1 and wrote most of this introduction.
423 J&uuml;rgen Pfeifer wrote all of the menu and forms code as well
424 as the <a href="http://www.adahome.com">Ada95</a> binding.
425 Ongoing work is being done by <a href=
426 "mailto:dickey@invisible-island.net">Thomas Dickey</a>
427 (maintainer). Contact the current maintainers at <a href=
428 "mailto:bug-ncurses@gnu.org">bug-ncurses@gnu.org</a>.</p>
429
430 <p>This document also describes the <a href="#panels">panels</a>
431 extension library, similarly modeled on the SVr4 panels facility.
432 This library allows you to associate backing store with each of a
433 stack or deck of overlapping windows, and provides operations for
434 moving windows around in the stack that change their visibility
435 in the natural way (handling window overlaps).</p>
436
437 <p>Finally, this document describes in detail the <a href=
438 "#menu">menus</a> and <a href="#form">forms</a> extension
439 libraries, also cloned from System V, which support easy
440 construction and sequences of menus and fill-in forms.</p>
441
442 <h3><a name="terminology" id="terminology">Terminology</a></h3>
443
444 <p>In this document, the following terminology is used with
445 reasonable consistency:</p>
446
447 <dl>
448 <dt>window</dt>
449
450 <dd>A data structure describing a sub-rectangle of the screen
451 (possibly the entire screen). You can write to a window as
452 though it were a miniature screen, scrolling independently of
453 other windows on the physical screen.</dd>
454
455 <dt>screens</dt>
456
457 <dd>A subset of windows which are as large as the terminal
458 screen, i.e., they start at the upper left hand corner and
459 encompass the lower right hand corner. One of these,
460 <code>stdscr</code>, is automatically provided for the
461 programmer.</dd>
462
463 <dt>terminal screen</dt>
464
465 <dd>The package's idea of what the terminal display currently
466 looks like, i.e., what the user sees now. This is a special
467 screen.</dd>
468 </dl>
469
470 <h2><a name="curses" id="curses">The Curses Library</a></h2>
471
472 <h3><a name="overview" id="overview">An Overview of Curses</a></h3>
473
474 <h4><a name="compiling" id="compiling">Compiling Programs using
475 Curses</a></h4>
476
477 <p>In order to use the library, it is necessary to have certain
478 types and variables defined. Therefore, the programmer must have
479 a line:</p>
480
481 <pre class="code-block">
482 #include &lt;curses.h&gt;
483</pre>
484 <p>at the top of the program source. The screen package uses the
485 Standard I/O library, so <code>&lt;curses.h&gt;</code> includes
486 <code>&lt;stdio.h&gt;</code>. <code>&lt;curses.h&gt;</code> also
487 includes <code>&lt;termios.h&gt;</code>,
488 <code>&lt;termio.h&gt;</code>, or <code>&lt;sgtty.h&gt;</code>
489 depending on your system. It is redundant (but harmless) for the
490 programmer to do these includes, too. In linking with
491 <code>curses</code> you need to have <code>-lncurses</code> in
492 your LDFLAGS or on the command line. There is no need for any
493 other libraries.</p>
494
495 <h4><a name="updating" id="updating">Updating the Screen</a></h4>
496
497 <p>In order to update the screen optimally, it is necessary for
498 the routines to know what the screen currently looks like and
499 what the programmer wants it to look like next. For this purpose,
500 a data type (structure) named WINDOW is defined which describes a
501 window image to the routines, including its starting position on
502 the screen (the (y, x) coordinates of the upper left hand corner)
503 and its size. One of these (called <code>curscr</code>, for
504 current screen) is a screen image of what the terminal currently
505 looks like. Another screen (called <code>stdscr</code>, for
506 standard screen) is provided by default to make changes on.</p>
507
508 <p>A window is a purely internal representation. It is used to
509 build and store a potential image of a portion of the terminal.
510 It does not bear any necessary relation to what is really on the
511 terminal screen; it is more like a scratchpad or write
512 buffer.</p>
513
514 <p>To make the section of physical screen corresponding to a
515 window reflect the contents of the window structure, the routine
516 <code>refresh()</code> (or <code>wrefresh()</code> if the window
517 is not <code>stdscr</code>) is called.</p>
518
519 <p>A given physical screen section may be within the scope of any
520 number of overlapping windows. Also, changes can be made to
521 windows in any order, without regard to motion efficiency. Then,
522 at will, the programmer can effectively say &ldquo;make it look
523 like this,&rdquo; and let the package implementation determine
524 the most efficient way to repaint the screen.</p>
525
526 <h4><a name="stdscr" id="stdscr">Standard Windows and Function
527 Naming Conventions</a></h4>
528
529 <p>As hinted above, the routines can use several windows, but two
530 are automatically given: <code>curscr</code>, which knows what
531 the terminal looks like, and <code>stdscr</code>, which is what
532 the programmer wants the terminal to look like next. The user
533 should never actually access <code>curscr</code> directly.
534 Changes should be made to through the API, and then the routine
535 <code>refresh()</code> (or <code>wrefresh()</code>) called.</p>
536
537 <p>Many functions are defined to use <code>stdscr</code> as a
538 default screen. For example, to add a character to
539 <code>stdscr</code>, one calls <code>addch()</code> with the
540 desired character as argument. To write to a different window.
541 use the routine <code>waddch()</code> (for
542 <strong>w</strong>indow-specific addch()) is provided. This
543 convention of prepending function names with a &ldquo;w&rdquo;
544 when they are to be applied to specific windows is consistent.
545 The only routines which do not follow it are those for which a
546 window must always be specified.</p>
547
548 <p>In order to move the current (y, x) coordinates from one point
549 to another, the routines <code>move()</code> and
550 <code>wmove()</code> are provided. However, it is often desirable
551 to first move and then perform some I/O operation. In order to
552 avoid clumsiness, most I/O routines can be preceded by the prefix
553 &ldquo;mv&rdquo; and the desired (y, x) coordinates prepended to
554 the arguments to the function. For example, the calls</p>
555
556 <pre class="code-block">
557 move(y, x);
558 addch(ch);
559</pre>
560 <p>can be replaced by</p>
561
562 <pre class="code-block">
563 mvaddch(y, x, ch);
564</pre>
565 <p>and</p>
566
567 <pre class="code-block">
568 wmove(win, y, x);
569 waddch(win, ch);
570</pre>
571 <p>can be replaced by</p>
572
573 <pre class="code-block">
574 mvwaddch(win, y, x, ch);
575</pre>
576 <p>Note that the window description pointer (win) comes before
577 the added (y, x) coordinates. If a function requires a window
578 pointer, it is always the first parameter passed.</p>
579
580 <h4><a name="variables" id="variables">Variables</a></h4>
581
582 <p>The <code>curses</code> library sets some variables describing
583 the terminal capabilities.</p>
584
585 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]586 type name description
587 ------------------------------------------------------------------
588 int LINES number of lines on the terminal
589 int COLS number of columns on the terminal
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]590</pre>
591 <p>The <code>curses.h</code> also introduces some
592 <code>#define</code> constants and types of general
593 usefulness:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]594
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]595 <dl>
596 <dt><code>bool</code>
597 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]598
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]599 <dd>boolean type, actually a &ldquo;char&rdquo; (e.g.,
600 <code>bool doneit;</code>)</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]601
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]602 <dt><code>TRUE</code>
603 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]604
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]605 <dd>boolean &ldquo;true&rdquo; flag (1).</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]606
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]607 <dt><code>FALSE</code>
608 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]609
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]610 <dd>boolean &ldquo;false&rdquo; flag (0).</dd>
611
612 <dt><code>ERR</code>
613 </dt>
614
615 <dd>error flag returned by routines on a failure (-1).</dd>
616
617 <dt><code>OK</code>
618 </dt>
619
620 <dd>error flag returned by routines when things go right.</dd>
621 </dl>
622
623 <h3><a name="using" id="using">Using the Library</a></h3>
624
625 <p>Now we describe how to actually use the screen package. In it,
626 we assume all updating, reading, etc. is applied to
627 <code>stdscr</code>. These instructions will work on any window,
628 providing you change the function names and parameters as
629 mentioned above.</p>
630
631 <p>Here is a sample program to motivate the discussion:</p>
632
633 <pre class="code-block">
Steve Kondikae271bc2015-11-15 02:50:53 +0100[diff] [blame]634#include &lt;stdlib.h&gt;
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]635#include &lt;curses.h&gt;
636#include &lt;signal.h&gt;
637
638static void finish(int sig);
639
640int
641main(int argc, char *argv[])
642{
643 int num = 0;
644
645 /* initialize your non-curses data structures here */
646
647 (void) signal(SIGINT, finish); /* arrange interrupts to terminate */
648
649 (void) initscr(); /* initialize the curses library */
650 keypad(stdscr, TRUE); /* enable keyboard mapping */
651 (void) nonl(); /* tell curses not to do NL-&gt;CR/NL on output */
652 (void) cbreak(); /* take input chars one at a time, no wait for \n */
653 (void) echo(); /* echo input - in color */
654
655 if (has_colors())
656 {
657 start_color();
658
659 /*
660 * Simple color assignment, often all we need. Color pair 0 cannot
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]661 * be redefined. This example uses the same value for the color
662 * pair as for the foreground color, though of course that is not
663 * necessary:
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]664 */
665 init_pair(1, COLOR_RED, COLOR_BLACK);
666 init_pair(2, COLOR_GREEN, COLOR_BLACK);
667 init_pair(3, COLOR_YELLOW, COLOR_BLACK);
668 init_pair(4, COLOR_BLUE, COLOR_BLACK);
669 init_pair(5, COLOR_CYAN, COLOR_BLACK);
670 init_pair(6, COLOR_MAGENTA, COLOR_BLACK);
671 init_pair(7, COLOR_WHITE, COLOR_BLACK);
672 }
673
674 for (;;)
675 {
676 int c = getch(); /* refresh, accept single keystroke of input */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]677 attrset(COLOR_PAIR(num % 8));
678 num++;
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]679
680 /* process the command keystroke */
681 }
682
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]683 finish(0); /* we are done */
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]684}
685
686static void finish(int sig)
687{
688 endwin();
689
690 /* do your non-curses wrapup here */
691
692 exit(0);
693}
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]694</pre>
695 <h4><a name="starting" id="starting">Starting up</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]696
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]697 <p>In order to use the screen package, the routines must know
698 about terminal characteristics, and the space for
699 <code>curscr</code> and <code>stdscr</code> must be allocated.
700 These function <code>initscr()</code> does both these things.
701 Since it must allocate space for the windows, it can overflow
702 memory when attempting to do so. On the rare occasions this
703 happens, <code>initscr()</code> will terminate the program with
704 an error message. <code>initscr()</code> must always be called
705 before any of the routines which affect windows are used. If it
706 is not, the program will core dump as soon as either
707 <code>curscr</code> or <code>stdscr</code> are referenced.
708 However, it is usually best to wait to call it until after you
709 are sure you will need it, like after checking for startup
710 errors. Terminal status changing routines like <code>nl()</code>
711 and <code>cbreak()</code> should be called after
712 <code>initscr()</code>.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]713
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]714 <p>Once the screen windows have been allocated, you can set them
715 up for your program. If you want to, say, allow a screen to
716 scroll, use <code>scrollok()</code>. If you want the cursor to be
717 left in place after the last change, use <code>leaveok()</code>.
718 If this is not done, <code>refresh()</code> will move the cursor
719 to the window's current (y, x) coordinates after updating it.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]720
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]721 <p>You can create new windows of your own using the functions
722 <code>newwin()</code>, <code>derwin()</code>, and
723 <code>subwin()</code>. The routine <code>delwin()</code> will
724 allow you to get rid of old windows. All the options described
725 above can be applied to any window.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]726
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]727 <h4><a name="output" id="output">Output</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]728
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]729 <p>Now that we have set things up, we will want to actually
730 update the terminal. The basic functions used to change what will
731 go on a window are <code>addch()</code> and <code>move()</code>.
732 <code>addch()</code> adds a character at the current (y, x)
733 coordinates. <code>move()</code> changes the current (y, x)
734 coordinates to whatever you want them to be. It returns
735 <code>ERR</code> if you try to move off the window. As mentioned
736 above, you can combine the two into <code>mvaddch()</code> to do
737 both things at once.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]738
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]739 <p>The other output functions, such as <code>addstr()</code> and
740 <code>printw()</code>, all call <code>addch()</code> to add
741 characters to the window.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]742
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]743 <p>After you have put on the window what you want there, when you
744 want the portion of the terminal covered by the window to be made
745 to look like it, you must call <code>refresh()</code>. In order
746 to optimize finding changes, <code>refresh()</code> assumes that
747 any part of the window not changed since the last
748 <code>refresh()</code> of that window has not been changed on the
749 terminal, i.e., that you have not refreshed a portion of the
750 terminal with an overlapping window. If this is not the case, the
751 routine <code>touchwin()</code> is provided to make it look like
752 the entire window has been changed, thus making
753 <code>refresh()</code> check the whole subsection of the terminal
754 for changes.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]755
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]756 <p>If you call <code>wrefresh()</code> with <code>curscr</code>
757 as its argument, it will make the screen look like
758 <code>curscr</code> thinks it looks like. This is useful for
759 implementing a command which would redraw the screen in case it
760 get messed up.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]761
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]762 <h4><a name="input" id="input">Input</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]763
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]764 <p>The complementary function to <code>addch()</code> is
765 <code>getch()</code> which, if echo is set, will call
766 <code>addch()</code> to echo the character. Since the screen
767 package needs to know what is on the terminal at all times, if
768 characters are to be echoed, the tty must be in raw or cbreak
769 mode. Since initially the terminal has echoing enabled and is in
770 ordinary &ldquo;cooked&rdquo; mode, one or the other has to
771 changed before calling <code>getch()</code>; otherwise, the
772 program's output will be unpredictable.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]773
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]774 <p>When you need to accept line-oriented input in a window, the
775 functions <code>wgetstr()</code> and friends are available. There
776 is even a <code>wscanw()</code> function that can do
777 <code>scanf()</code>(3)-style multi-field parsing on window
778 input. These pseudo-line-oriented functions turn on echoing while
779 they execute.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]780
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]781 <p>The example code above uses the call <code>keypad(stdscr,
782 TRUE)</code> to enable support for function-key mapping. With
783 this feature, the <code>getch()</code> code watches the input
784 stream for character sequences that correspond to arrow and
785 function keys. These sequences are returned as pseudo-character
786 values. The <code>#define</code> values returned are listed in
787 the <code>curses.h</code> The mapping from sequences to
788 <code>#define</code> values is determined by <code>key_</code>
789 capabilities in the terminal's terminfo entry.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]790
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]791 <h4><a name="formschars" id="formschars">Using Forms
792 Characters</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]793
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]794 <p>The <code>addch()</code> function (and some others, including
795 <code>box()</code> and <code>border()</code>) can accept some
796 pseudo-character arguments which are specially defined by
797 <code>ncurses</code>. These are <code>#define</code> values set
798 up in the <code>curses.h</code> header; see there for a complete
799 list (look for the prefix <code>ACS_</code>).</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]800
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]801 <p>The most useful of the ACS defines are the forms-drawing
802 characters. You can use these to draw boxes and simple graphs on
803 the screen. If the terminal does not have such characters,
804 <code>curses.h</code> will map them to a recognizable (though
805 ugly) set of ASCII defaults.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]806
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]807 <h4><a name="attributes" id="attributes">Character Attributes and
808 Color</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]809
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]810 <p>The <code>ncurses</code> package supports screen highlights
811 including standout, reverse-video, underline, and blink. It also
812 supports color, which is treated as another kind of
813 highlight.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]814
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]815 <p>Highlights are encoded, internally, as high bits of the
816 pseudo-character type (<code>chtype</code>) that
817 <code>curses.h</code> uses to represent the contents of a screen
818 cell. See the <code>curses.h</code> header file for a complete
819 list of highlight mask values (look for the prefix
820 <code>A_</code>).</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]821
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]822 <p>There are two ways to make highlights. One is to logical-or
823 the value of the highlights you want into the character argument
824 of an <code>addch()</code> call, or any other output call that
825 takes a <code>chtype</code> argument.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]826
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]827 <p>The other is to set the current-highlight value. This is
828 <em>logical-OR</em>ed with any highlight you specify the first
829 way. You do this with the functions <code>attron()</code>,
830 <code>attroff()</code>, and <code>attrset()</code>; see the
831 manual pages for details. Color is a special kind of highlight.
832 The package actually thinks in terms of color pairs, combinations
833 of foreground and background colors. The sample code above sets
834 up eight color pairs, all of the guaranteed-available colors on
835 black. Note that each color pair is, in effect, given the name of
836 its foreground color. Any other range of eight non-conflicting
837 values could have been used as the first arguments of the
838 <code>init_pair()</code> values.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]839
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]840 <p>Once you have done an <code>init_pair()</code> that creates
841 color-pair N, you can use <code>COLOR_PAIR(N)</code> as a
842 highlight that invokes that particular color combination. Note
843 that <code>COLOR_PAIR(N)</code>, for constant N, is itself a
844 compile-time constant and can be used in initializers.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]845
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]846 <h4><a name="mouse" id="mouse">Mouse Interfacing</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]847
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]848 <p>The <code>ncurses</code> library also provides a mouse
849 interface.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]850
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]851 <blockquote>
852 <strong>NOTE:</strong> this facility is specific to
853 <code>ncurses</code>, it is not part of either the XSI Curses
854 standard, nor of System V Release 4, nor BSD curses. System V
855 Release 4 curses contains code with similar interface
856 definitions, however it is not documented. Other than by
857 disassembling the library, we have no way to determine exactly
858 how that mouse code works. Thus, we recommend that you wrap
859 mouse-related code in an #ifdef using the feature macro
860 NCURSES_MOUSE_VERSION so it will not be compiled and linked on
861 non-ncurses systems.
862 </blockquote>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]863
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]864 <p>Presently, mouse event reporting works in the following
865 environments:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]866
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]867 <ul>
868 <li>xterm and similar programs such as rxvt.</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]869
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]870 <li>Linux console, when configured with <code>gpm</code>(1),
871 Alessandro Rubini's mouse server.</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]872
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]873 <li>FreeBSD sysmouse (console)</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]874
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]875 <li>OS/2 EMX</li>
876 </ul>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]877
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]878 <p>The mouse interface is very simple. To activate it, you use
879 the function <code>mousemask()</code>, passing it as first
880 argument a bit-mask that specifies what kinds of events you want
881 your program to be able to see. It will return the bit-mask of
882 events that actually become visible, which may differ from the
883 argument if the mouse device is not capable of reporting some of
884 the event types you specify.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]885
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]886 <p>Once the mouse is active, your application's command loop
887 should watch for a return value of <code>KEY_MOUSE</code> from
888 <code>wgetch()</code>. When you see this, a mouse event report
889 has been queued. To pick it off the queue, use the function
890 <code>getmouse()</code> (you must do this before the next
891 <code>wgetch()</code>, otherwise another mouse event might come
892 in and make the first one inaccessible).</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]893
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]894 <p>Each call to <code>getmouse()</code> fills a structure (the
895 address of which you will pass it) with mouse event data. The
896 event data includes zero-origin, screen-relative character-cell
897 coordinates of the mouse pointer. It also includes an event mask.
898 Bits in this mask will be set, corresponding to the event type
899 being reported.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]900
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]901 <p>The mouse structure contains two additional fields which may
902 be significant in the future as ncurses interfaces to new kinds
903 of pointing device. In addition to x and y coordinates, there is
904 a slot for a z coordinate; this might be useful with
905 touch-screens that can return a pressure or duration parameter.
906 There is also a device ID field, which could be used to
907 distinguish between multiple pointing devices.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]908
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]909 <p>The class of visible events may be changed at any time via
910 <code>mousemask()</code>. Events that can be reported include
911 presses, releases, single-, double- and triple-clicks (you can
912 set the maximum button-down time for clicks). If you do not make
913 clicks visible, they will be reported as press-release pairs. In
914 some environments, the event mask may include bits reporting the
915 state of shift, alt, and ctrl keys on the keyboard during the
916 event.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]917
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]918 <p>A function to check whether a mouse event fell within a given
919 window is also supplied. You can use this to see whether a given
920 window should consider a mouse event relevant to it.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]921
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]922 <p>Because mouse event reporting will not be available in all
923 environments, it would be unwise to build <code>ncurses</code>
924 applications that <em>require</em> the use of a mouse. Rather,
925 you should use the mouse as a shortcut for point-and-shoot
926 commands your application would normally accept from the
927 keyboard. Two of the test games in the <code>ncurses</code>
928 distribution (<code>bs</code> and <code>knight</code>) contain
929 code that illustrates how this can be done.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]930
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]931 <p>See the manual page <code>curs_mouse(3X)</code> for full
932 details of the mouse-interface functions.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]933
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]934 <h4><a name="finishing" id="finishing">Finishing Up</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]935
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]936 <p>In order to clean up after the <code>ncurses</code> routines,
937 the routine <code>endwin()</code> is provided. It restores tty
938 modes to what they were when <code>initscr()</code> was first
939 called, and moves the cursor down to the lower-left corner. Thus,
940 anytime after the call to initscr, <code>endwin()</code> should
941 be called before exiting.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]942
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]943 <h3><a name="functions" id="functions">Function Descriptions</a></h3>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]944
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]945 <p>We describe the detailed behavior of some important curses
946 functions here, as a supplement to the manual page
947 descriptions.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]948
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]949 <h4><a name="init" id="init">Initialization and Wrapup</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]950
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]951 <dl>
952 <dt><code>initscr()</code>
953 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]954
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]955 <dd>The first function called should almost always be
956 <code>initscr()</code>. This will determine the terminal type
957 and initialize curses data structures. <code>initscr()</code>
958 also arranges that the first call to <code>refresh()</code>
959 will clear the screen. If an error occurs a message is written
960 to standard error and the program exits. Otherwise it returns a
961 pointer to stdscr. A few functions may be called before initscr
962 (<code>slk_init()</code>, <code>filter()</code>,
963 <code>ripoffline()</code>, <code>use_env()</code>, and, if you
964 are using multiple terminals, <code>newterm()</code>.)</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]965
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]966 <dt><code>endwin()</code>
967 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]968
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]969 <dd>Your program should always call <code>endwin()</code>
970 before exiting or shelling out of the program. This function
971 will restore tty modes, move the cursor to the lower left
972 corner of the screen, reset the terminal into the proper
973 non-visual mode. Calling <code>refresh()</code> or
974 <code>doupdate()</code> after a temporary escape from the
975 program will restore the ncurses screen from before the
976 escape.</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]977
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]978 <dt><code>newterm(type, ofp, ifp)</code>
979 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]980
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]981 <dd>A program which outputs to more than one terminal should
982 use <code>newterm()</code> instead of <code>initscr()</code>.
983 <code>newterm()</code> should be called once for each terminal.
984 It returns a variable of type <code>SCREEN *</code> which
985 should be saved as a reference to that terminal. (NOTE: a
986 SCREEN variable is not a <em>screen</em> in the sense we are
987 describing in this introduction, but a collection of parameters
988 used to assist in optimizing the display.) The arguments are
989 the type of the terminal (a string) and <code>FILE</code>
990 pointers for the output and input of the terminal. If type is
991 NULL then the environment variable <code>$TERM</code> is used.
992 <code>endwin()</code> should called once at wrapup time for
993 each terminal opened using this function.</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]994
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]995 <dt><code>set_term(new)</code>
996 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]997
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]998 <dd>This function is used to switch to a different terminal
999 previously opened by <code>newterm()</code>. The screen
1000 reference for the new terminal is passed as the parameter. The
1001 previous terminal is returned by the function. All other calls
1002 affect only the current terminal.</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1003
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1004 <dt><code>delscreen(sp)</code>
1005 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1006
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1007 <dd>The inverse of <code>newterm()</code>; deallocates the data
1008 structures associated with a given <code>SCREEN</code>
1009 reference.</dd>
1010 </dl>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1011
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1012 <h4><a name="flush" id="flush">Causing Output to the Terminal</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1013
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1014 <dl>
1015 <dt><code>refresh()</code> and <code>wrefresh(win)</code></dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1016
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1017 <dd>These functions must be called to actually get any output
1018 on the terminal, as other routines merely manipulate data
1019 structures. <code>wrefresh()</code> copies the named window to
1020 the physical terminal screen, taking into account what is
1021 already there in order to do optimizations.
1022 <code>refresh()</code> does a refresh of <code>stdscr</code>.
1023 Unless <code>leaveok()</code> has been enabled, the physical
1024 cursor of the terminal is left at the location of the window's
1025 cursor.</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1026
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1027 <dt><code>doupdate()</code> and
1028 <code>wnoutrefresh(win)</code></dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1029
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1030 <dd>These two functions allow multiple updates with more
1031 efficiency than wrefresh. To use them, it is important to
1032 understand how curses works. In addition to all the window
1033 structures, curses keeps two data structures representing the
1034 terminal screen: a physical screen, describing what is actually
1035 on the screen, and a virtual screen, describing what the
1036 programmer wants to have on the screen. wrefresh works by first
1037 copying the named window to the virtual screen
1038 (<code>wnoutrefresh()</code>), and then calling the routine to
1039 update the screen (<code>doupdate()</code>). If the programmer
1040 wishes to output several windows at once, a series of calls to
1041 <code>wrefresh</code> will result in alternating calls to
1042 <code>wnoutrefresh()</code> and <code>doupdate()</code>,
1043 causing several bursts of output to the screen. By calling
1044 <code>wnoutrefresh()</code> for each window, it is then
1045 possible to call <code>doupdate()</code> once, resulting in
1046 only one burst of output, with fewer total characters
1047 transmitted (this also avoids a visually annoying flicker at
1048 each update).</dd>
1049 </dl>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1050
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1051 <h4><a name="lowlevel" id="lowlevel">Low-Level Capability
1052 Access</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1053
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1054 <dl>
1055 <dt><code>setupterm(term, filenum, errret)</code>
1056 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1057
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1058 <dd>
1059 This routine is called to initialize a terminal's
1060 description, without setting up the curses screen structures
1061 or changing the tty-driver mode bits. <code>term</code> is
1062 the character string representing the name of the terminal
1063 being used. <code>filenum</code> is the UNIX file descriptor
1064 of the terminal to be used for output. <code>errret</code> is
1065 a pointer to an integer, in which a success or failure
1066 indication is returned. The values returned can be 1 (all is
1067 well), 0 (no such terminal), or -1 (some problem locating the
1068 terminfo database).
1069 <p>The value of <code>term</code> can be given as NULL, which
1070 will cause the value of <code>TERM</code> in the environment
1071 to be used. The <code>errret</code> pointer can also be given
1072 as NULL, meaning no error code is wanted. If
1073 <code>errret</code> is defaulted, and something goes wrong,
1074 <code>setupterm()</code> will print an appropriate error
1075 message and exit, rather than returning. Thus, a simple
1076 program can call setupterm(0, 1, 0) and not worry about
1077 initialization errors.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1078
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1079 <p>After the call to <code>setupterm()</code>, the global
1080 variable <code>cur_term</code> is set to point to the current
1081 structure of terminal capabilities. By calling
1082 <code>setupterm()</code> for each terminal, and saving and
1083 restoring <code>cur_term</code>, it is possible for a program
1084 to use two or more terminals at once.
1085 <code>Setupterm()</code> also stores the names section of the
1086 terminal description in the global character array
1087 <code>ttytype[]</code>. Subsequent calls to
1088 <code>setupterm()</code> will overwrite this array, so you
1089 will have to save it yourself if need be.</p>
1090 </dd>
1091 </dl>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1092
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1093 <h4><a name="debugging" id="debugging">Debugging</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1094
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1095 <blockquote>
1096 <strong>NOTE:</strong> These functions are not part of the
1097 standard curses API!
1098 </blockquote>
1099
1100 <dl>
1101 <dt><code>trace()</code>
1102 </dt>
1103
1104 <dd>This function can be used to explicitly set a trace level.
1105 If the trace level is nonzero, execution of your program will
1106 generate a file called &ldquo;trace&rdquo; in the current
1107 working directory containing a report on the library's actions.
1108 Higher trace levels enable more detailed (and verbose)
1109 reporting -- see comments attached to <code>TRACE_</code>
1110 defines in the <code>curses.h</code> file for details. (It is
1111 also possible to set a trace level by assigning a trace level
1112 value to the environment variable
1113 <code>NCURSES_TRACE</code>).</dd>
1114
1115 <dt><code>_tracef()</code>
1116 </dt>
1117
1118 <dd>This function can be used to output your own debugging
1119 information. It is only available only if you link with
1120 -lncurses_g. It can be used the same way as
1121 <code>printf()</code>, only it outputs a newline after the end
1122 of arguments. The output goes to a file called
1123 <code>trace</code> in the current directory.</dd>
1124 </dl>
1125
1126 <p>Trace logs can be difficult to interpret due to the sheer
1127 volume of data dumped in them. There is a script called
1128 <strong>tracemunch</strong> included with the
1129 <code>ncurses</code> distribution that can alleviate this problem
1130 somewhat; it compacts long sequences of similar operations into
1131 more succinct single-line pseudo-operations. These pseudo-ops can
1132 be distinguished by the fact that they are named in capital
1133 letters.</p>
1134
1135 <h3><a name="hints" id="hints">Hints, Tips, and Tricks</a></h3>
1136
1137 <p>The <code>ncurses</code> manual pages are a complete reference
1138 for this library. In the remainder of this document, we discuss
1139 various useful methods that may not be obvious from the manual
1140 page descriptions.</p>
1141
1142 <h4><a name="caution" id="caution">Some Notes of Caution</a></h4>
1143
1144 <p>If you find yourself thinking you need to use
1145 <code>noraw()</code> or <code>nocbreak()</code>, think again and
1146 move carefully. It is probably better design to use
1147 <code>getstr()</code> or one of its relatives to simulate cooked
1148 mode. The <code>noraw()</code> and <code>nocbreak()</code>
1149 functions try to restore cooked mode, but they may end up
1150 clobbering some control bits set before you started your
1151 application. Also, they have always been poorly documented, and
1152 are likely to hurt your application's usability with other curses
1153 libraries.</p>
1154
1155 <p>Bear in mind that <code>refresh()</code> is a synonym for
1156 <code>wrefresh(stdscr)</code>. Do not try to mix use of
1157 <code>stdscr</code> with use of windows declared by
1158 <code>newwin()</code>; a <code>refresh()</code> call will blow
1159 them off the screen. The right way to handle this is to use
1160 <code>subwin()</code>, or not touch <code>stdscr</code> at all
1161 and tile your screen with declared windows which you then
1162 <code>wnoutrefresh()</code> somewhere in your program event loop,
1163 with a single <code>doupdate()</code> call to trigger actual
1164 repainting.</p>
1165
1166 <p>You are much less likely to run into problems if you design
1167 your screen layouts to use tiled rather than overlapping windows.
1168 Historically, curses support for overlapping windows has been
1169 weak, fragile, and poorly documented. The <code>ncurses</code>
1170 library is not yet an exception to this rule.</p>
1171
1172 <p>There is a panels library included in the <code>ncurses</code>
1173 distribution that does a pretty good job of strengthening the
1174 overlapping-windows facilities.</p>
1175
1176 <p>Try to avoid using the global variables LINES and COLS. Use
1177 <code>getmaxyx()</code> on the <code>stdscr</code> context
1178 instead. Reason: your code may be ported to run in an environment
1179 with window resizes, in which case several screens could be open
1180 with different sizes.</p>
1181
1182 <h4><a name="leaving" id="leaving">Temporarily Leaving NCURSES
1183 Mode</a></h4>
1184
1185 <p>Sometimes you will want to write a program that spends most of
1186 its time in screen mode, but occasionally returns to ordinary
1187 &ldquo;cooked&rdquo; mode. A common reason for this is to support
1188 shell-out. This behavior is simple to arrange in
1189 <code>ncurses</code>.</p>
1190
1191 <p>To leave <code>ncurses</code> mode, call <code>endwin()</code>
1192 as you would if you were intending to terminate the program. This
1193 will take the screen back to cooked mode; you can do your
1194 shell-out. When you want to return to <code>ncurses</code> mode,
1195 simply call <code>refresh()</code> or <code>doupdate()</code>.
1196 This will repaint the screen.</p>
1197
1198 <p>There is a boolean function, <code>isendwin()</code>, which
1199 code can use to test whether <code>ncurses</code> screen mode is
1200 active. It returns <code>TRUE</code> in the interval between an
1201 <code>endwin()</code> call and the following
1202 <code>refresh()</code>, <code>FALSE</code> otherwise.</p>
1203
1204 <p>Here is some sample code for shellout:</p>
1205
1206 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1207 addstr("Shelling out...");
1208 def_prog_mode(); /* save current tty modes */
1209 endwin(); /* restore original tty modes */
1210 system("sh"); /* run shell */
1211 addstr("returned.\n"); /* prepare return message */
1212 refresh(); /* restore save modes, repaint screen */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1213</pre>
1214 <h4><a name="xterm" id="xterm">Using NCURSES under XTERM</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1215
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1216 <p>A resize operation in X sends <code>SIGWINCH</code> to the
1217 application running under xterm. The easiest way to handle
1218 <code>SIGWINCH</code> is to do an <code>endwin</code>, followed
1219 by an <code>refresh</code> and a screen repaint you code
1220 yourself. The <code>refresh</code> will pick up the new screen
1221 size from the xterm's environment.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1222
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1223 <p>That is the standard way, of course (it even works with some
1224 vendor's curses implementations). Its drawback is that it clears
1225 the screen to reinitialize the display, and does not resize
1226 subwindows which must be shrunk. <code>Ncurses</code> provides an
1227 extension which works better, the <code>resizeterm</code>
1228 function. That function ensures that all windows are limited to
1229 the new screen dimensions, and pads <code>stdscr</code> with
1230 blanks if the screen is larger.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1231
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1232 <p>The <code>ncurses</code> library provides a SIGWINCH signal
1233 handler, which pushes a <code>KEY_RESIZE</code> via the wgetch()
1234 calls. When <code>ncurses</code> returns that code, it calls
1235 <code>resizeterm</code> to update the size of the standard
1236 screen's window, repainting that (filling with blanks or
1237 truncating as needed). It also resizes other windows, but its
1238 effect may be less satisfactory because it cannot know how you
1239 want the screen re-painted. You will usually have to write
1240 special-purpose code to handle <code>KEY_RESIZE</code>
1241 yourself.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1242
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1243 <h4><a name="screens" id="screens">Handling Multiple Terminal
1244 Screens</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1245
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1246 <p>The <code>initscr()</code> function actually calls a function
1247 named <code>newterm()</code> to do most of its work. If you are
1248 writing a program that opens multiple terminals, use
1249 <code>newterm()</code> directly.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1250
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1251 <p>For each call, you will have to specify a terminal type and a
1252 pair of file pointers; each call will return a screen reference,
1253 and <code>stdscr</code> will be set to the last one allocated.
1254 You will switch between screens with the <code>set_term</code>
1255 call. Note that you will also have to call
1256 <code>def_shell_mode</code> and <code>def_prog_mode</code> on
1257 each tty yourself.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1258
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1259 <h4><a name="testing" id="testing">Testing for Terminal
1260 Capabilities</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1261
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1262 <p>Sometimes you may want to write programs that test for the
1263 presence of various capabilities before deciding whether to go
1264 into <code>ncurses</code> mode. An easy way to do this is to call
1265 <code>setupterm()</code>, then use the functions
1266 <code>tigetflag()</code>, <code>tigetnum()</code>, and
1267 <code>tigetstr()</code> to do your testing.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1268
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1269 <p>A particularly useful case of this often comes up when you
1270 want to test whether a given terminal type should be treated as
1271 &ldquo;smart&rdquo; (cursor-addressable) or &ldquo;stupid&rdquo;.
1272 The right way to test this is to see if the return value of
1273 <code>tigetstr("cup")</code> is non-NULL. Alternatively, you can
1274 include the <code>term.h</code> file and test the value of the
1275 macro <code>cursor_address</code>.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1276
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1277 <h4><a name="tuning" id="tuning">Tuning for Speed</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1278
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1279 <p>Use the <code>addchstr()</code> family of functions for fast
1280 screen-painting of text when you know the text does not contain
1281 any control characters. Try to make attribute changes infrequent
1282 on your screens. Do not use the <code>immedok()</code>
1283 option!</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1284
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1285 <h4><a name="special" id="special">Special Features of
1286 NCURSES</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1287
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1288 <p>The <code>wresize()</code> function allows you to resize a
1289 window in place. The associated <code>resizeterm()</code>
1290 function simplifies the construction of <a href=
1291 "#xterm">SIGWINCH</a> handlers, for resizing all windows.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1292
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1293 <p>The <code>define_key()</code> function allows you to define at
1294 runtime function-key control sequences which are not in the
1295 terminal description. The <code>keyok()</code> function allows
1296 you to temporarily enable or disable interpretation of any
1297 function-key control sequence.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1298
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1299 <p>The <code>use_default_colors()</code> function allows you to
1300 construct applications which can use the terminal's default
1301 foreground and background colors as an additional "default"
1302 color. Several terminal emulators support this feature, which is
1303 based on ISO 6429.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1304
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1305 <p>Ncurses supports up 16 colors, unlike SVr4 curses which
1306 defines only 8. While most terminals which provide color allow
1307 only 8 colors, about a quarter (including XFree86 xterm) support
1308 16 colors.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1309
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1310 <h3><a name="compat" id="compat">Compatibility with Older
1311 Versions</a></h3>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1312
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1313 <p>Despite our best efforts, there are some differences between
1314 <code>ncurses</code> and the (undocumented!) behavior of older
1315 curses implementations. These arise from ambiguities or omissions
1316 in the documentation of the API.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1317
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1318 <h4><a name="refbug" id="refbug">Refresh of Overlapping
1319 Windows</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1320
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1321 <p>If you define two windows A and B that overlap, and then
1322 alternately scribble on and refresh them, the changes made to the
1323 overlapping region under historic <code>curses</code> versions
1324 were often not documented precisely.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1325
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1326 <p>To understand why this is a problem, remember that screen
1327 updates are calculated between two representations of the
1328 <em>entire</em> display. The documentation says that when you
1329 refresh a window, it is first copied to the virtual screen, and
1330 then changes are calculated to update the physical screen (and
1331 applied to the terminal). But "copied to" is not very specific,
1332 and subtle differences in how copying works can produce different
1333 behaviors in the case where two overlapping windows are each
1334 being refreshed at unpredictable intervals.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1335
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1336 <p>What happens to the overlapping region depends on what
1337 <code>wnoutrefresh()</code> does with its argument -- what
1338 portions of the argument window it copies to the virtual screen.
1339 Some implementations do "change copy", copying down only
1340 locations in the window that have changed (or been marked changed
1341 with <code>wtouchln()</code> and friends). Some implementations
1342 do "entire copy", copying <em>all</em> window locations to the
1343 virtual screen whether or not they have changed.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1344
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1345 <p>The <code>ncurses</code> library itself has not always been
1346 consistent on this score. Due to a bug, versions 1.8.7 to 1.9.8a
1347 did entire copy. Versions 1.8.6 and older, and versions 1.9.9 and
1348 newer, do change copy.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1349
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1350 <p>For most commercial curses implementations, it is not
1351 documented and not known for sure (at least not to the
1352 <code>ncurses</code> maintainers) whether they do change copy or
1353 entire copy. We know that System V release 3 curses has logic in
1354 it that looks like an attempt to do change copy, but the
1355 surrounding logic and data representations are sufficiently
1356 complex, and our knowledge sufficiently indirect, that it is hard
1357 to know whether this is reliable. It is not clear what the SVr4
1358 documentation and XSI standard intend. The XSI Curses standard
1359 barely mentions wnoutrefresh(); the SVr4 documents seem to be
1360 describing entire-copy, but it is possible with some effort and
1361 straining to read them the other way.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1362
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1363 <p>It might therefore be unwise to rely on either behavior in
1364 programs that might have to be linked with other curses
1365 implementations. Instead, you can do an explicit
1366 <code>touchwin()</code> before the <code>wnoutrefresh()</code>
1367 call to guarantee an entire-contents copy anywhere.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1368
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1369 <p>The really clean way to handle this is to use the panels
1370 library. If, when you want a screen update, you do
1371 <code>update_panels()</code>, it will do all the necessary
1372 <code>wnoutrefresh()</code> calls for whatever panel stacking
1373 order you have defined. Then you can do one
1374 <code>doupdate()</code> and there will be a <em>single</em> burst
1375 of physical I/O that will do all your updates.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1376
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1377 <h4><a name="backbug" id="backbug">Background Erase</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1378
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1379 <p>If you have been using a very old versions of
1380 <code>ncurses</code> (1.8.7 or older) you may be surprised by the
1381 behavior of the erase functions. In older versions, erased areas
1382 of a window were filled with a blank modified by the window's
1383 current attribute (as set by <strong>wattrset()</strong>,
1384 <strong>wattron()</strong>, <strong>wattroff()</strong> and
1385 friends).</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1386
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1387 <p>In newer versions, this is not so. Instead, the attribute of
1388 erased blanks is normal unless and until it is modified by the
1389 functions <code>bkgdset()</code> or <code>wbkgdset()</code>.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1390
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1391 <p>This change in behavior conforms <code>ncurses</code> to
1392 System V Release 4 and the XSI Curses standard.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1393
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1394 <h3><a name="xsifuncs" id="xsifuncs">XSI Curses Conformance</a></h3>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1395
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1396 <p>The <code>ncurses</code> library is intended to be base-level
1397 conformant with the XSI Curses standard from X/Open. Many
1398 extended-level features (in fact, almost all features not
1399 directly concerned with wide characters and internationalization)
1400 are also supported.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1401
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1402 <p>One effect of XSI conformance is the change in behavior
1403 described under <a href="#backbug">"Background Erase --
1404 Compatibility with Old Versions"</a>.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1405
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1406 <p>Also, <code>ncurses</code> meets the XSI requirement that
1407 every macro entry point have a corresponding function which may
1408 be linked (and will be prototype-checked) if the macro definition
1409 is disabled with <code>#undef</code>.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1410
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1411 <h2><a name="panels" id="panels">The Panels Library</a></h2>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1412
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1413 <p>The <code>ncurses</code> library by itself provides good
1414 support for screen displays in which the windows are tiled
1415 (non-overlapping). In the more general case that windows may
1416 overlap, you have to use a series of <code>wnoutrefresh()</code>
1417 calls followed by a <code>doupdate()</code>, and be careful about
1418 the order you do the window refreshes in. It has to be
1419 bottom-upwards, otherwise parts of windows that should be
1420 obscured will show through.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1421
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1422 <p>When your interface design is such that windows may dive
1423 deeper into the visibility stack or pop to the top at runtime,
1424 the resulting book-keeping can be tedious and difficult to get
1425 right. Hence the panels library.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1426
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1427 <p>The <code>panel</code> library first appeared in AT&amp;T
1428 System V. The version documented here is the <code>panel</code>
1429 code distributed with <code>ncurses</code>.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1430
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1431 <h3><a name="pcompile" id="pcompile">Compiling With the Panels
1432 Library</a></h3>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1433
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1434 <p>Your panels-using modules must import the panels library
1435 declarations with</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1436
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1437 <pre class="code-block">
1438 #include &lt;panel.h&gt;
1439</pre>
1440 <p>and must be linked explicitly with the panels library using an
1441 <code>-lpanel</code> argument. Note that they must also link the
1442 <code>ncurses</code> library with <code>-lncurses</code>. Many
1443 linkers are two-pass and will accept either order, but it is
1444 still good practice to put <code>-lpanel</code> first and
1445 <code>-lncurses</code> second.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1446
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1447 <h3><a name="poverview" id="poverview">Overview of Panels</a></h3>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1448
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1449 <p>A panel object is a window that is implicitly treated as part
1450 of a <dfn>deck</dfn> including all other panel objects. The deck
1451 has an implicit bottom-to-top visibility order. The panels
1452 library includes an update function (analogous to
1453 <code>refresh()</code>) that displays all panels in the deck in
1454 the proper order to resolve overlaps. The standard window,
1455 <code>stdscr</code>, is considered below all panels.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1456
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1457 <p>Details on the panels functions are available in the man
1458 pages. We will just hit the highlights here.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1459
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1460 <p>You create a panel from a window by calling
1461 <code>new_panel()</code> on a window pointer. It then becomes the
1462 top of the deck. The panel's window is available as the value of
1463 <code>panel_window()</code> called with the panel pointer as
1464 argument.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1465
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1466 <p>You can delete a panel (removing it from the deck) with
1467 <code>del_panel</code>. This will not deallocate the associated
1468 window; you have to do that yourself. You can replace a panel's
1469 window with a different window by calling
1470 <code>replace_window</code>. The new window may be of different
1471 size; the panel code will re-compute all overlaps. This operation
1472 does not change the panel's position in the deck.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1473
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1474 <p>To move a panel's window, use <code>move_panel()</code>. The
1475 <code>mvwin()</code> function on the panel's window is not
1476 sufficient because it does not update the panels library's
1477 representation of where the windows are. This operation leaves
1478 the panel's depth, contents, and size unchanged.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1479
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1480 <p>Two functions (<code>top_panel()</code>,
1481 <code>bottom_panel()</code>) are provided for rearranging the
1482 deck. The first pops its argument window to the top of the deck;
1483 the second sends it to the bottom. Either operation leaves the
1484 panel's screen location, contents, and size unchanged.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1485
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1486 <p>The function <code>update_panels()</code> does all the
1487 <code>wnoutrefresh()</code> calls needed to prepare for
1488 <code>doupdate()</code> (which you must call yourself,
1489 afterwards).</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1490
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1491 <p>Typically, you will want to call <code>update_panels()</code>
1492 and <code>doupdate()</code> just before accepting command input,
1493 once in each cycle of interaction with the user. If you call
1494 <code>update_panels()</code> after each and every panel write,
1495 you will generate a lot of unnecessary refresh activity and
1496 screen flicker.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1497
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1498 <h3><a name="pstdscr" id="pstdscr">Panels, Input, and the
1499 Standard Screen</a></h3>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1500
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1501 <p>You should not mix <code>wnoutrefresh()</code> or
1502 <code>wrefresh()</code> operations with panels code; this will
1503 work only if the argument window is either in the top panel or
1504 unobscured by any other panels.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1505
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1506 <p>The <code>stsdcr</code> window is a special case. It is
1507 considered below all panels. Because changes to panels may
1508 obscure parts of <code>stdscr</code>, though, you should call
1509 <code>update_panels()</code> before <code>doupdate()</code> even
1510 when you only change <code>stdscr</code>.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1511
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1512 <p>Note that <code>wgetch</code> automatically calls
1513 <code>wrefresh</code>. Therefore, before requesting input from a
1514 panel window, you need to be sure that the panel is totally
1515 unobscured.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1516
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1517 <p>There is presently no way to display changes to one obscured
1518 panel without repainting all panels.</p>
1519
1520 <h3><a name="hiding" id="hiding">Hiding Panels</a></h3>
1521
1522 <p>It is possible to remove a panel from the deck temporarily;
1523 use <code>hide_panel</code> for this. Use
1524 <code>show_panel()</code> to render it visible again. The
1525 predicate function <code>panel_hidden</code> tests whether or not
1526 a panel is hidden.</p>
1527
1528 <p>The <code>panel_update</code> code ignores hidden panels. You
1529 cannot do <code>top_panel()</code> or <code>bottom_panel</code>
1530 on a hidden panel(). Other panels operations are applicable.</p>
1531
1532 <h3><a name="pmisc" id="pmisc">Miscellaneous Other Facilities</a></h3>
1533
1534 <p>It is possible to navigate the deck using the functions
1535 <code>panel_above()</code> and <code>panel_below</code>. Handed a
1536 panel pointer, they return the panel above or below that panel.
1537 Handed <code>NULL</code>, they return the bottom-most or top-most
1538 panel.</p>
1539
1540 <p>Every panel has an associated user pointer, not used by the
1541 panel code, to which you can attach application data. See the man
1542 page documentation of <code>set_panel_userptr()</code> and
1543 <code>panel_userptr</code> for details.</p>
1544
1545 <h2><a name="menu" id="menu">The Menu Library</a></h2>
1546
1547 <p>A menu is a screen display that assists the user to choose
1548 some subset of a given set of items. The <code>menu</code>
1549 library is a curses extension that supports easy programming of
1550 menu hierarchies with a uniform but flexible interface.</p>
1551
1552 <p>The <code>menu</code> library first appeared in AT&amp;T
1553 System V. The version documented here is the <code>menu</code>
1554 code distributed with <code>ncurses</code>.</p>
1555
1556 <h3><a name="mcompile" id="mcompile">Compiling With the menu
1557 Library</a></h3>
1558
1559 <p>Your menu-using modules must import the menu library
1560 declarations with</p>
1561
1562 <pre class="code-block">
1563 #include &lt;menu.h&gt;
1564</pre>
1565 <p>and must be linked explicitly with the menus library using an
1566 <code>-lmenu</code> argument. Note that they must also link the
1567 <code>ncurses</code> library with <code>-lncurses</code>. Many
1568 linkers are two-pass and will accept either order, but it is
1569 still good practice to put <code>-lmenu</code> first and
1570 <code>-lncurses</code> second.</p>
1571
1572 <h3><a name="moverview" id="moverview">Overview of Menus</a></h3>
1573
1574 <p>The menus created by this library consist of collections of
1575 <dfn>items</dfn> including a name string part and a description
1576 string part. To make menus, you create groups of these items and
1577 connect them with menu frame objects.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1578
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1579 <p>The menu can then by <dfn>posted</dfn>, that is written to an
1580 associated window. Actually, each menu has two associated
1581 windows; a containing window in which the programmer can scribble
1582 titles or borders, and a subwindow in which the menu items proper
1583 are displayed. If this subwindow is too small to display all the
1584 items, it will be a scrollable viewport on the collection of
1585 items.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1586
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1587 <p>A menu may also be <dfn>unposted</dfn> (that is, undisplayed),
1588 and finally freed to make the storage associated with it and its
1589 items available for re-use.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1590
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1591 <p>The general flow of control of a menu program looks like
1592 this:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1593
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1594 <ol>
1595 <li>Initialize <code>curses</code>.</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1596
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1597 <li>Create the menu items, using <code>new_item()</code>.</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1598
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1599 <li>Create the menu using <code>new_menu()</code>.</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1600
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1601 <li>Post the menu using <code>post_menu()</code>.</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1602
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1603 <li>Refresh the screen.</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1604
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1605 <li>Process user requests via an input loop.</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1606
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1607 <li>Unpost the menu using <code>unpost_menu()</code>.</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1608
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1609 <li>Free the menu, using <code>free_menu()</code>.</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1610
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1611 <li>Free the items using <code>free_item()</code>.</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1612
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1613 <li>Terminate <code>curses</code>.</li>
1614 </ol>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1615
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1616 <h3><a name="mselect" id="mselect">Selecting items</a></h3>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1617
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1618 <p>Menus may be multi-valued or (the default) single-valued (see
1619 the manual page <code>menu_opts(3x)</code> to see how to change
1620 the default). Both types always have a <dfn>current
1621 item</dfn>.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1622
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1623 <p>From a single-valued menu you can read the selected value
1624 simply by looking at the current item. From a multi-valued menu,
1625 you get the selected set by looping through the items applying
1626 the <code>item_value()</code> predicate function. Your
1627 menu-processing code can use the function
1628 <code>set_item_value()</code> to flag the items in the select
1629 set.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1630
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1631 <p>Menu items can be made unselectable using
1632 <code>set_item_opts()</code> or <code>item_opts_off()</code> with
1633 the <code>O_SELECTABLE</code> argument. This is the only option
1634 so far defined for menus, but it is good practice to code as
1635 though other option bits might be on.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1636
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1637 <h3><a name="mdisplay" id="mdisplay">Menu Display</a></h3>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1638
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1639 <p>The menu library calculates a minimum display size for your
1640 window, based on the following variables:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1641
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1642 <ul>
1643 <li>The number and maximum length of the menu items</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1644
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1645 <li>Whether the O_ROWMAJOR option is enabled</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1646
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1647 <li>Whether display of descriptions is enabled</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1648
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1649 <li>Whatever menu format may have been set by the
1650 programmer</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1651
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1652 <li>The length of the menu mark string used for highlighting
1653 selected items</li>
1654 </ul>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1655
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1656 <p>The function <code>set_menu_format()</code> allows you to set
1657 the maximum size of the viewport or <dfn>menu page</dfn> that
1658 will be used to display menu items. You can retrieve any format
1659 associated with a menu with <code>menu_format()</code>. The
1660 default format is rows=16, columns=1.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1661
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1662 <p>The actual menu page may be smaller than the format size. This
1663 depends on the item number and size and whether O_ROWMAJOR is on.
1664 This option (on by default) causes menu items to be displayed in
1665 a &ldquo;raster-scan&rdquo; pattern, so that if more than one
1666 item will fit horizontally the first couple of items are
1667 side-by-side in the top row. The alternative is column-major
1668 display, which tries to put the first several items in the first
1669 column.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1670
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1671 <p>As mentioned above, a menu format not large enough to allow
1672 all items to fit on-screen will result in a menu display that is
1673 vertically scrollable.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1674
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1675 <p>You can scroll it with requests to the menu driver, which will
1676 be described in the section on <a href="#minput">menu input
1677 handling</a>.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1678
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1679 <p>Each menu has a <dfn>mark string</dfn> used to visually tag
1680 selected items; see the <code>menu_mark(3x)</code> manual page
1681 for details. The mark string length also influences the menu page
1682 size.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1683
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1684 <p>The function <code>scale_menu()</code> returns the minimum
1685 display size that the menu code computes from all these factors.
1686 There are other menu display attributes including a select
1687 attribute, an attribute for selectable items, an attribute for
1688 unselectable items, and a pad character used to separate item
1689 name text from description text. These have reasonable defaults
1690 which the library allows you to change (see the
1691 <code>menu_attribs(3x)</code> manual page.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1692
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1693 <h3><a name="mwindows" id="mwindows">Menu Windows</a></h3>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1694
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1695 <p>Each menu has, as mentioned previously, a pair of associated
1696 windows. Both these windows are painted when the menu is posted
1697 and erased when the menu is unposted.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1698
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1699 <p>The outer or frame window is not otherwise touched by the menu
1700 routines. It exists so the programmer can associate a title, a
1701 border, or perhaps help text with the menu and have it properly
1702 refreshed or erased at post/unpost time. The inner window or
1703 <dfn>subwindow</dfn> is where the current menu page is
1704 displayed.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1705
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1706 <p>By default, both windows are <code>stdscr</code>. You can set
1707 them with the functions in <code>menu_win(3x)</code>.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1708
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1709 <p>When you call <code>post_menu()</code>, you write the menu to
1710 its subwindow. When you call <code>unpost_menu()</code>, you
1711 erase the subwindow, However, neither of these actually modifies
1712 the screen. To do that, call <code>wrefresh()</code> or some
1713 equivalent.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1714
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1715 <h3><a name="minput" id="minput">Processing Menu Input</a></h3>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1716
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1717 <p>The main loop of your menu-processing code should call
1718 <code>menu_driver()</code> repeatedly. The first argument of this
1719 routine is a menu pointer; the second is a menu command code. You
1720 should write an input-fetching routine that maps input characters
1721 to menu command codes, and pass its output to
1722 <code>menu_driver()</code>. The menu command codes are fully
1723 documented in <code>menu_driver(3x)</code>.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1724
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1725 <p>The simplest group of command codes is
1726 <code>REQ_NEXT_ITEM</code>, <code>REQ_PREV_ITEM</code>,
1727 <code>REQ_FIRST_ITEM</code>, <code>REQ_LAST_ITEM</code>,
1728 <code>REQ_UP_ITEM</code>, <code>REQ_DOWN_ITEM</code>,
1729 <code>REQ_LEFT_ITEM</code>, <code>REQ_RIGHT_ITEM</code>. These
1730 change the currently selected item. These requests may cause
1731 scrolling of the menu page if it only partially displayed.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1732
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1733 <p>There are explicit requests for scrolling which also change
1734 the current item (because the select location does not change,
1735 but the item there does). These are <code>REQ_SCR_DLINE</code>,
1736 <code>REQ_SCR_ULINE</code>, <code>REQ_SCR_DPAGE</code>, and
1737 <code>REQ_SCR_UPAGE</code>.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1738
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1739 <p>The <code>REQ_TOGGLE_ITEM</code> selects or deselects the
1740 current item. It is for use in multi-valued menus; if you use it
1741 with <code>O_ONEVALUE</code> on, you will get an error return
1742 (<code>E_REQUEST_DENIED</code>).</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1743
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1744 <p>Each menu has an associated pattern buffer. The
1745 <code>menu_driver()</code> logic tries to accumulate printable
1746 ASCII characters passed in in that buffer; when it matches a
1747 prefix of an item name, that item (or the next matching item) is
1748 selected. If appending a character yields no new match, that
1749 character is deleted from the pattern buffer, and
1750 <code>menu_driver()</code> returns <code>E_NO_MATCH</code>.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1751
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1752 <p>Some requests change the pattern buffer directly:
1753 <code>REQ_CLEAR_PATTERN</code>, <code>REQ_BACK_PATTERN</code>,
1754 <code>REQ_NEXT_MATCH</code>, <code>REQ_PREV_MATCH</code>. The
1755 latter two are useful when pattern buffer input matches more than
1756 one item in a multi-valued menu.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1757
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1758 <p>Each successful scroll or item navigation request clears the
1759 pattern buffer. It is also possible to set the pattern buffer
1760 explicitly with <code>set_menu_pattern()</code>.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1761
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1762 <p>Finally, menu driver requests above the constant
1763 <code>MAX_COMMAND</code> are considered application-specific
1764 commands. The <code>menu_driver()</code> code ignores them and
1765 returns <code>E_UNKNOWN_COMMAND</code>.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1766
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1767 <h3><a name="mmisc" id="mmisc">Miscellaneous Other Features</a></h3>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1768
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1769 <p>Various menu options can affect the processing and visual
1770 appearance and input processing of menus. See <code>menu_opts(3x)
1771 for details.</code></p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1772
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1773 <p>It is possible to change the current item from application
1774 code; this is useful if you want to write your own navigation
1775 requests. It is also possible to explicitly set the top row of
1776 the menu display. See <code>mitem_current(3x)</code>. If your
1777 application needs to change the menu subwindow cursor for any
1778 reason, <code>pos_menu_cursor()</code> will restore it to the
1779 correct location for continuing menu driver processing.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1780
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1781 <p>It is possible to set hooks to be called at menu
1782 initialization and wrapup time, and whenever the selected item
1783 changes. See <code>menu_hook(3x)</code>.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1784
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1785 <p>Each item, and each menu, has an associated user pointer on
1786 which you can hang application data. See
1787 <code>mitem_userptr(3x)</code> and
1788 <code>menu_userptr(3x)</code>.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1789
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1790 <h2><a name="form" id="form">The Forms Library</a></h2>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1791
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1792 <p>The <code>form</code> library is a curses extension that
1793 supports easy programming of on-screen forms for data entry and
1794 program control.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1795
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1796 <p>The <code>form</code> library first appeared in AT&amp;T
1797 System V. The version documented here is the <code>form</code>
1798 code distributed with <code>ncurses</code>.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1799
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1800 <h3><a name="fcompile" id="fcompile">Compiling With the form
1801 Library</a></h3>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1802
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1803 <p>Your form-using modules must import the form library
1804 declarations with</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1805
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1806 <pre class="code-block">
1807 #include &lt;form.h&gt;
1808</pre>
1809 <p>and must be linked explicitly with the forms library using an
1810 <code>-lform</code> argument. Note that they must also link the
1811 <code>ncurses</code> library with <code>-lncurses</code>. Many
1812 linkers are two-pass and will accept either order, but it is
1813 still good practice to put <code>-lform</code> first and
1814 <code>-lncurses</code> second.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1815
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1816 <h3><a name="foverview" id="foverview">Overview of Forms</a></h3>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1817
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1818 <p>A form is a collection of fields; each field may be either a
1819 label (explanatory text) or a data-entry location. Long forms may
1820 be segmented into pages; each entry to a new page clears the
1821 screen.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1822
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1823 <p>To make forms, you create groups of fields and connect them
1824 with form frame objects; the form library makes this relatively
1825 simple.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1826
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1827 <p>Once defined, a form can be <dfn>posted</dfn>, that is written
1828 to an associated window. Actually, each form has two associated
1829 windows; a containing window in which the programmer can scribble
1830 titles or borders, and a subwindow in which the form fields
1831 proper are displayed.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1832
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1833 <p>As the form user fills out the posted form, navigation and
1834 editing keys support movement between fields, editing keys
1835 support modifying field, and plain text adds to or changes data
1836 in a current field. The form library allows you (the forms
1837 designer) to bind each navigation and editing key to any
1838 keystroke accepted by <code>curses</code> Fields may have
1839 validation conditions on them, so that they check input data for
1840 type and value. The form library supplies a rich set of
1841 pre-defined field types, and makes it relatively easy to define
1842 new ones.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1843
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1844 <p>Once its transaction is completed (or aborted), a form may be
1845 <dfn>unposted</dfn> (that is, undisplayed), and finally freed to
1846 make the storage associated with it and its items available for
1847 re-use.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1848
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1849 <p>The general flow of control of a form program looks like
1850 this:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1851
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1852 <ol>
1853 <li>Initialize <code>curses</code>.</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1854
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1855 <li>Create the form fields, using
1856 <code>new_field()</code>.</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1857
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1858 <li>Create the form using <code>new_form()</code>.</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1859
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1860 <li>Post the form using <code>post_form()</code>.</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1861
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1862 <li>Refresh the screen.</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1863
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1864 <li>Process user requests via an input loop.</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1865
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1866 <li>Unpost the form using <code>unpost_form()</code>.</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1867
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1868 <li>Free the form, using <code>free_form()</code>.</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1869
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1870 <li>Free the fields using <code>free_field()</code>.</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1871
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1872 <li>Terminate <code>curses</code>.</li>
1873 </ol>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1874
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1875 <p>Note that this looks much like a menu program; the form
1876 library handles tasks which are in many ways similar, and its
1877 interface was obviously designed to resemble that of the <a href=
1878 "#menu">menu library</a> wherever possible.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1879
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1880 <p>In forms programs, however, the &ldquo;process user
1881 requests&rdquo; is somewhat more complicated than for menus.
1882 Besides menu-like navigation operations, the menu driver loop has
1883 to support field editing and data validation.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1884
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1885 <h3><a name="fcreate" id="fcreate">Creating and Freeing Fields
1886 and Forms</a></h3>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1887
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1888 <p>The basic function for creating fields is
1889 <code>new_field()</code>:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1890
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1891 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1892FIELD *new_field(int height, int width, /* new field size */
1893 int top, int left, /* upper left corner */
1894 int offscreen, /* number of offscreen rows */
1895 int nbuf); /* number of working buffers */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1896</pre>
1897 <p>Menu items always occupy a single row, but forms fields may
1898 have multiple rows. So <code>new_field()</code> requires you to
1899 specify a width and height (the first two arguments, which mist
1900 both be greater than zero).</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1901
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1902 <p>You must also specify the location of the field's upper left
1903 corner on the screen (the third and fourth arguments, which must
1904 be zero or greater). Note that these coordinates are relative to
1905 the form subwindow, which will coincide with <code>stdscr</code>
1906 by default but need not be <code>stdscr</code> if you have done
1907 an explicit <code>set_form_win()</code> call.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1908
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1909 <p>The fifth argument allows you to specify a number of
1910 off-screen rows. If this is zero, the entire field will always be
1911 displayed. If it is nonzero, the form will be scrollable, with
1912 only one screen-full (initially the top part) displayed at any
1913 given time. If you make a field dynamic and grow it so it will no
1914 longer fit on the screen, the form will become scrollable even if
1915 the <code>offscreen</code> argument was initially zero.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1916
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1917 <p>The forms library allocates one working buffer per field; the
1918 size of each buffer is <code>((height + offscreen)*width +
1919 1</code>, one character for each position in the field plus a NUL
1920 terminator. The sixth argument is the number of additional data
1921 buffers to allocate for the field; your application can use them
1922 for its own purposes.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1923
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1924 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1925FIELD *dup_field(FIELD *field, /* field to copy */
1926 int top, int left); /* location of new copy */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1927</pre>
1928 <p>The function <code>dup_field()</code> duplicates an existing
1929 field at a new location. Size and buffering information are
1930 copied; some attribute flags and status bits are not (see the
1931 <code>form_field_new(3X)</code> for details).</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1932
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1933 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1934FIELD *link_field(FIELD *field, /* field to copy */
1935 int top, int left); /* location of new copy */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1936</pre>
1937 <p>The function <code>link_field()</code> also duplicates an
1938 existing field at a new location. The difference from
1939 <code>dup_field()</code> is that it arranges for the new field's
1940 buffer to be shared with the old one.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1941
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1942 <p>Besides the obvious use in making a field editable from two
1943 different form pages, linked fields give you a way to hack in
1944 dynamic labels. If you declare several fields linked to an
1945 original, and then make them inactive, changes from the original
1946 will still be propagated to the linked fields.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1947
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1948 <p>As with duplicated fields, linked fields have attribute bits
1949 separate from the original.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1950
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1951 <p>As you might guess, all these field-allocations return
1952 <code>NULL</code> if the field allocation is not possible due to
1953 an out-of-memory error or out-of-bounds arguments.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1954
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1955 <p>To connect fields to a form, use</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1956
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1957 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1958FORM *new_form(FIELD **fields);
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1959</pre>
1960 <p>This function expects to see a NULL-terminated array of field
1961 pointers. Said fields are connected to a newly-allocated form
1962 object; its address is returned (or else NULL if the allocation
1963 fails).</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1964
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1965 <p>Note that <code>new_field()</code> does <em>not</em> copy the
1966 pointer array into private storage; if you modify the contents of
1967 the pointer array during forms processing, all manner of bizarre
1968 things might happen. Also note that any given field may only be
1969 connected to one form.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1970
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1971 <p>The functions <code>free_field()</code> and
1972 <code>free_form</code> are available to free field and form
1973 objects. It is an error to attempt to free a field connected to a
1974 form, but not vice-versa; thus, you will generally free your form
1975 objects first.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1976
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1977 <h3><a name="fattributes" id="fattributes">Fetching and Changing
1978 Field Attributes</a></h3>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1979
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1980 <p>Each form field has a number of location and size attributes
1981 associated with it. There are other field attributes used to
1982 control display and editing of the field. Some (for example, the
1983 <code>O_STATIC</code> bit) involve sufficient complications to be
1984 covered in sections of their own later on. We cover the functions
1985 used to get and set several basic attributes here.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1986
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1987 <p>When a field is created, the attributes not specified by the
1988 <code>new_field</code> function are copied from an invisible
1989 system default field. In attribute-setting and -fetching
1990 functions, the argument NULL is taken to mean this field. Changes
1991 to it persist as defaults until your forms application
1992 terminates.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1993
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1994 <h4><a name="fsizes" id="fsizes">Fetching Size and Location
1995 Data</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1996
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1997 <p>You can retrieve field sizes and locations through:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]1998
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]1999 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2000int field_info(FIELD *field, /* field from which to fetch */
2001 int *height, *int width, /* field size */
2002 int *top, int *left, /* upper left corner */
2003 int *offscreen, /* number of offscreen rows */
2004 int *nbuf); /* number of working buffers */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2005</pre>
2006 <p>This function is a sort of inverse of
2007 <code>new_field()</code>; instead of setting size and location
2008 attributes of a new field, it fetches them from an existing
2009 one.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2010
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2011 <h4><a name="flocation" id="flocation">Changing the Field
2012 Location</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2013
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2014 <p>It is possible to move a field's location on the screen:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2015
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2016 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2017int move_field(FIELD *field, /* field to alter */
2018 int top, int left); /* new upper-left corner */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2019</pre>
2020 <p>You can, of course. query the current location through
2021 <code>field_info()</code>.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2022
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2023 <h4><a name="fjust" id="fjust">The Justification Attribute</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2024
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2025 <p>One-line fields may be unjustified, justified right, justified
2026 left, or centered. Here is how you manipulate this attribute:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2027
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2028 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2029int set_field_just(FIELD *field, /* field to alter */
2030 int justmode); /* mode to set */
2031
2032int field_just(FIELD *field); /* fetch mode of field */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2033</pre>
2034 <p>The mode values accepted and returned by this functions are
2035 preprocessor macros <code>NO_JUSTIFICATION</code>,
2036 <code>JUSTIFY_RIGHT</code>, <code>JUSTIFY_LEFT</code>, or
2037 <code>JUSTIFY_CENTER</code>.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2038
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2039 <h4><a name="fdispatts" id="fdispatts">Field Display
2040 Attributes</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2041
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2042 <p>For each field, you can set a foreground attribute for entered
2043 characters, a background attribute for the entire field, and a
2044 pad character for the unfilled portion of the field. You can also
2045 control pagination of the form.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2046
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2047 <p>This group of four field attributes controls the visual
2048 appearance of the field on the screen, without affecting in any
2049 way the data in the field buffer.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2050
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2051 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2052int set_field_fore(FIELD *field, /* field to alter */
2053 chtype attr); /* attribute to set */
2054
2055chtype field_fore(FIELD *field); /* field to query */
2056
2057int set_field_back(FIELD *field, /* field to alter */
2058 chtype attr); /* attribute to set */
2059
2060chtype field_back(FIELD *field); /* field to query */
2061
2062int set_field_pad(FIELD *field, /* field to alter */
2063 int pad); /* pad character to set */
2064
2065chtype field_pad(FIELD *field);
2066
2067int set_new_page(FIELD *field, /* field to alter */
2068 int flag); /* TRUE to force new page */
2069
2070chtype new_page(FIELD *field); /* field to query */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2071</pre>
2072 <p>The attributes set and returned by the first four functions
2073 are normal <code>curses(3x)</code> display attribute values
2074 (<code>A_STANDOUT</code>, <code>A_BOLD</code>,
2075 <code>A_REVERSE</code> etc). The page bit of a field controls
2076 whether it is displayed at the start of a new form screen.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2077
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2078 <h4><a name="foptions" id="foptions">Field Option Bits</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2079
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2080 <p>There is also a large collection of field option bits you can
2081 set to control various aspects of forms processing. You can
2082 manipulate them with these functions:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2083
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2084 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2085int set_field_opts(FIELD *field, /* field to alter */
2086 int attr); /* attribute to set */
2087
2088int field_opts_on(FIELD *field, /* field to alter */
2089 int attr); /* attributes to turn on */
2090
2091int field_opts_off(FIELD *field, /* field to alter */
2092 int attr); /* attributes to turn off */
2093
2094int field_opts(FIELD *field); /* field to query */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2095</pre>
2096 <p>By default, all options are on. Here are the available option
2097 bits:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2098
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2099 <dl>
2100 <dt>O_VISIBLE</dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2101
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2102 <dd>Controls whether the field is visible on the screen. Can be
2103 used during form processing to hide or pop up fields depending
2104 on the value of parent fields.</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2105
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2106 <dt>O_ACTIVE</dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2107
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2108 <dd>Controls whether the field is active during forms
2109 processing (i.e. visited by form navigation keys). Can be used
2110 to make labels or derived fields with buffer values alterable
2111 by the forms application, not the user.</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2112
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2113 <dt>O_PUBLIC</dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2114
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2115 <dd>Controls whether data is displayed during field entry. If
2116 this option is turned off on a field, the library will accept
2117 and edit data in that field, but it will not be displayed and
2118 the visible field cursor will not move. You can turn off the
2119 O_PUBLIC bit to define password fields.</dd>
2120
2121 <dt>O_EDIT</dt>
2122
2123 <dd>Controls whether the field's data can be modified. When
2124 this option is off, all editing requests except
2125 <code>REQ_PREV_CHOICE</code> and <code>REQ_NEXT_CHOICE</code>
2126 will fail. Such read-only fields may be useful for help
2127 messages.</dd>
2128
2129 <dt>O_WRAP</dt>
2130
2131 <dd>Controls word-wrapping in multi-line fields. Normally, when
2132 any character of a (blank-separated) word reaches the end of
2133 the current line, the entire word is wrapped to the next line
2134 (assuming there is one). When this option is off, the word will
2135 be split across the line break.</dd>
2136
2137 <dt>O_BLANK</dt>
2138
2139 <dd>Controls field blanking. When this option is on, entering a
2140 character at the first field position erases the entire field
2141 (except for the just-entered character).</dd>
2142
2143 <dt>O_AUTOSKIP</dt>
2144
2145 <dd>Controls automatic skip to next field when this one fills.
2146 Normally, when the forms user tries to type more data into a
2147 field than will fit, the editing location jumps to next field.
2148 When this option is off, the user's cursor will hang at the end
2149 of the field. This option is ignored in dynamic fields that
2150 have not reached their size limit.</dd>
2151
2152 <dt>O_NULLOK</dt>
2153
2154 <dd>Controls whether <a href="#fvalidation">validation</a> is
2155 applied to blank fields. Normally, it is not; the user can
2156 leave a field blank without invoking the usual validation check
2157 on exit. If this option is off on a field, exit from it will
2158 invoke a validation check.</dd>
2159
2160 <dt>O_PASSOK</dt>
2161
2162 <dd>Controls whether validation occurs on every exit, or only
2163 after the field is modified. Normally the latter is true.
2164 Setting O_PASSOK may be useful if your field's validation
2165 function may change during forms processing.</dd>
2166
2167 <dt>O_STATIC</dt>
2168
2169 <dd>Controls whether the field is fixed to its initial
2170 dimensions. If you turn this off, the field becomes <a href=
2171 "#fdynamic">dynamic</a> and will stretch to fit entered
2172 data.</dd>
2173 </dl>
2174
2175 <p>A field's options cannot be changed while the field is
2176 currently selected. However, options may be changed on posted
2177 fields that are not current.</p>
2178
2179 <p>The option values are bit-masks and can be composed with
2180 logical-or in the obvious way.</p>
2181
2182 <h3><a name="fstatus" id="fstatus">Field Status</a></h3>
2183
2184 <p>Every field has a status flag, which is set to FALSE when the
2185 field is created and TRUE when the value in field buffer 0
2186 changes. This flag can be queried and set directly:</p>
2187
2188 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2189int set_field_status(FIELD *field, /* field to alter */
2190 int status); /* mode to set */
2191
2192int field_status(FIELD *field); /* fetch mode of field */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2193</pre>
2194 <p>Setting this flag under program control can be useful if you
2195 use the same form repeatedly, looking for modified fields each
2196 time.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2197
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2198 <p>Calling <code>field_status()</code> on a field not currently
2199 selected for input will return a correct value. Calling
2200 <code>field_status()</code> on a field that is currently selected
2201 for input may not necessarily give a correct field status value,
2202 because entered data is not necessarily copied to buffer zero
2203 before the exit validation check. To guarantee that the returned
2204 status value reflects reality, call <code>field_status()</code>
2205 either (1) in the field's exit validation check routine, (2) from
2206 the field's or form's initialization or termination hooks, or (3)
2207 just after a <code>REQ_VALIDATION</code> request has been
2208 processed by the forms driver.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2209
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2210 <h3><a name="fuser" id="fuser">Field User Pointer</a></h3>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2211
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2212 <p>Each field structure contains one character pointer slot that
2213 is not used by the forms library. It is intended to be used by
2214 applications to store private per-field data. You can manipulate
2215 it with:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2216
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2217 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2218int set_field_userptr(FIELD *field, /* field to alter */
2219 char *userptr); /* mode to set */
2220
2221char *field_userptr(FIELD *field); /* fetch mode of field */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2222</pre>(Properly, this user pointer field ought to have <code>(void
2223*)</code> type. The <code>(char *)</code> type is retained for
2224System V compatibility.)
2225 <p>It is valid to set the user pointer of the default field (with
2226 a <code>set_field_userptr()</code> call passed a NULL field
2227 pointer.) When a new field is created, the default-field user
2228 pointer is copied to initialize the new field's user pointer.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2229
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2230 <h3><a name="fdynamic" id="fdynamic">Variable-Sized Fields</a></h3>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2231
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2232 <p>Normally, a field is fixed at the size specified for it at
2233 creation time. If, however, you turn off its O_STATIC bit, it
2234 becomes <dfn>dynamic</dfn> and will automatically resize itself
2235 to accommodate data as it is entered. If the field has extra
2236 buffers associated with it, they will grow right along with the
2237 main input buffer.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2238
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2239 <p>A one-line dynamic field will have a fixed height (1) but
2240 variable width, scrolling horizontally to display data within the
2241 field area as originally dimensioned and located. A multi-line
2242 dynamic field will have a fixed width, but variable height
2243 (number of rows), scrolling vertically to display data within the
2244 field area as originally dimensioned and located.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2245
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2246 <p>Normally, a dynamic field is allowed to grow without limit.
2247 But it is possible to set an upper limit on the size of a dynamic
2248 field. You do it with this function:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2249
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2250 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2251int set_max_field(FIELD *field, /* field to alter (may not be NULL) */
2252 int max_size); /* upper limit on field size */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2253</pre>
2254 <p>If the field is one-line, <code>max_size</code> is taken to be
2255 a column size limit; if it is multi-line, it is taken to be a
2256 line size limit. To disable any limit, use an argument of zero.
2257 The growth limit can be changed whether or not the O_STATIC bit
2258 is on, but has no effect until it is.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2259
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2260 <p>The following properties of a field change when it becomes
2261 dynamic:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2262
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2263 <ul>
2264 <li>If there is no growth limit, there is no final position of
2265 the field; therefore <code>O_AUTOSKIP</code> and
2266 <code>O_NL_OVERLOAD</code> are ignored.</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2267
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2268 <li>Field justification will be ignored (though whatever
2269 justification is set up will be retained internally and can be
2270 queried).</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2271
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2272 <li>The <code>dup_field()</code> and <code>link_field()</code>
2273 calls copy dynamic-buffer sizes. If the <code>O_STATIC</code>
2274 option is set on one of a collection of links, buffer resizing
2275 will occur only when the field is edited through that
2276 link.</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2277
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2278 <li>The call <code>field_info()</code> will retrieve the
2279 original static size of the field; use
2280 <code>dynamic_field_info()</code> to get the actual dynamic
2281 size.</li>
2282 </ul>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2283
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2284 <h3><a name="fvalidation" id="fvalidation">Field Validation</a></h3>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2285
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2286 <p>By default, a field will accept any data that will fit in its
2287 input buffer. However, it is possible to attach a validation type
2288 to a field. If you do this, any attempt to leave the field while
2289 it contains data that does not match the validation type will
2290 fail. Some validation types also have a character-validity check
2291 for each time a character is entered in the field.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2292
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2293 <p>A field's validation check (if any) is not called when
2294 <code>set_field_buffer()</code> modifies the input buffer, nor
2295 when that buffer is changed through a linked field.</p>
2296
2297 <p>The <code>form</code> library provides a rich set of
2298 pre-defined validation types, and gives you the capability to
2299 define custom ones of your own. You can examine and change field
2300 validation attributes with the following functions:</p>
2301
2302 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2303int set_field_type(FIELD *field, /* field to alter */
2304 FIELDTYPE *ftype, /* type to associate */
2305 ...); /* additional arguments*/
2306
2307FIELDTYPE *field_type(FIELD *field); /* field to query */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2308</pre>
2309 <p>The validation type of a field is considered an attribute of
2310 the field. As with other field attributes, Also, doing
2311 <code>set_field_type()</code> with a <code>NULL</code> field
2312 default will change the system default for validation of
2313 newly-created fields.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2314
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2315 <p>Here are the pre-defined validation types:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2316
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2317 <h4><a name="ftype_alpha" id="ftype_alpha">TYPE_ALPHA</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2318
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2319 <p>This field type accepts alphabetic data; no blanks, no digits,
2320 no special characters (this is checked at character-entry time).
2321 It is set up with:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2322
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2323 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2324int set_field_type(FIELD *field, /* field to alter */
2325 TYPE_ALPHA, /* type to associate */
2326 int width); /* maximum width of field */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2327</pre>
2328 <p>The <code>width</code> argument sets a minimum width of data.
2329 Typically you will want to set this to the field width; if it is
2330 greater than the field width, the validation check will always
2331 fail. A minimum width of zero makes field completion
2332 optional.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2333
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2334 <h4><a name="ftype_alnum" id="ftype_alnum">TYPE_ALNUM</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2335
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2336 <p>This field type accepts alphabetic data and digits; no blanks,
2337 no special characters (this is checked at character-entry time).
2338 It is set up with:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2339
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2340 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2341int set_field_type(FIELD *field, /* field to alter */
2342 TYPE_ALNUM, /* type to associate */
2343 int width); /* maximum width of field */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2344</pre>
2345 <p>The <code>width</code> argument sets a minimum width of data.
2346 As with TYPE_ALPHA, typically you will want to set this to the
2347 field width; if it is greater than the field width, the
2348 validation check will always fail. A minimum width of zero makes
2349 field completion optional.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2350
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2351 <h4><a name="ftype_enum" id="ftype_enum">TYPE_ENUM</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2352
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2353 <p>This type allows you to restrict a field's values to be among
2354 a specified set of string values (for example, the two-letter
2355 postal codes for U.S. states). It is set up with:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2356
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2357 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2358int set_field_type(FIELD *field, /* field to alter */
2359 TYPE_ENUM, /* type to associate */
2360 char **valuelist; /* list of possible values */
2361 int checkcase; /* case-sensitive? */
2362 int checkunique); /* must specify uniquely? */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2363</pre>
2364 <p>The <code>valuelist</code> parameter must point at a
2365 NULL-terminated list of valid strings. The <code>checkcase</code>
2366 argument, if true, makes comparison with the string
2367 case-sensitive.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2368
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2369 <p>When the user exits a TYPE_ENUM field, the validation
2370 procedure tries to complete the data in the buffer to a valid
2371 entry. If a complete choice string has been entered, it is of
2372 course valid. But it is also possible to enter a prefix of a
2373 valid string and have it completed for you.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2374
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2375 <p>By default, if you enter such a prefix and it matches more
2376 than one value in the string list, the prefix will be completed
2377 to the first matching value. But the <code>checkunique</code>
2378 argument, if true, requires prefix matches to be unique in order
2379 to be valid.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2380
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2381 <p>The <code>REQ_NEXT_CHOICE</code> and
2382 <code>REQ_PREV_CHOICE</code> input requests can be particularly
2383 useful with these fields.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2384
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2385 <h4><a name="ftype_integer" id="ftype_integer">TYPE_INTEGER</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2386
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2387 <p>This field type accepts an integer. It is set up as
2388 follows:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2389
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2390 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2391int set_field_type(FIELD *field, /* field to alter */
2392 TYPE_INTEGER, /* type to associate */
2393 int padding, /* # places to zero-pad to */
2394 int vmin, int vmax); /* valid range */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2395</pre>
2396 <p>Valid characters consist of an optional leading minus and
2397 digits. The range check is performed on exit. If the range
2398 maximum is less than or equal to the minimum, the range is
2399 ignored.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2400
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2401 <p>If the value passes its range check, it is padded with as many
2402 leading zero digits as necessary to meet the padding
2403 argument.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2404
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2405 <p>A <code>TYPE_INTEGER</code> value buffer can conveniently be
2406 interpreted with the C library function <code>atoi(3)</code>.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2407
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2408 <h4><a name="ftype_numeric" id="ftype_numeric">TYPE_NUMERIC</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2409
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2410 <p>This field type accepts a decimal number. It is set up as
2411 follows:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2412
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2413 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2414int set_field_type(FIELD *field, /* field to alter */
2415 TYPE_NUMERIC, /* type to associate */
2416 int padding, /* # places of precision */
2417 double vmin, double vmax); /* valid range */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2418</pre>
2419 <p>Valid characters consist of an optional leading minus and
2420 digits. possibly including a decimal point. If your system
2421 supports locale's, the decimal point character used must be the
2422 one defined by your locale. The range check is performed on exit.
2423 If the range maximum is less than or equal to the minimum, the
2424 range is ignored.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2425
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2426 <p>If the value passes its range check, it is padded with as many
2427 trailing zero digits as necessary to meet the padding
2428 argument.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2429
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2430 <p>A <code>TYPE_NUMERIC</code> value buffer can conveniently be
2431 interpreted with the C library function <code>atof(3)</code>.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2432
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2433 <h4><a name="ftype_regexp" id="ftype_regexp">TYPE_REGEXP</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2434
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2435 <p>This field type accepts data matching a regular expression. It
2436 is set up as follows:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2437
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2438 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2439int set_field_type(FIELD *field, /* field to alter */
2440 TYPE_REGEXP, /* type to associate */
2441 char *regexp); /* expression to match */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2442</pre>
2443 <p>The syntax for regular expressions is that of
2444 <code>regcomp(3)</code>. The check for regular-expression match
2445 is performed on exit.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2446
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2447 <h3><a name="fbuffer" id="fbuffer">Direct Field Buffer
2448 Manipulation</a></h3>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2449
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2450 <p>The chief attribute of a field is its buffer contents. When a
2451 form has been completed, your application usually needs to know
2452 the state of each field buffer. You can find this out with:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2453
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2454 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2455char *field_buffer(FIELD *field, /* field to query */
2456 int bufindex); /* number of buffer to query */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2457</pre>
2458 <p>Normally, the state of the zero-numbered buffer for each field
2459 is set by the user's editing actions on that field. It is
2460 sometimes useful to be able to set the value of the zero-numbered
2461 (or some other) buffer from your application:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2462
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2463 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2464int set_field_buffer(FIELD *field, /* field to alter */
2465 int bufindex, /* number of buffer to alter */
2466 char *value); /* string value to set */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2467</pre>
2468 <p>If the field is not large enough and cannot be resized to a
2469 sufficiently large size to contain the specified value, the value
2470 will be truncated to fit.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2471
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2472 <p>Calling <code>field_buffer()</code> with a null field pointer
2473 will raise an error. Calling <code>field_buffer()</code> on a
2474 field not currently selected for input will return a correct
2475 value. Calling <code>field_buffer()</code> on a field that is
2476 currently selected for input may not necessarily give a correct
2477 field buffer value, because entered data is not necessarily
2478 copied to buffer zero before the exit validation check. To
2479 guarantee that the returned buffer value reflects on-screen
2480 reality, call <code>field_buffer()</code> either (1) in the
2481 field's exit validation check routine, (2) from the field's or
2482 form's initialization or termination hooks, or (3) just after a
2483 <code>REQ_VALIDATION</code> request has been processed by the
2484 forms driver.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2485
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2486 <h3><a name="formattrs" id="formattrs">Attributes of Forms</a></h3>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2487
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2488 <p>As with field attributes, form attributes inherit a default
2489 from a system default form structure. These defaults can be
2490 queried or set by of these functions using a form-pointer
2491 argument of <code>NULL</code>.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2492
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2493 <p>The principal attribute of a form is its field list. You can
2494 query and change this list with:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2495
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2496 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2497int set_form_fields(FORM *form, /* form to alter */
2498 FIELD **fields); /* fields to connect */
2499
2500char *form_fields(FORM *form); /* fetch fields of form */
2501
2502int field_count(FORM *form); /* count connect fields */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2503</pre>
2504 <p>The second argument of <code>set_form_fields()</code> may be a
2505 NULL-terminated field pointer array like the one required by
2506 <code>new_form()</code>. In that case, the old fields of the form
2507 are disconnected but not freed (and eligible to be connected to
2508 other forms), then the new fields are connected.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2509
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2510 <p>It may also be null, in which case the old fields are
2511 disconnected (and not freed) but no new ones are connected.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2512
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2513 <p>The <code>field_count()</code> function simply counts the
2514 number of fields connected to a given from. It returns -1 if the
2515 form-pointer argument is NULL.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2516
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2517 <h3><a name="fdisplay" id="fdisplay">Control of Form Display</a></h3>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2518
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2519 <p>In the overview section, you saw that to display a form you
2520 normally start by defining its size (and fields), posting it, and
2521 refreshing the screen. There is an hidden step before posting,
2522 which is the association of the form with a frame window
2523 (actually, a pair of windows) within which it will be displayed.
2524 By default, the forms library associates every form with the
2525 full-screen window <code>stdscr</code>.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2526
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2527 <p>By making this step explicit, you can associate a form with a
2528 declared frame window on your screen display. This can be useful
2529 if you want to adapt the form display to different screen sizes,
2530 dynamically tile forms on the screen, or use a form as part of an
2531 interface layout managed by <a href="#panels">panels</a>.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2532
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2533 <p>The two windows associated with each form have the same
2534 functions as their analogues in the <a href="#menu">menu
2535 library</a>. Both these windows are painted when the form is
2536 posted and erased when the form is unposted.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2537
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2538 <p>The outer or frame window is not otherwise touched by the form
2539 routines. It exists so the programmer can associate a title, a
2540 border, or perhaps help text with the form and have it properly
2541 refreshed or erased at post/unpost time. The inner window or
2542 subwindow is where the current form page is actually
2543 displayed.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2544
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2545 <p>In order to declare your own frame window for a form, you will
2546 need to know the size of the form's bounding rectangle. You can
2547 get this information with:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2548
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2549 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2550int scale_form(FORM *form, /* form to query */
2551 int *rows, /* form rows */
2552 int *cols); /* form cols */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2553</pre>
2554 <p>The form dimensions are passed back in the locations pointed
2555 to by the arguments. Once you have this information, you can use
2556 it to declare of windows, then use one of these functions:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2557
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2558 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2559int set_form_win(FORM *form, /* form to alter */
2560 WINDOW *win); /* frame window to connect */
2561
2562WINDOW *form_win(FORM *form); /* fetch frame window of form */
2563
2564int set_form_sub(FORM *form, /* form to alter */
2565 WINDOW *win); /* form subwindow to connect */
2566
2567WINDOW *form_sub(FORM *form); /* fetch form subwindow of form */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2568</pre>
2569 <p>Note that curses operations, including <code>refresh()</code>,
2570 on the form, should be done on the frame window, not the form
2571 subwindow.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2572
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2573 <p>It is possible to check from your application whether all of a
2574 scrollable field is actually displayed within the menu subwindow.
2575 Use these functions:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2576
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2577 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2578int data_ahead(FORM *form); /* form to be queried */
2579
2580int data_behind(FORM *form); /* form to be queried */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2581</pre>
2582 <p>The function <code>data_ahead()</code> returns TRUE if (a) the
2583 current field is one-line and has undisplayed data off to the
2584 right, (b) the current field is multi-line and there is data
2585 off-screen below it.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2586
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2587 <p>The function <code>data_behind()</code> returns TRUE if the
2588 first (upper left hand) character position is off-screen (not
2589 being displayed).</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2590
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2591 <p>Finally, there is a function to restore the form window's
2592 cursor to the value expected by the forms driver:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2593
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2594 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2595int pos_form_cursor(FORM *) /* form to be queried */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2596</pre>
2597 <p>If your application changes the form window cursor, call this
2598 function before handing control back to the forms driver in order
2599 to re-synchronize it.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2600
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2601 <h3><a name="fdriver" id="fdriver">Input Processing in the Forms
2602 Driver</a></h3>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2603
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2604 <p>The function <code>form_driver()</code> handles virtualized
2605 input requests for form navigation, editing, and validation
2606 requests, just as <code>menu_driver</code> does for menus (see
2607 the section on <a href="#minput">menu input handling</a>).</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2608
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2609 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2610int form_driver(FORM *form, /* form to pass input to */
2611 int request); /* form request code */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2612</pre>
2613 <p>Your input virtualization function needs to take input and
2614 then convert it to either an alphanumeric character (which is
2615 treated as data to be entered in the currently-selected field),
2616 or a forms processing request.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2617
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2618 <p>The forms driver provides hooks (through input-validation and
2619 field-termination functions) with which your application code can
2620 check that the input taken by the driver matched what was
2621 expected.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2622
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2623 <h4><a name="fpage" id="fpage">Page Navigation Requests</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2624
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2625 <p>These requests cause page-level moves through the form,
2626 triggering display of a new form screen.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2627
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2628 <dl>
2629 <dt><code>REQ_NEXT_PAGE</code>
2630 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2631
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2632 <dd>Move to the next form page.</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2633
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2634 <dt><code>REQ_PREV_PAGE</code>
2635 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2636
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2637 <dd>Move to the previous form page.</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2638
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2639 <dt><code>REQ_FIRST_PAGE</code>
2640 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2641
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2642 <dd>Move to the first form page.</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2643
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2644 <dt><code>REQ_LAST_PAGE</code>
2645 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2646
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2647 <dd>Move to the last form page.</dd>
2648 </dl>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2649
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2650 <p>These requests treat the list as cyclic; that is,
2651 <code>REQ_NEXT_PAGE</code> from the last page goes to the first,
2652 and <code>REQ_PREV_PAGE</code> from the first page goes to the
2653 last.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2654
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2655 <h4><a name="ffield" id="ffield">Inter-Field Navigation
2656 Requests</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2657
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2658 <p>These requests handle navigation between fields on the same
2659 page.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2660
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2661 <dl>
2662 <dt><code>REQ_NEXT_FIELD</code>
2663 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2664
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2665 <dd>Move to next field.</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2666
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2667 <dt><code>REQ_PREV_FIELD</code>
2668 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2669
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2670 <dd>Move to previous field.</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2671
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2672 <dt><code>REQ_FIRST_FIELD</code>
2673 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2674
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2675 <dd>Move to the first field.</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2676
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2677 <dt><code>REQ_LAST_FIELD</code>
2678 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2679
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2680 <dd>Move to the last field.</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2681
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2682 <dt><code>REQ_SNEXT_FIELD</code>
2683 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2684
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2685 <dd>Move to sorted next field.</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2686
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2687 <dt><code>REQ_SPREV_FIELD</code>
2688 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2689
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2690 <dd>Move to sorted previous field.</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2691
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2692 <dt><code>REQ_SFIRST_FIELD</code>
2693 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2694
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2695 <dd>Move to the sorted first field.</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2696
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2697 <dt><code>REQ_SLAST_FIELD</code>
2698 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2699
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2700 <dd>Move to the sorted last field.</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2701
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2702 <dt><code>REQ_LEFT_FIELD</code>
2703 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2704
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2705 <dd>Move left to field.</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2706
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2707 <dt><code>REQ_RIGHT_FIELD</code>
2708 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2709
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2710 <dd>Move right to field.</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2711
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2712 <dt><code>REQ_UP_FIELD</code>
2713 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2714
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2715 <dd>Move up to field.</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2716
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2717 <dt><code>REQ_DOWN_FIELD</code>
2718 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2719
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2720 <dd>Move down to field.</dd>
2721 </dl>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2722
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2723 <p>These requests treat the list of fields on a page as cyclic;
2724 that is, <code>REQ_NEXT_FIELD</code> from the last field goes to
2725 the first, and <code>REQ_PREV_FIELD</code> from the first field
2726 goes to the last. The order of the fields for these (and the
2727 <code>REQ_FIRST_FIELD</code> and <code>REQ_LAST_FIELD</code>
2728 requests) is simply the order of the field pointers in the form
2729 array (as set up by <code>new_form()</code> or
2730 <code>set_form_fields()</code></p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2731
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2732 <p>It is also possible to traverse the fields as if they had been
2733 sorted in screen-position order, so the sequence goes
2734 left-to-right and top-to-bottom. To do this, use the second group
2735 of four sorted-movement requests.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2736
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2737 <p>Finally, it is possible to move between fields using visual
2738 directions up, down, right, and left. To accomplish this, use the
2739 third group of four requests. Note, however, that the position of
2740 a form for purposes of these requests is its upper-left
2741 corner.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2742
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2743 <p>For example, suppose you have a multi-line field B, and two
2744 single-line fields A and C on the same line with B, with A to the
2745 left of B and C to the right of B. A <code>REQ_MOVE_RIGHT</code>
2746 from A will go to B only if A, B, and C <em>all</em> share the
2747 same first line; otherwise it will skip over B to C.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]2748
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]2749 <h4><a name="fifield" id="fifield">Intra-Field Navigation
2750 Requests</a></h4>
2751
2752 <p>These requests drive movement of the edit cursor within the
2753 currently selected field.</p>
2754
2755 <dl>
2756 <dt><code>REQ_NEXT_CHAR</code>
2757 </dt>
2758
2759 <dd>Move to next character.</dd>
2760
2761 <dt><code>REQ_PREV_CHAR</code>
2762 </dt>
2763
2764 <dd>Move to previous character.</dd>
2765
2766 <dt><code>REQ_NEXT_LINE</code>
2767 </dt>
2768
2769 <dd>Move to next line.</dd>
2770
2771 <dt><code>REQ_PREV_LINE</code>
2772 </dt>
2773
2774 <dd>Move to previous line.</dd>
2775
2776 <dt><code>REQ_NEXT_WORD</code>
2777 </dt>
2778
2779 <dd>Move to next word.</dd>
2780
2781 <dt><code>REQ_PREV_WORD</code>
2782 </dt>
2783
2784 <dd>Move to previous word.</dd>
2785
2786 <dt><code>REQ_BEG_FIELD</code>
2787 </dt>
2788
2789 <dd>Move to beginning of field.</dd>
2790
2791 <dt><code>REQ_END_FIELD</code>
2792 </dt>
2793
2794 <dd>Move to end of field.</dd>
2795
2796 <dt><code>REQ_BEG_LINE</code>
2797 </dt>
2798
2799 <dd>Move to beginning of line.</dd>
2800
2801 <dt><code>REQ_END_LINE</code>
2802 </dt>
2803
2804 <dd>Move to end of line.</dd>
2805
2806 <dt><code>REQ_LEFT_CHAR</code>
2807 </dt>
2808
2809 <dd>Move left in field.</dd>
2810
2811 <dt><code>REQ_RIGHT_CHAR</code>
2812 </dt>
2813
2814 <dd>Move right in field.</dd>
2815
2816 <dt><code>REQ_UP_CHAR</code>
2817 </dt>
2818
2819 <dd>Move up in field.</dd>
2820
2821 <dt><code>REQ_DOWN_CHAR</code>
2822 </dt>
2823
2824 <dd>Move down in field.</dd>
2825 </dl>
2826
2827 <p>Each <em>word</em> is separated from the previous and next
2828 characters by whitespace. The commands to move to beginning and
2829 end of line or field look for the first or last non-pad character
2830 in their ranges.</p>
2831
2832 <h4><a name="fscroll" id="fscroll">Scrolling Requests</a></h4>
2833
2834 <p>Fields that are dynamic and have grown and fields explicitly
2835 created with offscreen rows are scrollable. One-line fields
2836 scroll horizontally; multi-line fields scroll vertically. Most
2837 scrolling is triggered by editing and intra-field movement (the
2838 library scrolls the field to keep the cursor visible). It is
2839 possible to explicitly request scrolling with the following
2840 requests:</p>
2841
2842 <dl>
2843 <dt><code>REQ_SCR_FLINE</code>
2844 </dt>
2845
2846 <dd>Scroll vertically forward a line.</dd>
2847
2848 <dt><code>REQ_SCR_BLINE</code>
2849 </dt>
2850
2851 <dd>Scroll vertically backward a line.</dd>
2852
2853 <dt><code>REQ_SCR_FPAGE</code>
2854 </dt>
2855
2856 <dd>Scroll vertically forward a page.</dd>
2857
2858 <dt><code>REQ_SCR_BPAGE</code>
2859 </dt>
2860
2861 <dd>Scroll vertically backward a page.</dd>
2862
2863 <dt><code>REQ_SCR_FHPAGE</code>
2864 </dt>
2865
2866 <dd>Scroll vertically forward half a page.</dd>
2867
2868 <dt><code>REQ_SCR_BHPAGE</code>
2869 </dt>
2870
2871 <dd>Scroll vertically backward half a page.</dd>
2872
2873 <dt><code>REQ_SCR_FCHAR</code>
2874 </dt>
2875
2876 <dd>Scroll horizontally forward a character.</dd>
2877
2878 <dt><code>REQ_SCR_BCHAR</code>
2879 </dt>
2880
2881 <dd>Scroll horizontally backward a character.</dd>
2882
2883 <dt><code>REQ_SCR_HFLINE</code>
2884 </dt>
2885
2886 <dd>Scroll horizontally one field width forward.</dd>
2887
2888 <dt><code>REQ_SCR_HBLINE</code>
2889 </dt>
2890
2891 <dd>Scroll horizontally one field width backward.</dd>
2892
2893 <dt><code>REQ_SCR_HFHALF</code>
2894 </dt>
2895
2896 <dd>Scroll horizontally one half field width forward.</dd>
2897
2898 <dt><code>REQ_SCR_HBHALF</code>
2899 </dt>
2900
2901 <dd>Scroll horizontally one half field width backward.</dd>
2902 </dl>
2903
2904 <p>For scrolling purposes, a <em>page</em> of a field is the
2905 height of its visible part.</p>
2906
2907 <h4><a name="fedit" id="fedit">Editing Requests</a></h4>
2908
2909 <p>When you pass the forms driver an ASCII character, it is
2910 treated as a request to add the character to the field's data
2911 buffer. Whether this is an insertion or a replacement depends on
2912 the field's edit mode (insertion is the default.</p>
2913
2914 <p>The following requests support editing the field and changing
2915 the edit mode:</p>
2916
2917 <dl>
2918 <dt><code>REQ_INS_MODE</code>
2919 </dt>
2920
2921 <dd>Set insertion mode.</dd>
2922
2923 <dt><code>REQ_OVL_MODE</code>
2924 </dt>
2925
2926 <dd>Set overlay mode.</dd>
2927
2928 <dt><code>REQ_NEW_LINE</code>
2929 </dt>
2930
2931 <dd>New line request (see below for explanation).</dd>
2932
2933 <dt><code>REQ_INS_CHAR</code>
2934 </dt>
2935
2936 <dd>Insert space at character location.</dd>
2937
2938 <dt><code>REQ_INS_LINE</code>
2939 </dt>
2940
2941 <dd>Insert blank line at character location.</dd>
2942
2943 <dt><code>REQ_DEL_CHAR</code>
2944 </dt>
2945
2946 <dd>Delete character at cursor.</dd>
2947
2948 <dt><code>REQ_DEL_PREV</code>
2949 </dt>
2950
2951 <dd>Delete previous word at cursor.</dd>
2952
2953 <dt><code>REQ_DEL_LINE</code>
2954 </dt>
2955
2956 <dd>Delete line at cursor.</dd>
2957
2958 <dt><code>REQ_DEL_WORD</code>
2959 </dt>
2960
2961 <dd>Delete word at cursor.</dd>
2962
2963 <dt><code>REQ_CLR_EOL</code>
2964 </dt>
2965
2966 <dd>Clear to end of line.</dd>
2967
2968 <dt><code>REQ_CLR_EOF</code>
2969 </dt>
2970
2971 <dd>Clear to end of field.</dd>
2972
2973 <dt><code>REQ_CLEAR_FIELD</code>
2974 </dt>
2975
2976 <dd>Clear entire field.</dd>
2977 </dl>
2978
2979 <p>The behavior of the <code>REQ_NEW_LINE</code> and
2980 <code>REQ_DEL_PREV</code> requests is complicated and partly
2981 controlled by a pair of forms options. The special cases are
2982 triggered when the cursor is at the beginning of a field, or on
2983 the last line of the field.</p>
2984
2985 <p>First, we consider <code>REQ_NEW_LINE</code>:</p>
2986
2987 <p>The normal behavior of <code>REQ_NEW_LINE</code> in insert
2988 mode is to break the current line at the position of the edit
2989 cursor, inserting the portion of the current line after the
2990 cursor as a new line following the current and moving the cursor
2991 to the beginning of that new line (you may think of this as
2992 inserting a newline in the field buffer).</p>
2993
2994 <p>The normal behavior of <code>REQ_NEW_LINE</code> in overlay
2995 mode is to clear the current line from the position of the edit
2996 cursor to end of line. The cursor is then moved to the beginning
2997 of the next line.</p>
2998
2999 <p>However, <code>REQ_NEW_LINE</code> at the beginning of a
3000 field, or on the last line of a field, instead does a
3001 <code>REQ_NEXT_FIELD</code>. <code>O_NL_OVERLOAD</code> option is
3002 off, this special action is disabled.</p>
3003
3004 <p>Now, let us consider <code>REQ_DEL_PREV</code>:</p>
3005
3006 <p>The normal behavior of <code>REQ_DEL_PREV</code> is to delete
3007 the previous character. If insert mode is on, and the cursor is
3008 at the start of a line, and the text on that line will fit on the
3009 previous one, it instead appends the contents of the current line
3010 to the previous one and deletes the current line (you may think
3011 of this as deleting a newline from the field buffer).</p>
3012
3013 <p>However, <code>REQ_DEL_PREV</code> at the beginning of a field
3014 is instead treated as a <code>REQ_PREV_FIELD</code>.</p>
3015
3016 <p>If the <code>O_BS_OVERLOAD</code> option is off, this special
3017 action is disabled and the forms driver just returns
3018 <code>E_REQUEST_DENIED</code>.</p>
3019
3020 <p>See <a href="#frmoptions">Form Options</a> for discussion of
3021 how to set and clear the overload options.</p>
3022
3023 <h4><a name="forder" id="forder">Order Requests</a></h4>
3024
3025 <p>If the type of your field is ordered, and has associated
3026 functions for getting the next and previous values of the type
3027 from a given value, there are requests that can fetch that value
3028 into the field buffer:</p>
3029
3030 <dl>
3031 <dt><code>REQ_NEXT_CHOICE</code>
3032 </dt>
3033
3034 <dd>Place the successor value of the current value in the
3035 buffer.</dd>
3036
3037 <dt><code>REQ_PREV_CHOICE</code>
3038 </dt>
3039
3040 <dd>Place the predecessor value of the current value in the
3041 buffer.</dd>
3042 </dl>
3043
3044 <p>Of the built-in field types, only <code>TYPE_ENUM</code> has
3045 built-in successor and predecessor functions. When you define a
3046 field type of your own (see <a href="#fcustom">Custom Validation
3047 Types</a>), you can associate our own ordering functions.</p>
3048
3049 <h4><a name="fappcmds" id="fappcmds">Application Commands</a></h4>
3050
3051 <p>Form requests are represented as integers above the
3052 <code>curses</code> value greater than <code>KEY_MAX</code> and
3053 less than or equal to the constant <code>MAX_COMMAND</code>. If
3054 your input-virtualization routine returns a value above
3055 <code>MAX_COMMAND</code>, the forms driver will ignore it.</p>
3056
3057 <h3><a name="fhooks" id="fhooks">Field Change Hooks</a></h3>
3058
3059 <p>It is possible to set function hooks to be executed whenever
3060 the current field or form changes. Here are the functions that
3061 support this:</p>
3062
3063 <pre class="code-block">
3064typedef void (*HOOK)(); /* pointer to function returning void */
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3065
3066int set_form_init(FORM *form, /* form to alter */
3067 HOOK hook); /* initialization hook */
3068
3069HOOK form_init(FORM *form); /* form to query */
3070
3071int set_form_term(FORM *form, /* form to alter */
3072 HOOK hook); /* termination hook */
3073
3074HOOK form_term(FORM *form); /* form to query */
3075
3076int set_field_init(FORM *form, /* form to alter */
3077 HOOK hook); /* initialization hook */
3078
3079HOOK field_init(FORM *form); /* form to query */
3080
3081int set_field_term(FORM *form, /* form to alter */
3082 HOOK hook); /* termination hook */
3083
3084HOOK field_term(FORM *form); /* form to query */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3085</pre>
3086 <p>These functions allow you to either set or query four
3087 different hooks. In each of the set functions, the second
3088 argument should be the address of a hook function. These
3089 functions differ only in the timing of the hook call.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3090
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3091 <dl>
3092 <dt>form_init</dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3093
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3094 <dd>This hook is called when the form is posted; also, just
3095 after each page change operation.</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3096
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3097 <dt>field_init</dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3098
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3099 <dd>This hook is called when the form is posted; also, just
3100 after each field change</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3101
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3102 <dt>field_term</dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3103
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3104 <dd>This hook is called just after field validation; that is,
3105 just before the field is altered. It is also called when the
3106 form is unposted.</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3107
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3108 <dt>form_term</dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3109
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3110 <dd>This hook is called when the form is unposted; also, just
3111 before each page change operation.</dd>
3112 </dl>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3113
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3114 <p>Calls to these hooks may be triggered</p>
3115
3116 <ol>
3117 <li>When user editing requests are processed by the forms
3118 driver</li>
3119
3120 <li>When the current page is changed by
3121 <code>set_current_field()</code> call</li>
3122
3123 <li>When the current field is changed by a
3124 <code>set_form_page()</code> call</li>
3125 </ol>
3126
3127 <p>See <a name="ffocus" id="ffocus">Field Change Commands</a> for
3128 discussion of the latter two cases.</p>
3129
3130 <p>You can set a default hook for all fields by passing one of
3131 the set functions a NULL first argument.</p>
3132
3133 <p>You can disable any of these hooks by (re)setting them to
3134 NULL, the default value.</p>
3135
3136 <h3><a href="#ffocus">Field Change Commands</a></h3>
3137
3138 <p>Normally, navigation through the form will be driven by the
3139 user's input requests. But sometimes it is useful to be able to
3140 move the focus for editing and viewing under control of your
3141 application, or ask which field it currently is in. The following
3142 functions help you accomplish this:</p>
3143
3144 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3145int set_current_field(FORM *form, /* form to alter */
3146 FIELD *field); /* field to shift to */
3147
3148FIELD *current_field(FORM *form); /* form to query */
3149
3150int field_index(FORM *form, /* form to query */
3151 FIELD *field); /* field to get index of */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3152</pre>
3153 <p>The function <code>field_index()</code> returns the index of
3154 the given field in the given form's field array (the array passed
3155 to <code>new_form()</code> or
3156 <code>set_form_fields()</code>).</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3157
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3158 <p>The initial current field of a form is the first active field
3159 on the first page. The function <code>set_form_fields()</code>
3160 resets this.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3161
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3162 <p>It is also possible to move around by pages.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3163
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3164 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3165int set_form_page(FORM *form, /* form to alter */
3166 int page); /* page to go to (0-origin) */
3167
3168int form_page(FORM *form); /* return form's current page */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3169</pre>
3170 <p>The initial page of a newly-created form is 0. The function
3171 <code>set_form_fields()</code> resets this.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3172
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3173 <h3><a name="frmoptions" id="frmoptions">Form Options</a></h3>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3174
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3175 <p>Like fields, forms may have control option bits. They can be
3176 changed or queried with these functions:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3177
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3178 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3179int set_form_opts(FORM *form, /* form to alter */
3180 int attr); /* attribute to set */
3181
3182int form_opts_on(FORM *form, /* form to alter */
3183 int attr); /* attributes to turn on */
3184
3185int form_opts_off(FORM *form, /* form to alter */
3186 int attr); /* attributes to turn off */
3187
3188int form_opts(FORM *form); /* form to query */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3189</pre>
3190 <p>By default, all options are on. Here are the available option
3191 bits:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3192
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3193 <dl>
3194 <dt>O_NL_OVERLOAD</dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3195
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3196 <dd>Enable overloading of <code>REQ_NEW_LINE</code> as
3197 described in <a href="#fedit">Editing Requests</a>. The value
3198 of this option is ignored on dynamic fields that have not
3199 reached their size limit; these have no last line, so the
3200 circumstances for triggering a <code>REQ_NEXT_FIELD</code>
3201 never arise.</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3202
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3203 <dt>O_BS_OVERLOAD</dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3204
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3205 <dd>Enable overloading of <code>REQ_DEL_PREV</code> as
3206 described in <a href="#fedit">Editing Requests</a>.</dd>
3207 </dl>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3208
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3209 <p>The option values are bit-masks and can be composed with
3210 logical-or in the obvious way.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3211
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3212 <h3><a name="fcustom" id="fcustom">Custom Validation Types</a></h3>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3213
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3214 <p>The <code>form</code> library gives you the capability to
3215 define custom validation types of your own. Further, the optional
3216 additional arguments of <code>set_field_type</code> effectively
3217 allow you to parameterize validation types. Most of the
3218 complications in the validation-type interface have to do with
3219 the handling of the additional arguments within custom validation
3220 functions.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3221
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3222 <h4><a name="flinktypes" id="flinktypes">Union Types</a></h4>
3223
3224 <p>The simplest way to create a custom data type is to compose it
3225 from two preexisting ones:</p>
3226
3227 <pre class="code-block">
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3228FIELD *link_fieldtype(FIELDTYPE *type1,
3229 FIELDTYPE *type2);
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3230</pre>
3231 <p>This function creates a field type that will accept any of the
3232 values legal for either of its argument field types (which may be
3233 either predefined or programmer-defined). If a
3234 <code>set_field_type()</code> call later requires arguments, the
3235 new composite type expects all arguments for the first type, than
3236 all arguments for the second. Order functions (see <a href=
3237 "#forder">Order Requests</a>) associated with the component types
3238 will work on the composite; what it does is check the validation
3239 function for the first type, then for the second, to figure what
3240 type the buffer contents should be treated as.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3241
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3242 <h4><a name="fnewtypes" id="fnewtypes">New Field Types</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3243
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3244 <p>To create a field type from scratch, you need to specify one
3245 or both of the following things:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3246
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3247 <ul>
3248 <li>A character-validation function, to check each character as
3249 it is entered.</li>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3250
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3251 <li>A field-validation function to be applied on exit from the
3252 field.</li>
3253 </ul>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3254
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3255 <p>Here is how you do that:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3256
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3257 <pre class="code-block">
3258typedef int (*HOOK)(); /* pointer to function returning int */
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3259
3260FIELDTYPE *new_fieldtype(HOOK f_validate, /* field validator */
3261 HOOK c_validate) /* character validator */
3262
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3263int free_fieldtype(FIELDTYPE *ftype); /* type to free */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3264</pre>
3265 <p>At least one of the arguments of <code>new_fieldtype()</code>
3266 must be non-NULL. The forms driver will automatically call the
3267 new type's validation functions at appropriate points in
3268 processing a field of the new type.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3269
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3270 <p>The function <code>free_fieldtype()</code> deallocates the
3271 argument fieldtype, freeing all storage associated with it.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3272
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3273 <p>Normally, a field validator is called when the user attempts
3274 to leave the field. Its first argument is a field pointer, from
3275 which it can get to field buffer 0 and test it. If the function
3276 returns TRUE, the operation succeeds; if it returns FALSE, the
3277 edit cursor stays in the field.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3278
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3279 <p>A character validator gets the character passed in as a first
3280 argument. It too should return TRUE if the character is valid,
3281 FALSE otherwise.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3282
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3283 <h4><a name="fcheckargs" id="fcheckargs">Validation Function
3284 Arguments</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3285
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3286 <p>Your field- and character- validation functions will be passed
3287 a second argument as well. This second argument is the address of
3288 a structure (which we will call a <em>pile</em>) built from any
3289 of the field-type-specific arguments passed to
3290 <code>set_field_type()</code>. If no such arguments are defined
3291 for the field type, this pile pointer argument will be NULL.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3292
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3293 <p>In order to arrange for such arguments to be passed to your
3294 validation functions, you must associate a small set of
3295 storage-management functions with the type. The forms driver will
3296 use these to synthesize a pile from the trailing arguments of
3297 each <code>set_field_type()</code> argument, and a pointer to the
3298 pile will be passed to the validation functions.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3299
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3300 <p>Here is how you make the association:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3301
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3302 <pre class="code-block">
3303typedef char *(*PTRHOOK)(); /* pointer to function returning (char *) */
3304typedef void (*VOIDHOOK)(); /* pointer to function returning void */
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3305
3306int set_fieldtype_arg(FIELDTYPE *type, /* type to alter */
3307 PTRHOOK make_str, /* make structure from args */
3308 PTRHOOK copy_str, /* make copy of structure */
3309 VOIDHOOK free_str); /* free structure storage */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3310</pre>
3311 <p>Here is how the storage-management hooks are used:</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3312
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3313 <dl>
3314 <dt><code>make_str</code>
3315 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3316
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3317 <dd>This function is called by <code>set_field_type()</code>.
3318 It gets one argument, a <code>va_list</code> of the
3319 type-specific arguments passed to
3320 <code>set_field_type()</code>. It is expected to return a pile
3321 pointer to a data structure that encapsulates those
3322 arguments.</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3323
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3324 <dt><code>copy_str</code>
3325 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3326
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3327 <dd>This function is called by form library functions that
3328 allocate new field instances. It is expected to take a pile
3329 pointer, copy the pile to allocated storage, and return the
3330 address of the pile copy.</dd>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3331
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3332 <dt><code>free_str</code>
3333 </dt>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3334
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3335 <dd>This function is called by field- and type-deallocation
3336 routines in the library. It takes a pile pointer argument, and
3337 is expected to free the storage of that pile.</dd>
3338 </dl>
3339
3340 <p>The <code>make_str</code> and <code>copy_str</code> functions
3341 may return NULL to signal allocation failure. The library
3342 routines will that call them will return error indication when
3343 this happens. Thus, your validation functions should never see a
3344 NULL file pointer and need not check specially for it.</p>
3345
3346 <h4><a name="fcustorder" id="fcustorder">Order Functions For
3347 Custom Types</a></h4>
3348
3349 <p>Some custom field types are simply ordered in the same
3350 well-defined way that <code>TYPE_ENUM</code> is. For such types,
3351 it is possible to define successor and predecessor functions to
3352 support the <code>REQ_NEXT_CHOICE</code> and
3353 <code>REQ_PREV_CHOICE</code> requests. Here is how:</p>
3354
3355 <pre class="code-block">
3356typedef int (*INTHOOK)(); /* pointer to function returning int */
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3357
3358int set_fieldtype_arg(FIELDTYPE *type, /* type to alter */
3359 INTHOOK succ, /* get successor value */
3360 INTHOOK pred); /* get predecessor value */
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3361</pre>
3362 <p>The successor and predecessor arguments will each be passed
3363 two arguments; a field pointer, and a pile pointer (as for the
3364 validation functions). They are expected to use the function
3365 <code>field_buffer()</code> to read the current value, and
3366 <code>set_field_buffer()</code> on buffer 0 to set the next or
3367 previous value. Either hook may return TRUE to indicate success
3368 (a legal next or previous value was set) or FALSE to indicate
3369 failure.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3370
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3371 <h4><a name="fcustprobs" id="fcustprobs">Avoiding Problems</a></h4>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3372
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3373 <p>The interface for defining custom types is complicated and
3374 tricky. Rather than attempting to create a custom type entirely
3375 from scratch, you should start by studying the library source
3376 code for whichever of the pre-defined types seems to be closest
3377 to what you want.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3378
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3379 <p>Use that code as a model, and evolve it towards what you
3380 really want. You will avoid many problems and annoyances that
3381 way. The code in the <code>ncurses</code> library has been
3382 specifically exempted from the package copyright to support
3383 this.</p>
Amit Daniel Kachhape6a01f52011-07-20 11:45:59 +0530[diff] [blame]3384
micky3879b9f5e72025-07-08 18:04:53 -0400[diff] [blame]3385 <p>If your custom type defines order functions, have do something
3386 intuitive with a blank field. A useful convention is to make the
3387 successor of a blank field the types minimum value, and its
3388 predecessor the maximum.</p>
3389</body>
3390</html>