RootZanli & NinjaCr3

: / usr / share / doc / ncurses-devel /
Filename : ncurses-intro.html
back
<!--
  $Id: ncurses-intro.html,v 1.54 2020/02/02 23:34:34 tom Exp $
  ****************************************************************************
  * Copyright 2019,2020 Thomas E. Dickey                                     *
  * Copyright 2000-2013,2017 Free Software Foundation, Inc.                  *
  *                                                                          *
  * Permission is hereby granted, free of charge, to any person obtaining a  *
  * copy of this software and associated documentation files (the            *
  * "Software"), to deal in the Software without restriction, including      *
  * without limitation the rights to use, copy, modify, merge, publish,      *
  * distribute, distribute with modifications, sublicense, and/or sell       *
  * copies of the Software, and to permit persons to whom the Software is    *
  * furnished to do so, subject to the following conditions:                 *
  *                                                                          *
  * The above copyright notice and this permission notice shall be included  *
  * in all copies or substantial portions of the Software.                   *
  *                                                                          *
  * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS  *
  * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF               *
  * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.   *
  * IN NO EVENT SHALL THE ABOVE COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM,   *
  * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR    *
  * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR    *
  * THE USE OR OTHER DEALINGS IN THE SOFTWARE.                               *
  *                                                                          *
  * Except as contained in this notice, the name(s) of the above copyright   *
  * holders shall not be used in advertising or otherwise to promote the     *
  * sale, use or other dealings in this Software without prior written       *
  * authorization.                                                           *
  ****************************************************************************
-->
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01//EN">

<html>
<head>
  <meta name="generator" content=
  "HTML Tidy for HTML5 for Linux version 5.2.0">

  <title>Writing Programs with NCURSES</title>
  <link rel="author" href="mailto:bugs-ncurses@gnu.org">
  <meta http-equiv="Content-Type" content=
  "text/html; charset=us-ascii">
</head>

<body>
  <h1>Writing Programs with NCURSES</h1>

  <blockquote>
    by Eric S. Raymond and Zeyd M. Ben-Halim<br>
    updates since release 1.9.9e by Thomas Dickey
  </blockquote>

  <h1>Contents</h1>

  <ul>
    <li>
      <a href="#introduction">Introduction</a>

      <ul>
        <li><a href="#history">A Brief History of Curses</a></li>

        <li><a href="#scope">Scope of This Document</a></li>

        <li><a href="#terminology">Terminology</a></li>
      </ul>
    </li>

    <li>
      <a href="#curses">The Curses Library</a>

      <ul>
        <li>
          <a href="#overview">An Overview of Curses</a>

          <ul>
            <li><a href="#compiling">Compiling Programs using
            Curses</a></li>

            <li><a href="#updating">Updating the Screen</a></li>

            <li><a href="#stdscr">Standard Windows and Function
            Naming Conventions</a></li>

            <li><a href="#variables">Variables</a></li>
          </ul>
        </li>

        <li>
          <a href="#using">Using the Library</a>

          <ul>
            <li><a href="#starting">Starting up</a></li>

            <li><a href="#output">Output</a></li>

            <li><a href="#input">Input</a></li>

            <li><a href="#formschars">Using Forms Characters</a></li>

            <li><a href="#attributes">Character Attributes and
            Color</a></li>

            <li><a href="#mouse">Mouse Interfacing</a></li>

            <li><a href="#finishing">Finishing Up</a></li>
          </ul>
        </li>

        <li>
          <a href="#functions">Function Descriptions</a>

          <ul>
            <li><a href="#init">Initialization and Wrapup</a></li>

            <li><a href="#flush">Causing Output to the Terminal</a></li>

            <li><a href="#lowlevel">Low-Level Capability Access</a></li>

            <li><a href="#debugging">Debugging</a></li>
          </ul>
        </li>

        <li>
          <a href="#hints">Hints, Tips, and Tricks</a>

          <ul>
            <li><a href="#caution">Some Notes of Caution</a></li>

            <li><a href="#leaving">Temporarily Leaving ncurses
            Mode</a></li>

            <li><a href="#xterm">Using <code>ncurses</code> under
            <code>xterm</code></a></li>

            <li><a href="#screens">Handling Multiple Terminal
            Screens</a></li>

            <li><a href="#testing">Testing for Terminal
            Capabilities</a></li>

            <li><a href="#tuning">Tuning for Speed</a></li>

            <li><a href="#special">Special Features of
            <code>ncurses</code></a></li>
          </ul>
        </li>

        <li>
          <a href="#compat">Compatibility with Older Versions</a>

          <ul>
            <li><a href="#refbug">Refresh of Overlapping
            Windows</a></li>

            <li><a href="#backbug">Background Erase</a></li>
          </ul>
        </li>

        <li><a href="#xsifuncs">XSI Curses Conformance</a></li>
      </ul>
    </li>

    <li>
      <a href="#panels">The Panels Library</a>

      <ul>
        <li><a href="#pcompile">Compiling With the Panels
        Library</a></li>

        <li><a href="#poverview">Overview of Panels</a></li>

        <li><a href="#pstdscr">Panels, Input, and the Standard
        Screen</a></li>

        <li><a href="#hiding">Hiding Panels</a></li>

        <li><a href="#pmisc">Miscellaneous Other Facilities</a></li>
      </ul>
    </li>

    <li>
      <a href="#menu">The Menu Library</a>

      <ul>
        <li><a href="#mcompile">Compiling with the menu Library</a></li>

        <li><a href="#moverview">Overview of Menus</a></li>

        <li><a href="#mselect">Selecting items</a></li>

        <li><a href="#mdisplay">Menu Display</a></li>

        <li><a href="#mwindows">Menu Windows</a></li>

        <li><a href="#minput">Processing Menu Input</a></li>

        <li><a href="#mmisc">Miscellaneous Other Features</a></li>
      </ul>
    </li>

    <li>
      <a href="#form">The Forms Library</a>

      <ul>
        <li><a href="#fcompile">Compiling with the forms
        Library</a></li>

        <li><a href="#foverview">Overview of Forms</a></li>

        <li><a href="#fcreate">Creating and Freeing Fields and
        Forms</a></li>

        <li>
          <a href="#fattributes">Fetching and Changing Field
          Attributes</a>

          <ul>
            <li><a href="#fsizes">Fetching Size and Location
            Data</a></li>

            <li><a href="#flocation">Changing the Field
            Location</a></li>

            <li><a href="#fjust">The Justification Attribute</a></li>

            <li><a href="#fdispatts">Field Display Attributes</a></li>

            <li><a href="#foptions">Field Option Bits</a></li>

            <li><a href="#fstatus">Field Status</a></li>

            <li><a href="#fuser">Field User Pointer</a></li>
          </ul>
        </li>

        <li><a href="#fdynamic">Variable-Sized Fields</a></li>

        <li>
          <a href="#fvalidation">Field Validation</a>

          <ul>
            <li><a href="#ftype_alpha">TYPE_ALPHA</a></li>

            <li><a href="#ftype_alnum">TYPE_ALNUM</a></li>

            <li><a href="#ftype_enum">TYPE_ENUM</a></li>

            <li><a href="#ftype_integer">TYPE_INTEGER</a></li>

            <li><a href="#ftype_numeric">TYPE_NUMERIC</a></li>

            <li><a href="#ftype_regexp">TYPE_REGEXP</a></li>
          </ul>
        </li>

        <li><a href="#fbuffer">Direct Field Buffer Manipulation</a></li>

        <li><a href="#formattrs">Attributes of Forms</a></li>

        <li><a href="#fdisplay">Control of Form Display</a></li>

        <li>
          <a href="#fdriver">Input Processing in the Forms
          Driver</a>

          <ul>
            <li><a href="#fpage">Page Navigation Requests</a></li>

            <li><a href="#ffield">Inter-Field Navigation
            Requests</a></li>

            <li><a href="#fifield">Intra-Field Navigation
            Requests</a></li>

            <li><a href="#fscroll">Scrolling Requests</a></li>

            <li><a href="#fedit">Field Editing Requests</a></li>

            <li><a href="#forder">Order Requests</a></li>

            <li><a href="#fappcmds">Application Commands</a></li>
          </ul>
        </li>

        <li><a href="#fhooks">Field Change Hooks</a></li>

        <li><a href="#ffocus">Field Change Commands</a></li>

        <li><a href="#frmoptions">Form Options</a></li>

        <li>
          <a href="#fcustom">Custom Validation Types</a>

          <ul>
            <li><a href="#flinktypes">Union Types</a></li>

            <li><a href="#fnewtypes">New Field Types</a></li>

            <li><a href="#fcheckargs">Validation Function
            Arguments</a></li>

            <li><a href="#fcustorder">Order Functions For Custom
            Types</a></li>

            <li><a href="#fcustprobs">Avoiding Problems</a></li>
          </ul>
        </li>
      </ul>
    </li>
  </ul>

  <hr>

  <h1><a name="introduction" id="introduction">Introduction</a></h1>

  <p>This document is an introduction to programming with
  <code>curses</code>. It is not an exhaustive reference for the
  curses Application Programming Interface (API); that role is
  filled by the <code>curses</code> manual pages. Rather, it is
  intended to help C programmers ease into using the package.</p>

  <p>This document is aimed at C applications programmers not yet
  specifically familiar with ncurses. If you are already an
  experienced <code>curses</code> programmer, you should
  nevertheless read the sections on <a href="#mouse">Mouse
  Interfacing</a>, <a href="#debugging">Debugging</a>, <a href=
  "#compat">Compatibility with Older Versions</a>, and <a href=
  "#hints">Hints, Tips, and Tricks</a>. These will bring you up to
  speed on the special features and quirks of the
  <code>ncurses</code> implementation. If you are not so
  experienced, keep reading.</p>

  <p>The <code>curses</code> package is a subroutine library for
  terminal-independent screen-painting and input-event handling
  which presents a high level screen model to the programmer,
  hiding differences between terminal types and doing automatic
  optimization of output to change one screen full of text into
  another. <code>Curses</code> uses terminfo, which is a database
  format that can describe the capabilities of thousands of
  different terminals.</p>

  <p>The <code>curses</code> API may seem something of an archaism
  on UNIX desktops increasingly dominated by X, Motif, and Tcl/Tk.
  Nevertheless, UNIX still supports tty lines and X supports
  <em>xterm(1)</em>; the <code>curses</code> API has the advantage
  of (a) back-portability to character-cell terminals, and (b)
  simplicity. For an application that does not require bit-mapped
  graphics and multiple fonts, an interface implementation using
  <code>curses</code> will typically be a great deal simpler and
  less expensive than one using an X toolkit.</p>

  <h2><a name="history" id="history">A Brief History of Curses</a></h2>

  <p>Historically, the first ancestor of <code>curses</code> was
  the routines written to provide screen-handling for the
  <code>vi</code> editor; these used the <code>termcap</code>
  database facility (both released in 3BSD) for describing terminal
  capabilities. These routines were abstracted into a documented
  library and first released with the early BSD UNIX versions. All
  of this work was done by students at the University of California
  (Berkeley campus). The curses library was first published in
  4.0BSD, a year after 3BSD (i.e., late 1980).</p>

  <p>After graduation, one of those students went to work at
  AT&amp;T Bell Labs, and made an improved <code>termcap</code>
  library called <code>terminfo</code> (i.e.,
  &ldquo;libterm&rdquo;), and adapted the curses library to use
  this. That was subsequently released in System V Release 2 (early
  1984). Thereafter, other developers added to the curses and
  terminfo libraries. For instance, a student at Cornell University
  wrote an improved terminfo library as well as a tool
  (<code>tic</code>) to compile the terminal descriptions. As a
  general rule, AT&amp;T did not identify the developers in the
  source-code or documentation; the <code>tic</code> and
  <code>infocmp</code> programs are the exceptions.</p>

  <p>System V Release 3 (System III UNIX) from Bell Labs featured a
  rewritten and much-improved <code>curses</code> library, along
  with the <code>tic</code> program (late 1986).</p>

  <p>To recap, terminfo is based on Berkeley's termcap database,
  but contains a number of improvements and extensions.
  Parameterized capabilities strings were introduced, making it
  possible to describe multiple video attributes, and colors and to
  handle far more unusual terminals than possible with termcap. In
  the later AT&amp;T System V releases, <code>curses</code> evolved
  to use more facilities and offer more capabilities, going far
  beyond BSD curses in power and flexibility.</p>

  <h2><a name="scope" id="scope">Scope of This Document</a></h2>

  <p>This document describes <code>ncurses</code>, a free
  implementation of the System V <code>curses</code> API with some
  clearly marked extensions. It includes the following System V
  curses features:</p>

  <ul>
    <li>Support for multiple screen highlights (BSD curses could
    only handle one &ldquo;standout&rdquo; highlight, usually
    reverse-video).</li>

    <li>Support for line- and box-drawing using forms
    characters.</li>

    <li>Recognition of function keys on input.</li>

    <li>Color support.</li>

    <li>Support for pads (windows of larger than screen size on
    which the screen or a subwindow defines a viewport).</li>
  </ul>

  <p>Also, this package makes use of the insert and delete line and
  character features of terminals so equipped, and determines how
  to optimally use these features with no help from the programmer.
  It allows arbitrary combinations of video attributes to be
  displayed, even on terminals that leave &ldquo;magic
  cookies&rdquo; on the screen to mark changes in attributes.</p>

  <p>The <code>ncurses</code> package can also capture and use
  event reports from a mouse in some environments (notably, xterm
  under the X window system). This document includes tips for using
  the mouse.</p>

  <p>The <code>ncurses</code> package was originated by Pavel
  Curtis. The original maintainer of this package is <a href=
  "mailto:zmbenhal@netcom.com">Zeyd Ben-Halim</a>
  &lt;zmbenhal@netcom.com&gt;. <a href=
  "mailto:esr@snark.thyrsus.com">Eric S. Raymond</a>
  &lt;esr@snark.thyrsus.com&gt; wrote many of the new features in
  versions after 1.8.1 and wrote most of this introduction.
  J&uuml;rgen Pfeifer wrote all of the menu and forms code as well
  as the <a href="http://www.adahome.com">Ada95</a> binding.
  Ongoing work is being done by <a href=
  "mailto:dickey@invisible-island.net">Thomas Dickey</a>
  (maintainer). Contact the current maintainers at <a href=
  "mailto:bug-ncurses@gnu.org">bug-ncurses@gnu.org</a>.</p>

  <p>This document also describes the <a href="#panels">panels</a>
  extension library, similarly modeled on the SVr4 panels facility.
  This library allows you to associate backing store with each of a
  stack or deck of overlapping windows, and provides operations for
  moving windows around in the stack that change their visibility
  in the natural way (handling window overlaps).</p>

  <p>Finally, this document describes in detail the <a href=
  "#menu">menus</a> and <a href="#form">forms</a> extension
  libraries, also cloned from System V, which support easy
  construction and sequences of menus and fill-in forms.</p>

  <h2><a name="terminology" id="terminology">Terminology</a></h2>

  <p>In this document, the following terminology is used with
  reasonable consistency:</p>

  <dl>
    <dt>window</dt>

    <dd>A data structure describing a sub-rectangle of the screen
    (possibly the entire screen). You can write to a window as
    though it were a miniature screen, scrolling independently of
    other windows on the physical screen.</dd>

    <dt>screens</dt>

    <dd>A subset of windows which are as large as the terminal
    screen, i.e., they start at the upper left hand corner and
    encompass the lower right hand corner. One of these,
    <code>stdscr</code>, is automatically provided for the
    programmer.</dd>

    <dt>terminal screen</dt>

    <dd>The package's idea of what the terminal display currently
    looks like, i.e., what the user sees now. This is a special
    screen.</dd>
  </dl>

  <h1><a name="curses" id="curses">The Curses Library</a></h1>

  <h2><a name="overview" id="overview">An Overview of Curses</a></h2>

  <h3><a name="compiling" id="compiling">Compiling Programs using
  Curses</a></h3>

  <p>In order to use the library, it is necessary to have certain
  types and variables defined. Therefore, the programmer must have
  a line:</p>

  <pre>
          #include &lt;curses.h&gt;
</pre>

  <p>at the top of the program source. The screen package uses the
  Standard I/O library, so <code>&lt;curses.h&gt;</code> includes
  <code>&lt;stdio.h&gt;</code>. <code>&lt;curses.h&gt;</code> also
  includes <code>&lt;termios.h&gt;</code>,
  <code>&lt;termio.h&gt;</code>, or <code>&lt;sgtty.h&gt;</code>
  depending on your system. It is redundant (but harmless) for the
  programmer to do these includes, too. In linking with
  <code>curses</code> you need to have <code>-lncurses</code> in
  your LDFLAGS or on the command line. There is no need for any
  other libraries.</p>

  <h3><a name="updating" id="updating">Updating the Screen</a></h3>

  <p>In order to update the screen optimally, it is necessary for
  the routines to know what the screen currently looks like and
  what the programmer wants it to look like next. For this purpose,
  a data type (structure) named WINDOW is defined which describes a
  window image to the routines, including its starting position on
  the screen (the (y, x) coordinates of the upper left hand corner)
  and its size. One of these (called <code>curscr</code>, for
  current screen) is a screen image of what the terminal currently
  looks like. Another screen (called <code>stdscr</code>, for
  standard screen) is provided by default to make changes on.</p>

  <p>A window is a purely internal representation. It is used to
  build and store a potential image of a portion of the terminal.
  It does not bear any necessary relation to what is really on the
  terminal screen; it is more like a scratchpad or write
  buffer.</p>

  <p>To make the section of physical screen corresponding to a
  window reflect the contents of the window structure, the routine
  <code>refresh()</code> (or <code>wrefresh()</code> if the window
  is not <code>stdscr</code>) is called.</p>

  <p>A given physical screen section may be within the scope of any
  number of overlapping windows. Also, changes can be made to
  windows in any order, without regard to motion efficiency. Then,
  at will, the programmer can effectively say &ldquo;make it look
  like this,&rdquo; and let the package implementation determine
  the most efficient way to repaint the screen.</p>

  <h3><a name="stdscr" id="stdscr">Standard Windows and Function
  Naming Conventions</a></h3>

  <p>As hinted above, the routines can use several windows, but two
  are automatically given: <code>curscr</code>, which knows what
  the terminal looks like, and <code>stdscr</code>, which is what
  the programmer wants the terminal to look like next. The user
  should never actually access <code>curscr</code> directly.
  Changes should be made to through the API, and then the routine
  <code>refresh()</code> (or <code>wrefresh()</code>) called.</p>

  <p>Many functions are defined to use <code>stdscr</code> as a
  default screen. For example, to add a character to
  <code>stdscr</code>, one calls <code>addch()</code> with the
  desired character as argument. To write to a different window.
  use the routine <code>waddch()</code> (for
  <strong>w</strong>indow-specific addch()) is provided. This
  convention of prepending function names with a &ldquo;w&rdquo;
  when they are to be applied to specific windows is consistent.
  The only routines which do not follow it are those for which a
  window must always be specified.</p>

  <p>In order to move the current (y, x) coordinates from one point
  to another, the routines <code>move()</code> and
  <code>wmove()</code> are provided. However, it is often desirable
  to first move and then perform some I/O operation. In order to
  avoid clumsiness, most I/O routines can be preceded by the prefix
  &ldquo;mv&rdquo; and the desired (y, x) coordinates prepended to
  the arguments to the function. For example, the calls</p>

  <pre>
          move(y, x);
          addch(ch);
</pre>

  <p>can be replaced by</p>

  <pre>
          mvaddch(y, x, ch);
</pre>

  <p>and</p>

  <pre>
          wmove(win, y, x);
          waddch(win, ch);
</pre>

  <p>can be replaced by</p>

  <pre>
          mvwaddch(win, y, x, ch);
</pre>

  <p>Note that the window description pointer (win) comes before
  the added (y, x) coordinates. If a function requires a window
  pointer, it is always the first parameter passed.</p>

  <h3><a name="variables" id="variables">Variables</a></h3>

  <p>The <code>curses</code> library sets some variables describing
  the terminal capabilities.</p>

  <pre>
      type   name      description
      ------------------------------------------------------------------
      int    LINES     number of lines on the terminal
      int    COLS      number of columns on the terminal
</pre>

  <p>The <code>curses.h</code> also introduces some
  <code>#define</code> constants and types of general
  usefulness:</p>

  <dl>
    <dt><code>bool</code>
    </dt>

    <dd>boolean type, actually a &ldquo;char&rdquo; (e.g.,
    <code>bool doneit;</code>)</dd>

    <dt><code>TRUE</code>
    </dt>

    <dd>boolean &ldquo;true&rdquo; flag (1).</dd>

    <dt><code>FALSE</code>
    </dt>

    <dd>boolean &ldquo;false&rdquo; flag (0).</dd>

    <dt><code>ERR</code>
    </dt>

    <dd>error flag returned by routines on a failure (-1).</dd>

    <dt><code>OK</code>
    </dt>

    <dd>error flag returned by routines when things go right.</dd>
  </dl>

  <h2><a name="using" id="using">Using the Library</a></h2>

  <p>Now we describe how to actually use the screen package. In it,
  we assume all updating, reading, etc. is applied to
  <code>stdscr</code>. These instructions will work on any window,
  providing you change the function names and parameters as
  mentioned above.</p>

  <p>Here is a sample program to motivate the discussion:</p>

  <pre>
#include &lt;stdlib.h&gt;
#include &lt;curses.h&gt;
#include &lt;signal.h&gt;

static void finish(int sig);

int
main(int argc, char *argv[])
{
    int num = 0;

    /* initialize your non-curses data structures here */

    (void) signal(SIGINT, finish);      /* arrange interrupts to terminate */

    (void) initscr();      /* initialize the curses library */
    keypad(stdscr, TRUE);  /* enable keyboard mapping */
    (void) nonl();         /* tell curses not to do NL-&gt;CR/NL on output */
    (void) cbreak();       /* take input chars one at a time, no wait for \n */
    (void) echo();         /* echo input - in color */

    if (has_colors())
    {
        start_color();

        /*
         * Simple color assignment, often all we need.  Color pair 0 cannot
         * be redefined.  This example uses the same value for the color
         * pair as for the foreground color, though of course that is not
         * necessary:
         */
        init_pair(1, COLOR_RED,     COLOR_BLACK);
        init_pair(2, COLOR_GREEN,   COLOR_BLACK);
        init_pair(3, COLOR_YELLOW,  COLOR_BLACK);
        init_pair(4, COLOR_BLUE,    COLOR_BLACK);
        init_pair(5, COLOR_CYAN,    COLOR_BLACK);
        init_pair(6, COLOR_MAGENTA, COLOR_BLACK);
        init_pair(7, COLOR_WHITE,   COLOR_BLACK);
    }

    for (;;)
    {
        int c = getch();     /* refresh, accept single keystroke of input */
        attrset(COLOR_PAIR(num % 8));
        num++;

        /* process the command keystroke */
    }

    finish(0);               /* we are done */
}

static void finish(int sig)
{
    endwin();

    /* do your non-curses wrapup here */

    exit(0);
}
</pre>

  <h3><a name="starting" id="starting">Starting up</a></h3>

  <p>In order to use the screen package, the routines must know
  about terminal characteristics, and the space for
  <code>curscr</code> and <code>stdscr</code> must be allocated.
  These function <code>initscr()</code> does both these things.
  Since it must allocate space for the windows, it can overflow
  memory when attempting to do so. On the rare occasions this
  happens, <code>initscr()</code> will terminate the program with
  an error message. <code>initscr()</code> must always be called
  before any of the routines which affect windows are used. If it
  is not, the program will core dump as soon as either
  <code>curscr</code> or <code>stdscr</code> are referenced.
  However, it is usually best to wait to call it until after you
  are sure you will need it, like after checking for startup
  errors. Terminal status changing routines like <code>nl()</code>
  and <code>cbreak()</code> should be called after
  <code>initscr()</code>.</p>

  <p>Once the screen windows have been allocated, you can set them
  up for your program. If you want to, say, allow a screen to
  scroll, use <code>scrollok()</code>. If you want the cursor to be
  left in place after the last change, use <code>leaveok()</code>.
  If this is not done, <code>refresh()</code> will move the cursor
  to the window's current (y, x) coordinates after updating it.</p>

  <p>You can create new windows of your own using the functions
  <code>newwin()</code>, <code>derwin()</code>, and
  <code>subwin()</code>. The routine <code>delwin()</code> will
  allow you to get rid of old windows. All the options described
  above can be applied to any window.</p>

  <h3><a name="output" id="output">Output</a></h3>

  <p>Now that we have set things up, we will want to actually
  update the terminal. The basic functions used to change what will
  go on a window are <code>addch()</code> and <code>move()</code>.
  <code>addch()</code> adds a character at the current (y, x)
  coordinates. <code>move()</code> changes the current (y, x)
  coordinates to whatever you want them to be. It returns
  <code>ERR</code> if you try to move off the window. As mentioned
  above, you can combine the two into <code>mvaddch()</code> to do
  both things at once.</p>

  <p>The other output functions, such as <code>addstr()</code> and
  <code>printw()</code>, all call <code>addch()</code> to add
  characters to the window.</p>

  <p>After you have put on the window what you want there, when you
  want the portion of the terminal covered by the window to be made
  to look like it, you must call <code>refresh()</code>. In order
  to optimize finding changes, <code>refresh()</code> assumes that
  any part of the window not changed since the last
  <code>refresh()</code> of that window has not been changed on the
  terminal, i.e., that you have not refreshed a portion of the
  terminal with an overlapping window. If this is not the case, the
  routine <code>touchwin()</code> is provided to make it look like
  the entire window has been changed, thus making
  <code>refresh()</code> check the whole subsection of the terminal
  for changes.</p>

  <p>If you call <code>wrefresh()</code> with <code>curscr</code>
  as its argument, it will make the screen look like
  <code>curscr</code> thinks it looks like. This is useful for
  implementing a command which would redraw the screen in case it
  get messed up.</p>

  <h3><a name="input" id="input">Input</a></h3>

  <p>The complementary function to <code>addch()</code> is
  <code>getch()</code> which, if echo is set, will call
  <code>addch()</code> to echo the character. Since the screen
  package needs to know what is on the terminal at all times, if
  characters are to be echoed, the tty must be in raw or cbreak
  mode. Since initially the terminal has echoing enabled and is in
  ordinary &ldquo;cooked&rdquo; mode, one or the other has to
  changed before calling <code>getch()</code>; otherwise, the
  program's output will be unpredictable.</p>

  <p>When you need to accept line-oriented input in a window, the
  functions <code>wgetstr()</code> and friends are available. There
  is even a <code>wscanw()</code> function that can do
  <code>scanf()</code>(3)-style multi-field parsing on window
  input. These pseudo-line-oriented functions turn on echoing while
  they execute.</p>

  <p>The example code above uses the call <code>keypad(stdscr,
  TRUE)</code> to enable support for function-key mapping. With
  this feature, the <code>getch()</code> code watches the input
  stream for character sequences that correspond to arrow and
  function keys. These sequences are returned as pseudo-character
  values. The <code>#define</code> values returned are listed in
  the <code>curses.h</code> The mapping from sequences to
  <code>#define</code> values is determined by <code>key_</code>
  capabilities in the terminal's terminfo entry.</p>

  <h3><a name="formschars" id="formschars">Using Forms
  Characters</a></h3>

  <p>The <code>addch()</code> function (and some others, including
  <code>box()</code> and <code>border()</code>) can accept some
  pseudo-character arguments which are specially defined by
  <code>ncurses</code>. These are <code>#define</code> values set
  up in the <code>curses.h</code> header; see there for a complete
  list (look for the prefix <code>ACS_</code>).</p>

  <p>The most useful of the ACS defines are the forms-drawing
  characters. You can use these to draw boxes and simple graphs on
  the screen. If the terminal does not have such characters,
  <code>curses.h</code> will map them to a recognizable (though
  ugly) set of ASCII defaults.</p>

  <h3><a name="attributes" id="attributes">Character Attributes and
  Color</a></h3>

  <p>The <code>ncurses</code> package supports screen highlights
  including standout, reverse-video, underline, and blink. It also
  supports color, which is treated as another kind of
  highlight.</p>

  <p>Highlights are encoded, internally, as high bits of the
  pseudo-character type (<code>chtype</code>) that
  <code>curses.h</code> uses to represent the contents of a screen
  cell. See the <code>curses.h</code> header file for a complete
  list of highlight mask values (look for the prefix
  <code>A_</code>).</p>

  <p>There are two ways to make highlights. One is to logical-or
  the value of the highlights you want into the character argument
  of an <code>addch()</code> call, or any other output call that
  takes a <code>chtype</code> argument.</p>

  <p>The other is to set the current-highlight value. This is
  <em>logical-OR</em>ed with any highlight you specify the first
  way. You do this with the functions <code>attron()</code>,
  <code>attroff()</code>, and <code>attrset()</code>; see the
  manual pages for details. Color is a special kind of highlight.
  The package actually thinks in terms of color pairs, combinations
  of foreground and background colors. The sample code above sets
  up eight color pairs, all of the guaranteed-available colors on
  black. Note that each color pair is, in effect, given the name of
  its foreground color. Any other range of eight non-conflicting
  values could have been used as the first arguments of the
  <code>init_pair()</code> values.</p>

  <p>Once you have done an <code>init_pair()</code> that creates
  color-pair N, you can use <code>COLOR_PAIR(N)</code> as a
  highlight that invokes that particular color combination. Note
  that <code>COLOR_PAIR(N)</code>, for constant N, is itself a
  compile-time constant and can be used in initializers.</p>

  <h3><a name="mouse" id="mouse">Mouse Interfacing</a></h3>

  <p>The <code>ncurses</code> library also provides a mouse
  interface.</p>

  <blockquote>
    <strong>NOTE:</strong> this facility is specific to
    <code>ncurses</code>, it is not part of either the XSI Curses
    standard, nor of System V Release 4, nor BSD curses. System V
    Release 4 curses contains code with similar interface
    definitions, however it is not documented. Other than by
    disassembling the library, we have no way to determine exactly
    how that mouse code works. Thus, we recommend that you wrap
    mouse-related code in an #ifdef using the feature macro
    NCURSES_MOUSE_VERSION so it will not be compiled and linked on
    non-ncurses systems.
  </blockquote>

  <p>Presently, mouse event reporting works in the following
  environments:</p>

  <ul>
    <li>xterm and similar programs such as rxvt.</li>

    <li>Linux console, when configured with <code>gpm</code>(1),
    Alessandro Rubini's mouse server.</li>

    <li>FreeBSD sysmouse (console)</li>

    <li>OS/2 EMX</li>
  </ul>

  <p>The mouse interface is very simple. To activate it, you use
  the function <code>mousemask()</code>, passing it as first
  argument a bit-mask that specifies what kinds of events you want
  your program to be able to see. It will return the bit-mask of
  events that actually become visible, which may differ from the
  argument if the mouse device is not capable of reporting some of
  the event types you specify.</p>

  <p>Once the mouse is active, your application's command loop
  should watch for a return value of <code>KEY_MOUSE</code> from
  <code>wgetch()</code>. When you see this, a mouse event report
  has been queued. To pick it off the queue, use the function
  <code>getmouse()</code> (you must do this before the next
  <code>wgetch()</code>, otherwise another mouse event might come
  in and make the first one inaccessible).</p>

  <p>Each call to <code>getmouse()</code> fills a structure (the
  address of which you will pass it) with mouse event data. The
  event data includes zero-origin, screen-relative character-cell
  coordinates of the mouse pointer. It also includes an event mask.
  Bits in this mask will be set, corresponding to the event type
  being reported.</p>

  <p>The mouse structure contains two additional fields which may
  be significant in the future as ncurses interfaces to new kinds
  of pointing device. In addition to x and y coordinates, there is
  a slot for a z coordinate; this might be useful with
  touch-screens that can return a pressure or duration parameter.
  There is also a device ID field, which could be used to
  distinguish between multiple pointing devices.</p>

  <p>The class of visible events may be changed at any time via
  <code>mousemask()</code>. Events that can be reported include
  presses, releases, single-, double- and triple-clicks (you can
  set the maximum button-down time for clicks). If you do not make
  clicks visible, they will be reported as press-release pairs. In
  some environments, the event mask may include bits reporting the
  state of shift, alt, and ctrl keys on the keyboard during the
  event.</p>

  <p>A function to check whether a mouse event fell within a given
  window is also supplied. You can use this to see whether a given
  window should consider a mouse event relevant to it.</p>

  <p>Because mouse event reporting will not be available in all
  environments, it would be unwise to build <code>ncurses</code>
  applications that <em>require</em> the use of a mouse. Rather,
  you should use the mouse as a shortcut for point-and-shoot
  commands your application would normally accept from the
  keyboard. Two of the test games in the <code>ncurses</code>
  distribution (<code>bs</code> and <code>knight</code>) contain
  code that illustrates how this can be done.</p>

  <p>See the manual page <code>curs_mouse(3X)</code> for full
  details of the mouse-interface functions.</p>

  <h3><a name="finishing" id="finishing">Finishing Up</a></h3>

  <p>In order to clean up after the <code>ncurses</code> routines,
  the routine <code>endwin()</code> is provided. It restores tty
  modes to what they were when <code>initscr()</code> was first
  called, and moves the cursor down to the lower-left corner. Thus,
  anytime after the call to initscr, <code>endwin()</code> should
  be called before exiting.</p>

  <h2><a name="functions" id="functions">Function Descriptions</a></h2>

  <p>We describe the detailed behavior of some important curses
  functions here, as a supplement to the manual page
  descriptions.</p>

  <h3><a name="init" id="init">Initialization and Wrapup</a></h3>

  <dl>
    <dt><code>initscr()</code>
    </dt>

    <dd>The first function called should almost always be
    <code>initscr()</code>. This will determine the terminal type
    and initialize curses data structures. <code>initscr()</code>
    also arranges that the first call to <code>refresh()</code>
    will clear the screen. If an error occurs a message is written
    to standard error and the program exits. Otherwise it returns a
    pointer to stdscr. A few functions may be called before initscr
    (<code>slk_init()</code>, <code>filter()</code>,
    <code>ripoffline()</code>, <code>use_env()</code>, and, if you
    are using multiple terminals, <code>newterm()</code>.)</dd>

    <dt><code>endwin()</code>
    </dt>

    <dd>Your program should always call <code>endwin()</code>
    before exiting or shelling out of the program. This function
    will restore tty modes, move the cursor to the lower left
    corner of the screen, reset the terminal into the proper
    non-visual mode. Calling <code>refresh()</code> or
    <code>doupdate()</code> after a temporary escape from the
    program will restore the ncurses screen from before the
    escape.</dd>

    <dt><code>newterm(type, ofp, ifp)</code>
    </dt>

    <dd>A program which outputs to more than one terminal should
    use <code>newterm()</code> instead of <code>initscr()</code>.
    <code>newterm()</code> should be called once for each terminal.
    It returns a variable of type <code>SCREEN *</code> which
    should be saved as a reference to that terminal. (NOTE: a
    SCREEN variable is not a <em>screen</em> in the sense we are
    describing in this introduction, but a collection of parameters
    used to assist in optimizing the display.) The arguments are
    the type of the terminal (a string) and <code>FILE</code>
    pointers for the output and input of the terminal. If type is
    NULL then the environment variable <code>$TERM</code> is used.
    <code>endwin()</code> should called once at wrapup time for
    each terminal opened using this function.</dd>

    <dt><code>set_term(new)</code>
    </dt>

    <dd>This function is used to switch to a different terminal
    previously opened by <code>newterm()</code>. The screen
    reference for the new terminal is passed as the parameter. The
    previous terminal is returned by the function. All other calls
    affect only the current terminal.</dd>

    <dt><code>delscreen(sp)</code>
    </dt>

    <dd>The inverse of <code>newterm()</code>; deallocates the data
    structures associated with a given <code>SCREEN</code>
    reference.</dd>
  </dl>

  <h3><a name="flush" id="flush">Causing Output to the Terminal</a></h3>

  <dl>
    <dt><code>refresh()</code> and <code>wrefresh(win)</code></dt>

    <dd>These functions must be called to actually get any output
    on the terminal, as other routines merely manipulate data
    structures. <code>wrefresh()</code> copies the named window to
    the physical terminal screen, taking into account what is
    already there in order to do optimizations.
    <code>refresh()</code> does a refresh of <code>stdscr</code>.
    Unless <code>leaveok()</code> has been enabled, the physical
    cursor of the terminal is left at the location of the window's
    cursor.</dd>

    <dt><code>doupdate()</code> and
    <code>wnoutrefresh(win)</code></dt>

    <dd>These two functions allow multiple updates with more
    efficiency than wrefresh. To use them, it is important to
    understand how curses works. In addition to all the window
    structures, curses keeps two data structures representing the
    terminal screen: a physical screen, describing what is actually
    on the screen, and a virtual screen, describing what the
    programmer wants to have on the screen. wrefresh works by first
    copying the named window to the virtual screen
    (<code>wnoutrefresh()</code>), and then calling the routine to
    update the screen (<code>doupdate()</code>). If the programmer
    wishes to output several windows at once, a series of calls to
    <code>wrefresh</code> will result in alternating calls to
    <code>wnoutrefresh()</code> and <code>doupdate()</code>,
    causing several bursts of output to the screen. By calling
    <code>wnoutrefresh()</code> for each window, it is then
    possible to call <code>doupdate()</code> once, resulting in
    only one burst of output, with fewer total characters
    transmitted (this also avoids a visually annoying flicker at
    each update).</dd>
  </dl>

  <h3><a name="lowlevel" id="lowlevel">Low-Level Capability
  Access</a></h3>

  <dl>
    <dt><code>setupterm(term, filenum, errret)</code>
    </dt>

    <dd>
      This routine is called to initialize a terminal's
      description, without setting up the curses screen structures
      or changing the tty-driver mode bits. <code>term</code> is
      the character string representing the name of the terminal
      being used. <code>filenum</code> is the UNIX file descriptor
      of the terminal to be used for output. <code>errret</code> is
      a pointer to an integer, in which a success or failure
      indication is returned. The values returned can be 1 (all is
      well), 0 (no such terminal), or -1 (some problem locating the
      terminfo database).

      <p>The value of <code>term</code> can be given as NULL, which
      will cause the value of <code>TERM</code> in the environment
      to be used. The <code>errret</code> pointer can also be given
      as NULL, meaning no error code is wanted. If
      <code>errret</code> is defaulted, and something goes wrong,
      <code>setupterm()</code> will print an appropriate error
      message and exit, rather than returning. Thus, a simple
      program can call setupterm(0, 1, 0) and not worry about
      initialization errors.</p>

      <p>After the call to <code>setupterm()</code>, the global
      variable <code>cur_term</code> is set to point to the current
      structure of terminal capabilities. By calling
      <code>setupterm()</code> for each terminal, and saving and
      restoring <code>cur_term</code>, it is possible for a program
      to use two or more terminals at once.
      <code>Setupterm()</code> also stores the names section of the
      terminal description in the global character array
      <code>ttytype[]</code>. Subsequent calls to
      <code>setupterm()</code> will overwrite this array, so you
      will have to save it yourself if need be.</p>
    </dd>
  </dl>

  <h3><a name="debugging" id="debugging">Debugging</a></h3>

  <blockquote>
    <strong>NOTE:</strong> These functions are not part of the
    standard curses API!
  </blockquote>

  <dl>
    <dt><code>trace()</code>
    </dt>

    <dd>This function can be used to explicitly set a trace level.
    If the trace level is nonzero, execution of your program will
    generate a file called &ldquo;trace&rdquo; in the current
    working directory containing a report on the library's actions.
    Higher trace levels enable more detailed (and verbose)
    reporting -- see comments attached to <code>TRACE_</code>
    defines in the <code>curses.h</code> file for details. (It is
    also possible to set a trace level by assigning a trace level
    value to the environment variable
    <code>NCURSES_TRACE</code>).</dd>

    <dt><code>_tracef()</code>
    </dt>

    <dd>This function can be used to output your own debugging
    information. It is only available only if you link with
    -lncurses_g. It can be used the same way as
    <code>printf()</code>, only it outputs a newline after the end
    of arguments. The output goes to a file called
    <code>trace</code> in the current directory.</dd>
  </dl>

  <p>Trace logs can be difficult to interpret due to the sheer
  volume of data dumped in them. There is a script called
  <strong>tracemunch</strong> included with the
  <code>ncurses</code> distribution that can alleviate this problem
  somewhat; it compacts long sequences of similar operations into
  more succinct single-line pseudo-operations. These pseudo-ops can
  be distinguished by the fact that they are named in capital
  letters.</p>

  <h2><a name="hints" id="hints">Hints, Tips, and Tricks</a></h2>

  <p>The <code>ncurses</code> manual pages are a complete reference
  for this library. In the remainder of this document, we discuss
  various useful methods that may not be obvious from the manual
  page descriptions.</p>

  <h3><a name="caution" id="caution">Some Notes of Caution</a></h3>

  <p>If you find yourself thinking you need to use
  <code>noraw()</code> or <code>nocbreak()</code>, think again and
  move carefully. It is probably better design to use
  <code>getstr()</code> or one of its relatives to simulate cooked
  mode. The <code>noraw()</code> and <code>nocbreak()</code>
  functions try to restore cooked mode, but they may end up
  clobbering some control bits set before you started your
  application. Also, they have always been poorly documented, and
  are likely to hurt your application's usability with other curses
  libraries.</p>

  <p>Bear in mind that <code>refresh()</code> is a synonym for
  <code>wrefresh(stdscr)</code>. Do not try to mix use of
  <code>stdscr</code> with use of windows declared by
  <code>newwin()</code>; a <code>refresh()</code> call will blow
  them off the screen. The right way to handle this is to use
  <code>subwin()</code>, or not touch <code>stdscr</code> at all
  and tile your screen with declared windows which you then
  <code>wnoutrefresh()</code> somewhere in your program event loop,
  with a single <code>doupdate()</code> call to trigger actual
  repainting.</p>

  <p>You are much less likely to run into problems if you design
  your screen layouts to use tiled rather than overlapping windows.
  Historically, curses support for overlapping windows has been
  weak, fragile, and poorly documented. The <code>ncurses</code>
  library is not yet an exception to this rule.</p>

  <p>There is a panels library included in the <code>ncurses</code>
  distribution that does a pretty good job of strengthening the
  overlapping-windows facilities.</p>

  <p>Try to avoid using the global variables LINES and COLS. Use
  <code>getmaxyx()</code> on the <code>stdscr</code> context
  instead. Reason: your code may be ported to run in an environment
  with window resizes, in which case several screens could be open
  with different sizes.</p>

  <h3><a name="leaving" id="leaving">Temporarily Leaving NCURSES
  Mode</a></h3>

  <p>Sometimes you will want to write a program that spends most of
  its time in screen mode, but occasionally returns to ordinary
  &ldquo;cooked&rdquo; mode. A common reason for this is to support
  shell-out. This behavior is simple to arrange in
  <code>ncurses</code>.</p>

  <p>To leave <code>ncurses</code> mode, call <code>endwin()</code>
  as you would if you were intending to terminate the program. This
  will take the screen back to cooked mode; you can do your
  shell-out. When you want to return to <code>ncurses</code> mode,
  simply call <code>refresh()</code> or <code>doupdate()</code>.
  This will repaint the screen.</p>

  <p>There is a boolean function, <code>isendwin()</code>, which
  code can use to test whether <code>ncurses</code> screen mode is
  active. It returns <code>TRUE</code> in the interval between an
  <code>endwin()</code> call and the following
  <code>refresh()</code>, <code>FALSE</code> otherwise.</p>

  <p>Here is some sample code for shellout:</p>

  <pre>
    addstr("Shelling out...");
    def_prog_mode();           /* save current tty modes */
    endwin();                  /* restore original tty modes */
    system("sh");              /* run shell */
    addstr("returned.\n");     /* prepare return message */
    refresh();                 /* restore save modes, repaint screen */
</pre>

  <h3><a name="xterm" id="xterm">Using NCURSES under XTERM</a></h3>

  <p>A resize operation in X sends <code>SIGWINCH</code> to the
  application running under xterm. The easiest way to handle
  <code>SIGWINCH</code> is to do an <code>endwin</code>, followed
  by an <code>refresh</code> and a screen repaint you code
  yourself. The <code>refresh</code> will pick up the new screen
  size from the xterm's environment.</p>

  <p>That is the standard way, of course (it even works with some
  vendor's curses implementations). Its drawback is that it clears
  the screen to reinitialize the display, and does not resize
  subwindows which must be shrunk. <code>Ncurses</code> provides an
  extension which works better, the <code>resizeterm</code>
  function. That function ensures that all windows are limited to
  the new screen dimensions, and pads <code>stdscr</code> with
  blanks if the screen is larger.</p>

  <p>The <code>ncurses</code> library provides a SIGWINCH signal
  handler, which pushes a <code>KEY_RESIZE</code> via the wgetch()
  calls. When <code>ncurses</code> returns that code, it calls
  <code>resizeterm</code> to update the size of the standard
  screen's window, repainting that (filling with blanks or
  truncating as needed). It also resizes other windows, but its
  effect may be less satisfactory because it cannot know how you
  want the screen re-painted. You will usually have to write
  special-purpose code to handle <code>KEY_RESIZE</code>
  yourself.</p>

  <h3><a name="screens" id="screens">Handling Multiple Terminal
  Screens</a></h3>

  <p>The <code>initscr()</code> function actually calls a function
  named <code>newterm()</code> to do most of its work. If you are
  writing a program that opens multiple terminals, use
  <code>newterm()</code> directly.</p>

  <p>For each call, you will have to specify a terminal type and a
  pair of file pointers; each call will return a screen reference,
  and <code>stdscr</code> will be set to the last one allocated.
  You will switch between screens with the <code>set_term</code>
  call. Note that you will also have to call
  <code>def_shell_mode</code> and <code>def_prog_mode</code> on
  each tty yourself.</p>

  <h3><a name="testing" id="testing">Testing for Terminal
  Capabilities</a></h3>

  <p>Sometimes you may want to write programs that test for the
  presence of various capabilities before deciding whether to go
  into <code>ncurses</code> mode. An easy way to do this is to call
  <code>setupterm()</code>, then use the functions
  <code>tigetflag()</code>, <code>tigetnum()</code>, and
  <code>tigetstr()</code> to do your testing.</p>

  <p>A particularly useful case of this often comes up when you
  want to test whether a given terminal type should be treated as
  &ldquo;smart&rdquo; (cursor-addressable) or &ldquo;stupid&rdquo;.
  The right way to test this is to see if the return value of
  <code>tigetstr("cup")</code> is non-NULL. Alternatively, you can
  include the <code>term.h</code> file and test the value of the
  macro <code>cursor_address</code>.</p>

  <h3><a name="tuning" id="tuning">Tuning for Speed</a></h3>

  <p>Use the <code>addchstr()</code> family of functions for fast
  screen-painting of text when you know the text does not contain
  any control characters. Try to make attribute changes infrequent
  on your screens. Do not use the <code>immedok()</code>
  option!</p>

  <h3><a name="special" id="special">Special Features of
  NCURSES</a></h3>

  <p>The <code>wresize()</code> function allows you to resize a
  window in place. The associated <code>resizeterm()</code>
  function simplifies the construction of <a href=
  "#xterm">SIGWINCH</a> handlers, for resizing all windows.</p>

  <p>The <code>define_key()</code> function allows you to define at
  runtime function-key control sequences which are not in the
  terminal description. The <code>keyok()</code> function allows
  you to temporarily enable or disable interpretation of any
  function-key control sequence.</p>

  <p>The <code>use_default_colors()</code> function allows you to
  construct applications which can use the terminal's default
  foreground and background colors as an additional "default"
  color. Several terminal emulators support this feature, which is
  based on ISO 6429.</p>

  <p>Ncurses supports up 16 colors, unlike SVr4 curses which
  defines only 8. While most terminals which provide color allow
  only 8 colors, about a quarter (including XFree86 xterm) support
  16 colors.</p>

  <h2><a name="compat" id="compat">Compatibility with Older
  Versions</a></h2>

  <p>Despite our best efforts, there are some differences between
  <code>ncurses</code> and the (undocumented!) behavior of older
  curses implementations. These arise from ambiguities or omissions
  in the documentation of the API.</p>

  <h3><a name="refbug" id="refbug">Refresh of Overlapping
  Windows</a></h3>

  <p>If you define two windows A and B that overlap, and then
  alternately scribble on and refresh them, the changes made to the
  overlapping region under historic <code>curses</code> versions
  were often not documented precisely.</p>

  <p>To understand why this is a problem, remember that screen
  updates are calculated between two representations of the
  <em>entire</em> display. The documentation says that when you
  refresh a window, it is first copied to the virtual screen, and
  then changes are calculated to update the physical screen (and
  applied to the terminal). But "copied to" is not very specific,
  and subtle differences in how copying works can produce different
  behaviors in the case where two overlapping windows are each
  being refreshed at unpredictable intervals.</p>

  <p>What happens to the overlapping region depends on what
  <code>wnoutrefresh()</code> does with its argument -- what
  portions of the argument window it copies to the virtual screen.
  Some implementations do "change copy", copying down only
  locations in the window that have changed (or been marked changed
  with <code>wtouchln()</code> and friends). Some implementations
  do "entire copy", copying <em>all</em> window locations to the
  virtual screen whether or not they have changed.</p>

  <p>The <code>ncurses</code> library itself has not always been
  consistent on this score. Due to a bug, versions 1.8.7 to 1.9.8a
  did entire copy. Versions 1.8.6 and older, and versions 1.9.9 and
  newer, do change copy.</p>

  <p>For most commercial curses implementations, it is not
  documented and not known for sure (at least not to the
  <code>ncurses</code> maintainers) whether they do change copy or
  entire copy. We know that System V release 3 curses has logic in
  it that looks like an attempt to do change copy, but the
  surrounding logic and data representations are sufficiently
  complex, and our knowledge sufficiently indirect, that it is hard
  to know whether this is reliable. It is not clear what the SVr4
  documentation and XSI standard intend. The XSI Curses standard
  barely mentions wnoutrefresh(); the SVr4 documents seem to be
  describing entire-copy, but it is possible with some effort and
  straining to read them the other way.</p>

  <p>It might therefore be unwise to rely on either behavior in
  programs that might have to be linked with other curses
  implementations. Instead, you can do an explicit
  <code>touchwin()</code> before the <code>wnoutrefresh()</code>
  call to guarantee an entire-contents copy anywhere.</p>

  <p>The really clean way to handle this is to use the panels
  library. If, when you want a screen update, you do
  <code>update_panels()</code>, it will do all the necessary
  <code>wnoutrefresh()</code> calls for whatever panel stacking
  order you have defined. Then you can do one
  <code>doupdate()</code> and there will be a <em>single</em> burst
  of physical I/O that will do all your updates.</p>

  <h3><a name="backbug" id="backbug">Background Erase</a></h3>

  <p>If you have been using a very old versions of
  <code>ncurses</code> (1.8.7 or older) you may be surprised by the
  behavior of the erase functions. In older versions, erased areas
  of a window were filled with a blank modified by the window's
  current attribute (as set by <strong>wattrset()</strong>,
  <strong>wattron()</strong>, <strong>wattroff()</strong> and
  friends).</p>

  <p>In newer versions, this is not so. Instead, the attribute of
  erased blanks is normal unless and until it is modified by the
  functions <code>bkgdset()</code> or <code>wbkgdset()</code>.</p>

  <p>This change in behavior conforms <code>ncurses</code> to
  System V Release 4 and the XSI Curses standard.</p>

  <h2><a name="xsifuncs" id="xsifuncs">XSI Curses Conformance</a></h2>

  <p>The <code>ncurses</code> library is intended to be base-level
  conformant with the XSI Curses standard from X/Open. Many
  extended-level features (in fact, almost all features not
  directly concerned with wide characters and internationalization)
  are also supported.</p>

  <p>One effect of XSI conformance is the change in behavior
  described under <a href="#backbug">"Background Erase --
  Compatibility with Old Versions"</a>.</p>

  <p>Also, <code>ncurses</code> meets the XSI requirement that
  every macro entry point have a corresponding function which may
  be linked (and will be prototype-checked) if the macro definition
  is disabled with <code>#undef</code>.</p>

  <h1><a name="panels" id="panels">The Panels Library</a></h1>

  <p>The <code>ncurses</code> library by itself provides good
  support for screen displays in which the windows are tiled
  (non-overlapping). In the more general case that windows may
  overlap, you have to use a series of <code>wnoutrefresh()</code>
  calls followed by a <code>doupdate()</code>, and be careful about
  the order you do the window refreshes in. It has to be
  bottom-upwards, otherwise parts of windows that should be
  obscured will show through.</p>

  <p>When your interface design is such that windows may dive
  deeper into the visibility stack or pop to the top at runtime,
  the resulting book-keeping can be tedious and difficult to get
  right. Hence the panels library.</p>

  <p>The <code>panel</code> library first appeared in AT&amp;T
  System V. The version documented here is the <code>panel</code>
  code distributed with <code>ncurses</code>.</p>

  <h2><a name="pcompile" id="pcompile">Compiling With the Panels
  Library</a></h2>

  <p>Your panels-using modules must import the panels library
  declarations with</p>

  <pre>
          #include &lt;panel.h&gt;
</pre>

  <p>and must be linked explicitly with the panels library using an
  <code>-lpanel</code> argument. Note that they must also link the
  <code>ncurses</code> library with <code>-lncurses</code>. Many
  linkers are two-pass and will accept either order, but it is
  still good practice to put <code>-lpanel</code> first and
  <code>-lncurses</code> second.</p>

  <h2><a name="poverview" id="poverview">Overview of Panels</a></h2>

  <p>A panel object is a window that is implicitly treated as part
  of a <dfn>deck</dfn> including all other panel objects. The deck
  has an implicit bottom-to-top visibility order. The panels
  library includes an update function (analogous to
  <code>refresh()</code>) that displays all panels in the deck in
  the proper order to resolve overlaps. The standard window,
  <code>stdscr</code>, is considered below all panels.</p>

  <p>Details on the panels functions are available in the man
  pages. We will just hit the highlights here.</p>

  <p>You create a panel from a window by calling
  <code>new_panel()</code> on a window pointer. It then becomes the
  top of the deck. The panel's window is available as the value of
  <code>panel_window()</code> called with the panel pointer as
  argument.</p>

  <p>You can delete a panel (removing it from the deck) with
  <code>del_panel</code>. This will not deallocate the associated
  window; you have to do that yourself. You can replace a panel's
  window with a different window by calling
  <code>replace_window</code>. The new window may be of different
  size; the panel code will re-compute all overlaps. This operation
  does not change the panel's position in the deck.</p>

  <p>To move a panel's window, use <code>move_panel()</code>. The
  <code>mvwin()</code> function on the panel's window is not
  sufficient because it does not update the panels library's
  representation of where the windows are. This operation leaves
  the panel's depth, contents, and size unchanged.</p>

  <p>Two functions (<code>top_panel()</code>,
  <code>bottom_panel()</code>) are provided for rearranging the
  deck. The first pops its argument window to the top of the deck;
  the second sends it to the bottom. Either operation leaves the
  panel's screen location, contents, and size unchanged.</p>

  <p>The function <code>update_panels()</code> does all the
  <code>wnoutrefresh()</code> calls needed to prepare for
  <code>doupdate()</code> (which you must call yourself,
  afterwards).</p>

  <p>Typically, you will want to call <code>update_panels()</code>
  and <code>doupdate()</code> just before accepting command input,
  once in each cycle of interaction with the user. If you call
  <code>update_panels()</code> after each and every panel write,
  you will generate a lot of unnecessary refresh activity and
  screen flicker.</p>

  <h2><a name="pstdscr" id="pstdscr">Panels, Input, and the
  Standard Screen</a></h2>

  <p>You should not mix <code>wnoutrefresh()</code> or
  <code>wrefresh()</code> operations with panels code; this will
  work only if the argument window is either in the top panel or
  unobscured by any other panels.</p>

  <p>The <code>stsdcr</code> window is a special case. It is
  considered below all panels. Because changes to panels may
  obscure parts of <code>stdscr</code>, though, you should call
  <code>update_panels()</code> before <code>doupdate()</code> even
  when you only change <code>stdscr</code>.</p>

  <p>Note that <code>wgetch</code> automatically calls
  <code>wrefresh</code>. Therefore, before requesting input from a
  panel window, you need to be sure that the panel is totally
  unobscured.</p>

  <p>There is presently no way to display changes to one obscured
  panel without repainting all panels.</p>

  <h2><a name="hiding" id="hiding">Hiding Panels</a></h2>

  <p>It is possible to remove a panel from the deck temporarily;
  use <code>hide_panel</code> for this. Use
  <code>show_panel()</code> to render it visible again. The
  predicate function <code>panel_hidden</code> tests whether or not
  a panel is hidden.</p>

  <p>The <code>panel_update</code> code ignores hidden panels. You
  cannot do <code>top_panel()</code> or <code>bottom_panel</code>
  on a hidden panel(). Other panels operations are applicable.</p>

  <h2><a name="pmisc" id="pmisc">Miscellaneous Other Facilities</a></h2>

  <p>It is possible to navigate the deck using the functions
  <code>panel_above()</code> and <code>panel_below</code>. Handed a
  panel pointer, they return the panel above or below that panel.
  Handed <code>NULL</code>, they return the bottom-most or top-most
  panel.</p>

  <p>Every panel has an associated user pointer, not used by the
  panel code, to which you can attach application data. See the man
  page documentation of <code>set_panel_userptr()</code> and
  <code>panel_userptr</code> for details.</p>

  <h1><a name="menu" id="menu">The Menu Library</a></h1>

  <p>A menu is a screen display that assists the user to choose
  some subset of a given set of items. The <code>menu</code>
  library is a curses extension that supports easy programming of
  menu hierarchies with a uniform but flexible interface.</p>

  <p>The <code>menu</code> library first appeared in AT&amp;T
  System V. The version documented here is the <code>menu</code>
  code distributed with <code>ncurses</code>.</p>

  <h2><a name="mcompile" id="mcompile">Compiling With the menu
  Library</a></h2>

  <p>Your menu-using modules must import the menu library
  declarations with</p>

  <pre>
          #include &lt;menu.h&gt;
</pre>

  <p>and must be linked explicitly with the menus library using an
  <code>-lmenu</code> argument. Note that they must also link the
  <code>ncurses</code> library with <code>-lncurses</code>. Many
  linkers are two-pass and will accept either order, but it is
  still good practice to put <code>-lmenu</code> first and
  <code>-lncurses</code> second.</p>

  <h2><a name="moverview" id="moverview">Overview of Menus</a></h2>

  <p>The menus created by this library consist of collections of
  <dfn>items</dfn> including a name string part and a description
  string part. To make menus, you create groups of these items and
  connect them with menu frame objects.</p>

  <p>The menu can then by <dfn>posted</dfn>, that is written to an
  associated window. Actually, each menu has two associated
  windows; a containing window in which the programmer can scribble
  titles or borders, and a subwindow in which the menu items proper
  are displayed. If this subwindow is too small to display all the
  items, it will be a scrollable viewport on the collection of
  items.</p>

  <p>A menu may also be <dfn>unposted</dfn> (that is, undisplayed),
  and finally freed to make the storage associated with it and its
  items available for re-use.</p>

  <p>The general flow of control of a menu program looks like
  this:</p>

  <ol>
    <li>Initialize <code>curses</code>.</li>

    <li>Create the menu items, using <code>new_item()</code>.</li>

    <li>Create the menu using <code>new_menu()</code>.</li>

    <li>Post the menu using <code>post_menu()</code>.</li>

    <li>Refresh the screen.</li>

    <li>Process user requests via an input loop.</li>

    <li>Unpost the menu using <code>unpost_menu()</code>.</li>

    <li>Free the menu, using <code>free_menu()</code>.</li>

    <li>Free the items using <code>free_item()</code>.</li>

    <li>Terminate <code>curses</code>.</li>
  </ol>

  <h2><a name="mselect" id="mselect">Selecting items</a></h2>

  <p>Menus may be multi-valued or (the default) single-valued (see
  the manual page <code>menu_opts(3x)</code> to see how to change
  the default). Both types always have a <dfn>current
  item</dfn>.</p>

  <p>From a single-valued menu you can read the selected value
  simply by looking at the current item. From a multi-valued menu,
  you get the selected set by looping through the items applying
  the <code>item_value()</code> predicate function. Your
  menu-processing code can use the function
  <code>set_item_value()</code> to flag the items in the select
  set.</p>

  <p>Menu items can be made unselectable using
  <code>set_item_opts()</code> or <code>item_opts_off()</code> with
  the <code>O_SELECTABLE</code> argument. This is the only option
  so far defined for menus, but it is good practice to code as
  though other option bits might be on.</p>

  <h2><a name="mdisplay" id="mdisplay">Menu Display</a></h2>

  <p>The menu library calculates a minimum display size for your
  window, based on the following variables:</p>

  <ul>
    <li>The number and maximum length of the menu items</li>

    <li>Whether the O_ROWMAJOR option is enabled</li>

    <li>Whether display of descriptions is enabled</li>

    <li>Whatever menu format may have been set by the
    programmer</li>

    <li>The length of the menu mark string used for highlighting
    selected items</li>
  </ul>

  <p>The function <code>set_menu_format()</code> allows you to set
  the maximum size of the viewport or <dfn>menu page</dfn> that
  will be used to display menu items. You can retrieve any format
  associated with a menu with <code>menu_format()</code>. The
  default format is rows=16, columns=1.</p>

  <p>The actual menu page may be smaller than the format size. This
  depends on the item number and size and whether O_ROWMAJOR is on.
  This option (on by default) causes menu items to be displayed in
  a &ldquo;raster-scan&rdquo; pattern, so that if more than one
  item will fit horizontally the first couple of items are
  side-by-side in the top row. The alternative is column-major
  display, which tries to put the first several items in the first
  column.</p>

  <p>As mentioned above, a menu format not large enough to allow
  all items to fit on-screen will result in a menu display that is
  vertically scrollable.</p>

  <p>You can scroll it with requests to the menu driver, which will
  be described in the section on <a href="#minput">menu input
  handling</a>.</p>

  <p>Each menu has a <dfn>mark string</dfn> used to visually tag
  selected items; see the <code>menu_mark(3x)</code> manual page
  for details. The mark string length also influences the menu page
  size.</p>

  <p>The function <code>scale_menu()</code> returns the minimum
  display size that the menu code computes from all these factors.
  There are other menu display attributes including a select
  attribute, an attribute for selectable items, an attribute for
  unselectable items, and a pad character used to separate item
  name text from description text. These have reasonable defaults
  which the library allows you to change (see the
  <code>menu_attribs(3x)</code> manual page.</p>

  <h2><a name="mwindows" id="mwindows">Menu Windows</a></h2>

  <p>Each menu has, as mentioned previously, a pair of associated
  windows. Both these windows are painted when the menu is posted
  and erased when the menu is unposted.</p>

  <p>The outer or frame window is not otherwise touched by the menu
  routines. It exists so the programmer can associate a title, a
  border, or perhaps help text with the menu and have it properly
  refreshed or erased at post/unpost time. The inner window or
  <dfn>subwindow</dfn> is where the current menu page is
  displayed.</p>

  <p>By default, both windows are <code>stdscr</code>. You can set
  them with the functions in <code>menu_win(3x)</code>.</p>

  <p>When you call <code>post_menu()</code>, you write the menu to
  its subwindow. When you call <code>unpost_menu()</code>, you
  erase the subwindow, However, neither of these actually modifies
  the screen. To do that, call <code>wrefresh()</code> or some
  equivalent.</p>

  <h2><a name="minput" id="minput">Processing Menu Input</a></h2>

  <p>The main loop of your menu-processing code should call
  <code>menu_driver()</code> repeatedly. The first argument of this
  routine is a menu pointer; the second is a menu command code. You
  should write an input-fetching routine that maps input characters
  to menu command codes, and pass its output to
  <code>menu_driver()</code>. The menu command codes are fully
  documented in <code>menu_driver(3x)</code>.</p>

  <p>The simplest group of command codes is
  <code>REQ_NEXT_ITEM</code>, <code>REQ_PREV_ITEM</code>,
  <code>REQ_FIRST_ITEM</code>, <code>REQ_LAST_ITEM</code>,
  <code>REQ_UP_ITEM</code>, <code>REQ_DOWN_ITEM</code>,
  <code>REQ_LEFT_ITEM</code>, <code>REQ_RIGHT_ITEM</code>. These
  change the currently selected item. These requests may cause
  scrolling of the menu page if it only partially displayed.</p>

  <p>There are explicit requests for scrolling which also change
  the current item (because the select location does not change,
  but the item there does). These are <code>REQ_SCR_DLINE</code>,
  <code>REQ_SCR_ULINE</code>, <code>REQ_SCR_DPAGE</code>, and
  <code>REQ_SCR_UPAGE</code>.</p>

  <p>The <code>REQ_TOGGLE_ITEM</code> selects or deselects the
  current item. It is for use in multi-valued menus; if you use it
  with <code>O_ONEVALUE</code> on, you will get an error return
  (<code>E_REQUEST_DENIED</code>).</p>

  <p>Each menu has an associated pattern buffer. The
  <code>menu_driver()</code> logic tries to accumulate printable
  ASCII characters passed in in that buffer; when it matches a
  prefix of an item name, that item (or the next matching item) is
  selected. If appending a character yields no new match, that
  character is deleted from the pattern buffer, and
  <code>menu_driver()</code> returns <code>E_NO_MATCH</code>.</p>

  <p>Some requests change the pattern buffer directly:
  <code>REQ_CLEAR_PATTERN</code>, <code>REQ_BACK_PATTERN</code>,
  <code>REQ_NEXT_MATCH</code>, <code>REQ_PREV_MATCH</code>. The
  latter two are useful when pattern buffer input matches more than
  one item in a multi-valued menu.</p>

  <p>Each successful scroll or item navigation request clears the
  pattern buffer. It is also possible to set the pattern buffer
  explicitly with <code>set_menu_pattern()</code>.</p>

  <p>Finally, menu driver requests above the constant
  <code>MAX_COMMAND</code> are considered application-specific
  commands. The <code>menu_driver()</code> code ignores them and
  returns <code>E_UNKNOWN_COMMAND</code>.</p>

  <h2><a name="mmisc" id="mmisc">Miscellaneous Other Features</a></h2>

  <p>Various menu options can affect the processing and visual
  appearance and input processing of menus. See <code>menu_opts(3x)
  for details.</code></p>

  <p>It is possible to change the current item from application
  code; this is useful if you want to write your own navigation
  requests. It is also possible to explicitly set the top row of
  the menu display. See <code>mitem_current(3x)</code>. If your
  application needs to change the menu subwindow cursor for any
  reason, <code>pos_menu_cursor()</code> will restore it to the
  correct location for continuing menu driver processing.</p>

  <p>It is possible to set hooks to be called at menu
  initialization and wrapup time, and whenever the selected item
  changes. See <code>menu_hook(3x)</code>.</p>

  <p>Each item, and each menu, has an associated user pointer on
  which you can hang application data. See
  <code>mitem_userptr(3x)</code> and
  <code>menu_userptr(3x)</code>.</p>

  <h1><a name="form" id="form">The Forms Library</a></h1>

  <p>The <code>form</code> library is a curses extension that
  supports easy programming of on-screen forms for data entry and
  program control.</p>

  <p>The <code>form</code> library first appeared in AT&amp;T
  System V. The version documented here is the <code>form</code>
  code distributed with <code>ncurses</code>.</p>

  <h2><a name="fcompile" id="fcompile">Compiling With the form
  Library</a></h2>

  <p>Your form-using modules must import the form library
  declarations with</p>

  <pre>
          #include &lt;form.h&gt;
</pre>

  <p>and must be linked explicitly with the forms library using an
  <code>-lform</code> argument. Note that they must also link the
  <code>ncurses</code> library with <code>-lncurses</code>. Many
  linkers are two-pass and will accept either order, but it is
  still good practice to put <code>-lform</code> first and
  <code>-lncurses</code> second.</p>

  <h2><a name="foverview" id="foverview">Overview of Forms</a></h2>

  <p>A form is a collection of fields; each field may be either a
  label (explanatory text) or a data-entry location. Long forms may
  be segmented into pages; each entry to a new page clears the
  screen.</p>

  <p>To make forms, you create groups of fields and connect them
  with form frame objects; the form library makes this relatively
  simple.</p>

  <p>Once defined, a form can be <dfn>posted</dfn>, that is written
  to an associated window. Actually, each form has two associated
  windows; a containing window in which the programmer can scribble
  titles or borders, and a subwindow in which the form fields
  proper are displayed.</p>

  <p>As the form user fills out the posted form, navigation and
  editing keys support movement between fields, editing keys
  support modifying field, and plain text adds to or changes data
  in a current field. The form library allows you (the forms
  designer) to bind each navigation and editing key to any
  keystroke accepted by <code>curses</code> Fields may have
  validation conditions on them, so that they check input data for
  type and value. The form library supplies a rich set of
  pre-defined field types, and makes it relatively easy to define
  new ones.</p>

  <p>Once its transaction is completed (or aborted), a form may be
  <dfn>unposted</dfn> (that is, undisplayed), and finally freed to
  make the storage associated with it and its items available for
  re-use.</p>

  <p>The general flow of control of a form program looks like
  this:</p>

  <ol>
    <li>Initialize <code>curses</code>.</li>

    <li>Create the form fields, using
    <code>new_field()</code>.</li>

    <li>Create the form using <code>new_form()</code>.</li>

    <li>Post the form using <code>post_form()</code>.</li>

    <li>Refresh the screen.</li>

    <li>Process user requests via an input loop.</li>

    <li>Unpost the form using <code>unpost_form()</code>.</li>

    <li>Free the form, using <code>free_form()</code>.</li>

    <li>Free the fields using <code>free_field()</code>.</li>

    <li>Terminate <code>curses</code>.</li>
  </ol>

  <p>Note that this looks much like a menu program; the form
  library handles tasks which are in many ways similar, and its
  interface was obviously designed to resemble that of the <a href=
  "#menu">menu library</a> wherever possible.</p>

  <p>In forms programs, however, the &ldquo;process user
  requests&rdquo; is somewhat more complicated than for menus.
  Besides menu-like navigation operations, the menu driver loop has
  to support field editing and data validation.</p>

  <h2><a name="fcreate" id="fcreate">Creating and Freeing Fields
  and Forms</a></h2>

  <p>The basic function for creating fields is
  <code>new_field()</code>:</p>

  <pre>
FIELD *new_field(int height, int width,   /* new field size */
                 int top, int left,       /* upper left corner */
                 int offscreen,           /* number of offscreen rows */
                 int nbuf);               /* number of working buffers */
</pre>

  <p>Menu items always occupy a single row, but forms fields may
  have multiple rows. So <code>new_field()</code> requires you to
  specify a width and height (the first two arguments, which mist
  both be greater than zero).</p>

  <p>You must also specify the location of the field's upper left
  corner on the screen (the third and fourth arguments, which must
  be zero or greater). Note that these coordinates are relative to
  the form subwindow, which will coincide with <code>stdscr</code>
  by default but need not be <code>stdscr</code> if you have done
  an explicit <code>set_form_win()</code> call.</p>

  <p>The fifth argument allows you to specify a number of
  off-screen rows. If this is zero, the entire field will always be
  displayed. If it is nonzero, the form will be scrollable, with
  only one screen-full (initially the top part) displayed at any
  given time. If you make a field dynamic and grow it so it will no
  longer fit on the screen, the form will become scrollable even if
  the <code>offscreen</code> argument was initially zero.</p>

  <p>The forms library allocates one working buffer per field; the
  size of each buffer is <code>((height + offscreen)*width +
  1</code>, one character for each position in the field plus a NUL
  terminator. The sixth argument is the number of additional data
  buffers to allocate for the field; your application can use them
  for its own purposes.</p>

  <pre>
FIELD *dup_field(FIELD *field,            /* field to copy */
                 int top, int left);      /* location of new copy */
</pre>

  <p>The function <code>dup_field()</code> duplicates an existing
  field at a new location. Size and buffering information are
  copied; some attribute flags and status bits are not (see the
  <code>form_field_new(3X)</code> for details).</p>

  <pre>
FIELD *link_field(FIELD *field,           /* field to copy */
                  int top, int left);     /* location of new copy */
</pre>

  <p>The function <code>link_field()</code> also duplicates an
  existing field at a new location. The difference from
  <code>dup_field()</code> is that it arranges for the new field's
  buffer to be shared with the old one.</p>

  <p>Besides the obvious use in making a field editable from two
  different form pages, linked fields give you a way to hack in
  dynamic labels. If you declare several fields linked to an
  original, and then make them inactive, changes from the original
  will still be propagated to the linked fields.</p>

  <p>As with duplicated fields, linked fields have attribute bits
  separate from the original.</p>

  <p>As you might guess, all these field-allocations return
  <code>NULL</code> if the field allocation is not possible due to
  an out-of-memory error or out-of-bounds arguments.</p>

  <p>To connect fields to a form, use</p>

  <pre>
FORM *new_form(FIELD **fields);
</pre>

  <p>This function expects to see a NULL-terminated array of field
  pointers. Said fields are connected to a newly-allocated form
  object; its address is returned (or else NULL if the allocation
  fails).</p>

  <p>Note that <code>new_field()</code> does <em>not</em> copy the
  pointer array into private storage; if you modify the contents of
  the pointer array during forms processing, all manner of bizarre
  things might happen. Also note that any given field may only be
  connected to one form.</p>

  <p>The functions <code>free_field()</code> and
  <code>free_form</code> are available to free field and form
  objects. It is an error to attempt to free a field connected to a
  form, but not vice-versa; thus, you will generally free your form
  objects first.</p>

  <h2><a name="fattributes" id="fattributes">Fetching and Changing
  Field Attributes</a></h2>

  <p>Each form field has a number of location and size attributes
  associated with it. There are other field attributes used to
  control display and editing of the field. Some (for example, the
  <code>O_STATIC</code> bit) involve sufficient complications to be
  covered in sections of their own later on. We cover the functions
  used to get and set several basic attributes here.</p>

  <p>When a field is created, the attributes not specified by the
  <code>new_field</code> function are copied from an invisible
  system default field. In attribute-setting and -fetching
  functions, the argument NULL is taken to mean this field. Changes
  to it persist as defaults until your forms application
  terminates.</p>

  <h3><a name="fsizes" id="fsizes">Fetching Size and Location
  Data</a></h3>

  <p>You can retrieve field sizes and locations through:</p>

  <pre>
int field_info(FIELD *field,              /* field from which to fetch */
               int *height, *int width,   /* field size */
               int *top, int *left,       /* upper left corner */
               int *offscreen,            /* number of offscreen rows */
               int *nbuf);                /* number of working buffers */
</pre>

  <p>This function is a sort of inverse of
  <code>new_field()</code>; instead of setting size and location
  attributes of a new field, it fetches them from an existing
  one.</p>

  <h3><a name="flocation" id="flocation">Changing the Field
  Location</a></h3>

  <p>It is possible to move a field's location on the screen:</p>

  <pre>
int move_field(FIELD *field,              /* field to alter */
               int top, int left);        /* new upper-left corner */
</pre>

  <p>You can, of course. query the current location through
  <code>field_info()</code>.</p>

  <h3><a name="fjust" id="fjust">The Justification Attribute</a></h3>

  <p>One-line fields may be unjustified, justified right, justified
  left, or centered. Here is how you manipulate this attribute:</p>

  <pre>
int set_field_just(FIELD *field,          /* field to alter */
                   int justmode);         /* mode to set */

int field_just(FIELD *field);             /* fetch mode of field */
</pre>

  <p>The mode values accepted and returned by this functions are
  preprocessor macros <code>NO_JUSTIFICATION</code>,
  <code>JUSTIFY_RIGHT</code>, <code>JUSTIFY_LEFT</code>, or
  <code>JUSTIFY_CENTER</code>.</p>

  <h3><a name="fdispatts" id="fdispatts">Field Display
  Attributes</a></h3>

  <p>For each field, you can set a foreground attribute for entered
  characters, a background attribute for the entire field, and a
  pad character for the unfilled portion of the field. You can also
  control pagination of the form.</p>

  <p>This group of four field attributes controls the visual
  appearance of the field on the screen, without affecting in any
  way the data in the field buffer.</p>

  <pre>
int set_field_fore(FIELD *field,          /* field to alter */
                   chtype attr);          /* attribute to set */

chtype field_fore(FIELD *field);          /* field to query */

int set_field_back(FIELD *field,          /* field to alter */
                   chtype attr);          /* attribute to set */

chtype field_back(FIELD *field);          /* field to query */

int set_field_pad(FIELD *field,           /* field to alter */
                 int pad);                /* pad character to set */

chtype field_pad(FIELD *field);

int set_new_page(FIELD *field,            /* field to alter */
                 int flag);               /* TRUE to force new page */

chtype new_page(FIELD *field);            /* field to query */
</pre>

  <p>The attributes set and returned by the first four functions
  are normal <code>curses(3x)</code> display attribute values
  (<code>A_STANDOUT</code>, <code>A_BOLD</code>,
  <code>A_REVERSE</code> etc). The page bit of a field controls
  whether it is displayed at the start of a new form screen.</p>

  <h3><a name="foptions" id="foptions">Field Option Bits</a></h3>

  <p>There is also a large collection of field option bits you can
  set to control various aspects of forms processing. You can
  manipulate them with these functions:</p>

  <pre>
int set_field_opts(FIELD *field,          /* field to alter */
                   int attr);             /* attribute to set */

int field_opts_on(FIELD *field,           /* field to alter */
                  int attr);              /* attributes to turn on */

int field_opts_off(FIELD *field,          /* field to alter */
                   int attr);             /* attributes to turn off */

int field_opts(FIELD *field);             /* field to query */
</pre>

  <p>By default, all options are on. Here are the available option
  bits:</p>

  <dl>
    <dt>O_VISIBLE</dt>

    <dd>Controls whether the field is visible on the screen. Can be
    used during form processing to hide or pop up fields depending
    on the value of parent fields.</dd>

    <dt>O_ACTIVE</dt>

    <dd>Controls whether the field is active during forms
    processing (i.e. visited by form navigation keys). Can be used
    to make labels or derived fields with buffer values alterable
    by the forms application, not the user.</dd>

    <dt>O_PUBLIC</dt>

    <dd>Controls whether data is displayed during field entry. If
    this option is turned off on a field, the library will accept
    and edit data in that field, but it will not be displayed and
    the visible field cursor will not move. You can turn off the
    O_PUBLIC bit to define password fields.</dd>

    <dt>O_EDIT</dt>

    <dd>Controls whether the field's data can be modified. When
    this option is off, all editing requests except
    <code>REQ_PREV_CHOICE</code> and <code>REQ_NEXT_CHOICE</code>
    will fail. Such read-only fields may be useful for help
    messages.</dd>

    <dt>O_WRAP</dt>

    <dd>Controls word-wrapping in multi-line fields. Normally, when
    any character of a (blank-separated) word reaches the end of
    the current line, the entire word is wrapped to the next line
    (assuming there is one). When this option is off, the word will
    be split across the line break.</dd>

    <dt>O_BLANK</dt>

    <dd>Controls field blanking. When this option is on, entering a
    character at the first field position erases the entire field
    (except for the just-entered character).</dd>

    <dt>O_AUTOSKIP</dt>

    <dd>Controls automatic skip to next field when this one fills.
    Normally, when the forms user tries to type more data into a
    field than will fit, the editing location jumps to next field.
    When this option is off, the user's cursor will hang at the end
    of the field. This option is ignored in dynamic fields that
    have not reached their size limit.</dd>

    <dt>O_NULLOK</dt>

    <dd>Controls whether <a href="#fvalidation">validation</a> is
    applied to blank fields. Normally, it is not; the user can
    leave a field blank without invoking the usual validation check
    on exit. If this option is off on a field, exit from it will
    invoke a validation check.</dd>

    <dt>O_PASSOK</dt>

    <dd>Controls whether validation occurs on every exit, or only
    after the field is modified. Normally the latter is true.
    Setting O_PASSOK may be useful if your field's validation
    function may change during forms processing.</dd>

    <dt>O_STATIC</dt>

    <dd>Controls whether the field is fixed to its initial
    dimensions. If you turn this off, the field becomes <a href=
    "#fdynamic">dynamic</a> and will stretch to fit entered
    data.</dd>
  </dl>

  <p>A field's options cannot be changed while the field is
  currently selected. However, options may be changed on posted
  fields that are not current.</p>

  <p>The option values are bit-masks and can be composed with
  logical-or in the obvious way.</p>

  <h2><a name="fstatus" id="fstatus">Field Status</a></h2>

  <p>Every field has a status flag, which is set to FALSE when the
  field is created and TRUE when the value in field buffer 0
  changes. This flag can be queried and set directly:</p>

  <pre>
int set_field_status(FIELD *field,      /* field to alter */
                   int status);         /* mode to set */

int field_status(FIELD *field);         /* fetch mode of field */
</pre>

  <p>Setting this flag under program control can be useful if you
  use the same form repeatedly, looking for modified fields each
  time.</p>

  <p>Calling <code>field_status()</code> on a field not currently
  selected for input will return a correct value. Calling
  <code>field_status()</code> on a field that is currently selected
  for input may not necessarily give a correct field status value,
  because entered data is not necessarily copied to buffer zero
  before the exit validation check. To guarantee that the returned
  status value reflects reality, call <code>field_status()</code>
  either (1) in the field's exit validation check routine, (2) from
  the field's or form's initialization or termination hooks, or (3)
  just after a <code>REQ_VALIDATION</code> request has been
  processed by the forms driver.</p>

  <h2><a name="fuser" id="fuser">Field User Pointer</a></h2>

  <p>Each field structure contains one character pointer slot that
  is not used by the forms library. It is intended to be used by
  applications to store private per-field data. You can manipulate
  it with:</p>

  <pre>
int set_field_userptr(FIELD *field,       /* field to alter */
                   char *userptr);        /* mode to set */

char *field_userptr(FIELD *field);        /* fetch mode of field */
</pre>(Properly, this user pointer field ought to have <code>(void
*)</code> type. The <code>(char *)</code> type is retained for
System V compatibility.)

  <p>It is valid to set the user pointer of the default field (with
  a <code>set_field_userptr()</code> call passed a NULL field
  pointer.) When a new field is created, the default-field user
  pointer is copied to initialize the new field's user pointer.</p>

  <h2><a name="fdynamic" id="fdynamic">Variable-Sized Fields</a></h2>

  <p>Normally, a field is fixed at the size specified for it at
  creation time. If, however, you turn off its O_STATIC bit, it
  becomes <dfn>dynamic</dfn> and will automatically resize itself
  to accommodate data as it is entered. If the field has extra
  buffers associated with it, they will grow right along with the
  main input buffer.</p>

  <p>A one-line dynamic field will have a fixed height (1) but
  variable width, scrolling horizontally to display data within the
  field area as originally dimensioned and located. A multi-line
  dynamic field will have a fixed width, but variable height
  (number of rows), scrolling vertically to display data within the
  field area as originally dimensioned and located.</p>

  <p>Normally, a dynamic field is allowed to grow without limit.
  But it is possible to set an upper limit on the size of a dynamic
  field. You do it with this function:</p>

  <pre>
int set_max_field(FIELD *field,     /* field to alter (may not be NULL) */
                   int max_size);   /* upper limit on field size */
</pre>

  <p>If the field is one-line, <code>max_size</code> is taken to be
  a column size limit; if it is multi-line, it is taken to be a
  line size limit. To disable any limit, use an argument of zero.
  The growth limit can be changed whether or not the O_STATIC bit
  is on, but has no effect until it is.</p>

  <p>The following properties of a field change when it becomes
  dynamic:</p>

  <ul>
    <li>If there is no growth limit, there is no final position of
    the field; therefore <code>O_AUTOSKIP</code> and
    <code>O_NL_OVERLOAD</code> are ignored.</li>

    <li>Field justification will be ignored (though whatever
    justification is set up will be retained internally and can be
    queried).</li>

    <li>The <code>dup_field()</code> and <code>link_field()</code>
    calls copy dynamic-buffer sizes. If the <code>O_STATIC</code>
    option is set on one of a collection of links, buffer resizing
    will occur only when the field is edited through that
    link.</li>

    <li>The call <code>field_info()</code> will retrieve the
    original static size of the field; use
    <code>dynamic_field_info()</code> to get the actual dynamic
    size.</li>
  </ul>

  <h2><a name="fvalidation" id="fvalidation">Field Validation</a></h2>

  <p>By default, a field will accept any data that will fit in its
  input buffer. However, it is possible to attach a validation type
  to a field. If you do this, any attempt to leave the field while
  it contains data that does not match the validation type will
  fail. Some validation types also have a character-validity check
  for each time a character is entered in the field.</p>

  <p>A field's validation check (if any) is not called when
  <code>set_field_buffer()</code> modifies the input buffer, nor
  when that buffer is changed through a linked field.</p>

  <p>The <code>form</code> library provides a rich set of
  pre-defined validation types, and gives you the capability to
  define custom ones of your own. You can examine and change field
  validation attributes with the following functions:</p>

  <pre>
int set_field_type(FIELD *field,          /* field to alter */
                   FIELDTYPE *ftype,      /* type to associate */
                   ...);                  /* additional arguments*/

FIELDTYPE *field_type(FIELD *field);      /* field to query */
</pre>

  <p>The validation type of a field is considered an attribute of
  the field. As with other field attributes, Also, doing
  <code>set_field_type()</code> with a <code>NULL</code> field
  default will change the system default for validation of
  newly-created fields.</p>

  <p>Here are the pre-defined validation types:</p>

  <h3><a name="ftype_alpha" id="ftype_alpha">TYPE_ALPHA</a></h3>

  <p>This field type accepts alphabetic data; no blanks, no digits,
  no special characters (this is checked at character-entry time).
  It is set up with:</p>

  <pre>
int set_field_type(FIELD *field,          /* field to alter */
                   TYPE_ALPHA,            /* type to associate */
                   int width);            /* maximum width of field */
</pre>

  <p>The <code>width</code> argument sets a minimum width of data.
  Typically you will want to set this to the field width; if it is
  greater than the field width, the validation check will always
  fail. A minimum width of zero makes field completion
  optional.</p>

  <h3><a name="ftype_alnum" id="ftype_alnum">TYPE_ALNUM</a></h3>

  <p>This field type accepts alphabetic data and digits; no blanks,
  no special characters (this is checked at character-entry time).
  It is set up with:</p>

  <pre>
int set_field_type(FIELD *field,          /* field to alter */
                   TYPE_ALNUM,            /* type to associate */
                   int width);            /* maximum width of field */
</pre>

  <p>The <code>width</code> argument sets a minimum width of data.
  As with TYPE_ALPHA, typically you will want to set this to the
  field width; if it is greater than the field width, the
  validation check will always fail. A minimum width of zero makes
  field completion optional.</p>

  <h3><a name="ftype_enum" id="ftype_enum">TYPE_ENUM</a></h3>

  <p>This type allows you to restrict a field's values to be among
  a specified set of string values (for example, the two-letter
  postal codes for U.S. states). It is set up with:</p>

  <pre>
int set_field_type(FIELD *field,          /* field to alter */
                   TYPE_ENUM,             /* type to associate */
                   char **valuelist;      /* list of possible values */
                   int checkcase;         /* case-sensitive? */
                   int checkunique);      /* must specify uniquely? */
</pre>

  <p>The <code>valuelist</code> parameter must point at a
  NULL-terminated list of valid strings. The <code>checkcase</code>
  argument, if true, makes comparison with the string
  case-sensitive.</p>

  <p>When the user exits a TYPE_ENUM field, the validation
  procedure tries to complete the data in the buffer to a valid
  entry. If a complete choice string has been entered, it is of
  course valid. But it is also possible to enter a prefix of a
  valid string and have it completed for you.</p>

  <p>By default, if you enter such a prefix and it matches more
  than one value in the string list, the prefix will be completed
  to the first matching value. But the <code>checkunique</code>
  argument, if true, requires prefix matches to be unique in order
  to be valid.</p>

  <p>The <code>REQ_NEXT_CHOICE</code> and
  <code>REQ_PREV_CHOICE</code> input requests can be particularly
  useful with these fields.</p>

  <h3><a name="ftype_integer" id="ftype_integer">TYPE_INTEGER</a></h3>

  <p>This field type accepts an integer. It is set up as
  follows:</p>

  <pre>
int set_field_type(FIELD *field,          /* field to alter */
                   TYPE_INTEGER,          /* type to associate */
                   int padding,           /* # places to zero-pad to */
                   int vmin, int vmax);   /* valid range */
</pre>

  <p>Valid characters consist of an optional leading minus and
  digits. The range check is performed on exit. If the range
  maximum is less than or equal to the minimum, the range is
  ignored.</p>

  <p>If the value passes its range check, it is padded with as many
  leading zero digits as necessary to meet the padding
  argument.</p>

  <p>A <code>TYPE_INTEGER</code> value buffer can conveniently be
  interpreted with the C library function <code>atoi(3)</code>.</p>

  <h3><a name="ftype_numeric" id="ftype_numeric">TYPE_NUMERIC</a></h3>

  <p>This field type accepts a decimal number. It is set up as
  follows:</p>

  <pre>
int set_field_type(FIELD *field,              /* field to alter */
                   TYPE_NUMERIC,              /* type to associate */
                   int padding,               /* # places of precision */
                   double vmin, double vmax); /* valid range */
</pre>

  <p>Valid characters consist of an optional leading minus and
  digits. possibly including a decimal point. If your system
  supports locale's, the decimal point character used must be the
  one defined by your locale. The range check is performed on exit.
  If the range maximum is less than or equal to the minimum, the
  range is ignored.</p>

  <p>If the value passes its range check, it is padded with as many
  trailing zero digits as necessary to meet the padding
  argument.</p>

  <p>A <code>TYPE_NUMERIC</code> value buffer can conveniently be
  interpreted with the C library function <code>atof(3)</code>.</p>

  <h3><a name="ftype_regexp" id="ftype_regexp">TYPE_REGEXP</a></h3>

  <p>This field type accepts data matching a regular expression. It
  is set up as follows:</p>

  <pre>
int set_field_type(FIELD *field,          /* field to alter */
                   TYPE_REGEXP,           /* type to associate */
                   char *regexp);         /* expression to match */
</pre>

  <p>The syntax for regular expressions is that of
  <code>regcomp(3)</code>. The check for regular-expression match
  is performed on exit.</p>

  <h2><a name="fbuffer" id="fbuffer">Direct Field Buffer
  Manipulation</a></h2>

  <p>The chief attribute of a field is its buffer contents. When a
  form has been completed, your application usually needs to know
  the state of each field buffer. You can find this out with:</p>

  <pre>
char *field_buffer(FIELD *field,          /* field to query */
                   int bufindex);         /* number of buffer to query */
</pre>

  <p>Normally, the state of the zero-numbered buffer for each field
  is set by the user's editing actions on that field. It is
  sometimes useful to be able to set the value of the zero-numbered
  (or some other) buffer from your application:</p>

  <pre>
int set_field_buffer(FIELD *field,        /* field to alter */
                   int bufindex,          /* number of buffer to alter */
                   char *value);          /* string value to set */
</pre>

  <p>If the field is not large enough and cannot be resized to a
  sufficiently large size to contain the specified value, the value
  will be truncated to fit.</p>

  <p>Calling <code>field_buffer()</code> with a null field pointer
  will raise an error. Calling <code>field_buffer()</code> on a
  field not currently selected for input will return a correct
  value. Calling <code>field_buffer()</code> on a field that is
  currently selected for input may not necessarily give a correct
  field buffer value, because entered data is not necessarily
  copied to buffer zero before the exit validation check. To
  guarantee that the returned buffer value reflects on-screen
  reality, call <code>field_buffer()</code> either (1) in the
  field's exit validation check routine, (2) from the field's or
  form's initialization or termination hooks, or (3) just after a
  <code>REQ_VALIDATION</code> request has been processed by the
  forms driver.</p>

  <h2><a name="formattrs" id="formattrs">Attributes of Forms</a></h2>

  <p>As with field attributes, form attributes inherit a default
  from a system default form structure. These defaults can be
  queried or set by of these functions using a form-pointer
  argument of <code>NULL</code>.</p>

  <p>The principal attribute of a form is its field list. You can
  query and change this list with:</p>

  <pre>
int set_form_fields(FORM *form,           /* form to alter */
                    FIELD **fields);      /* fields to connect */

char *form_fields(FORM *form);            /* fetch fields of form */

int field_count(FORM *form);              /* count connect fields */
</pre>

  <p>The second argument of <code>set_form_fields()</code> may be a
  NULL-terminated field pointer array like the one required by
  <code>new_form()</code>. In that case, the old fields of the form
  are disconnected but not freed (and eligible to be connected to
  other forms), then the new fields are connected.</p>

  <p>It may also be null, in which case the old fields are
  disconnected (and not freed) but no new ones are connected.</p>

  <p>The <code>field_count()</code> function simply counts the
  number of fields connected to a given from. It returns -1 if the
  form-pointer argument is NULL.</p>

  <h2><a name="fdisplay" id="fdisplay">Control of Form Display</a></h2>

  <p>In the overview section, you saw that to display a form you
  normally start by defining its size (and fields), posting it, and
  refreshing the screen. There is an hidden step before posting,
  which is the association of the form with a frame window
  (actually, a pair of windows) within which it will be displayed.
  By default, the forms library associates every form with the
  full-screen window <code>stdscr</code>.</p>

  <p>By making this step explicit, you can associate a form with a
  declared frame window on your screen display. This can be useful
  if you want to adapt the form display to different screen sizes,
  dynamically tile forms on the screen, or use a form as part of an
  interface layout managed by <a href="#panels">panels</a>.</p>

  <p>The two windows associated with each form have the same
  functions as their analogues in the <a href="#menu">menu
  library</a>. Both these windows are painted when the form is
  posted and erased when the form is unposted.</p>

  <p>The outer or frame window is not otherwise touched by the form
  routines. It exists so the programmer can associate a title, a
  border, or perhaps help text with the form and have it properly
  refreshed or erased at post/unpost time. The inner window or
  subwindow is where the current form page is actually
  displayed.</p>

  <p>In order to declare your own frame window for a form, you will
  need to know the size of the form's bounding rectangle. You can
  get this information with:</p>

  <pre>
int scale_form(FORM *form,                /* form to query */
               int *rows,                 /* form rows */
               int *cols);                /* form cols */
</pre>

  <p>The form dimensions are passed back in the locations pointed
  to by the arguments. Once you have this information, you can use
  it to declare of windows, then use one of these functions:</p>

  <pre>
int set_form_win(FORM *form,              /* form to alter */
                 WINDOW *win);            /* frame window to connect */

WINDOW *form_win(FORM *form);             /* fetch frame window of form */

int set_form_sub(FORM *form,              /* form to alter */
                 WINDOW *win);            /* form subwindow to connect */

WINDOW *form_sub(FORM *form);             /* fetch form subwindow of form */
</pre>

  <p>Note that curses operations, including <code>refresh()</code>,
  on the form, should be done on the frame window, not the form
  subwindow.</p>

  <p>It is possible to check from your application whether all of a
  scrollable field is actually displayed within the menu subwindow.
  Use these functions:</p>

  <pre>
int data_ahead(FORM *form);               /* form to be queried */

int data_behind(FORM *form);              /* form to be queried */
</pre>

  <p>The function <code>data_ahead()</code> returns TRUE if (a) the
  current field is one-line and has undisplayed data off to the
  right, (b) the current field is multi-line and there is data
  off-screen below it.</p>

  <p>The function <code>data_behind()</code> returns TRUE if the
  first (upper left hand) character position is off-screen (not
  being displayed).</p>

  <p>Finally, there is a function to restore the form window's
  cursor to the value expected by the forms driver:</p>

  <pre>
int pos_form_cursor(FORM *)               /* form to be queried */
</pre>

  <p>If your application changes the form window cursor, call this
  function before handing control back to the forms driver in order
  to re-synchronize it.</p>

  <h2><a name="fdriver" id="fdriver">Input Processing in the Forms
  Driver</a></h2>

  <p>The function <code>form_driver()</code> handles virtualized
  input requests for form navigation, editing, and validation
  requests, just as <code>menu_driver</code> does for menus (see
  the section on <a href="#minput">menu input handling</a>).</p>

  <pre>
int form_driver(FORM *form,               /* form to pass input to */
                int request);             /* form request code */
</pre>

  <p>Your input virtualization function needs to take input and
  then convert it to either an alphanumeric character (which is
  treated as data to be entered in the currently-selected field),
  or a forms processing request.</p>

  <p>The forms driver provides hooks (through input-validation and
  field-termination functions) with which your application code can
  check that the input taken by the driver matched what was
  expected.</p>

  <h3><a name="fpage" id="fpage">Page Navigation Requests</a></h3>

  <p>These requests cause page-level moves through the form,
  triggering display of a new form screen.</p>

  <dl>
    <dt><code>REQ_NEXT_PAGE</code>
    </dt>

    <dd>Move to the next form page.</dd>

    <dt><code>REQ_PREV_PAGE</code>
    </dt>

    <dd>Move to the previous form page.</dd>

    <dt><code>REQ_FIRST_PAGE</code>
    </dt>

    <dd>Move to the first form page.</dd>

    <dt><code>REQ_LAST_PAGE</code>
    </dt>

    <dd>Move to the last form page.</dd>
  </dl>

  <p>These requests treat the list as cyclic; that is,
  <code>REQ_NEXT_PAGE</code> from the last page goes to the first,
  and <code>REQ_PREV_PAGE</code> from the first page goes to the
  last.</p>

  <h3><a name="ffield" id="ffield">Inter-Field Navigation
  Requests</a></h3>

  <p>These requests handle navigation between fields on the same
  page.</p>

  <dl>
    <dt><code>REQ_NEXT_FIELD</code>
    </dt>

    <dd>Move to next field.</dd>

    <dt><code>REQ_PREV_FIELD</code>
    </dt>

    <dd>Move to previous field.</dd>

    <dt><code>REQ_FIRST_FIELD</code>
    </dt>

    <dd>Move to the first field.</dd>

    <dt><code>REQ_LAST_FIELD</code>
    </dt>

    <dd>Move to the last field.</dd>

    <dt><code>REQ_SNEXT_FIELD</code>
    </dt>

    <dd>Move to sorted next field.</dd>

    <dt><code>REQ_SPREV_FIELD</code>
    </dt>

    <dd>Move to sorted previous field.</dd>

    <dt><code>REQ_SFIRST_FIELD</code>
    </dt>

    <dd>Move to the sorted first field.</dd>

    <dt><code>REQ_SLAST_FIELD</code>
    </dt>

    <dd>Move to the sorted last field.</dd>

    <dt><code>REQ_LEFT_FIELD</code>
    </dt>

    <dd>Move left to field.</dd>

    <dt><code>REQ_RIGHT_FIELD</code>
    </dt>

    <dd>Move right to field.</dd>

    <dt><code>REQ_UP_FIELD</code>
    </dt>

    <dd>Move up to field.</dd>

    <dt><code>REQ_DOWN_FIELD</code>
    </dt>

    <dd>Move down to field.</dd>
  </dl>

  <p>These requests treat the list of fields on a page as cyclic;
  that is, <code>REQ_NEXT_FIELD</code> from the last field goes to
  the first, and <code>REQ_PREV_FIELD</code> from the first field
  goes to the last. The order of the fields for these (and the
  <code>REQ_FIRST_FIELD</code> and <code>REQ_LAST_FIELD</code>
  requests) is simply the order of the field pointers in the form
  array (as set up by <code>new_form()</code> or
  <code>set_form_fields()</code></p>

  <p>It is also possible to traverse the fields as if they had been
  sorted in screen-position order, so the sequence goes
  left-to-right and top-to-bottom. To do this, use the second group
  of four sorted-movement requests.</p>

  <p>Finally, it is possible to move between fields using visual
  directions up, down, right, and left. To accomplish this, use the
  third group of four requests. Note, however, that the position of
  a form for purposes of these requests is its upper-left
  corner.</p>

  <p>For example, suppose you have a multi-line field B, and two
  single-line fields A and C on the same line with B, with A to the
  left of B and C to the right of B. A <code>REQ_MOVE_RIGHT</code>
  from A will go to B only if A, B, and C <em>all</em> share the
  same first line; otherwise it will skip over B to C.</p>

  <h3><a name="fifield" id="fifield">Intra-Field Navigation
  Requests</a></h3>

  <p>These requests drive movement of the edit cursor within the
  currently selected field.</p>

  <dl>
    <dt><code>REQ_NEXT_CHAR</code>
    </dt>

    <dd>Move to next character.</dd>

    <dt><code>REQ_PREV_CHAR</code>
    </dt>

    <dd>Move to previous character.</dd>

    <dt><code>REQ_NEXT_LINE</code>
    </dt>

    <dd>Move to next line.</dd>

    <dt><code>REQ_PREV_LINE</code>
    </dt>

    <dd>Move to previous line.</dd>

    <dt><code>REQ_NEXT_WORD</code>
    </dt>

    <dd>Move to next word.</dd>

    <dt><code>REQ_PREV_WORD</code>
    </dt>

    <dd>Move to previous word.</dd>

    <dt><code>REQ_BEG_FIELD</code>
    </dt>

    <dd>Move to beginning of field.</dd>

    <dt><code>REQ_END_FIELD</code>
    </dt>

    <dd>Move to end of field.</dd>

    <dt><code>REQ_BEG_LINE</code>
    </dt>

    <dd>Move to beginning of line.</dd>

    <dt><code>REQ_END_LINE</code>
    </dt>

    <dd>Move to end of line.</dd>

    <dt><code>REQ_LEFT_CHAR</code>
    </dt>

    <dd>Move left in field.</dd>

    <dt><code>REQ_RIGHT_CHAR</code>
    </dt>

    <dd>Move right in field.</dd>

    <dt><code>REQ_UP_CHAR</code>
    </dt>

    <dd>Move up in field.</dd>

    <dt><code>REQ_DOWN_CHAR</code>
    </dt>

    <dd>Move down in field.</dd>
  </dl>

  <p>Each <em>word</em> is separated from the previous and next
  characters by whitespace. The commands to move to beginning and
  end of line or field look for the first or last non-pad character
  in their ranges.</p>

  <h3><a name="fscroll" id="fscroll">Scrolling Requests</a></h3>

  <p>Fields that are dynamic and have grown and fields explicitly
  created with offscreen rows are scrollable. One-line fields
  scroll horizontally; multi-line fields scroll vertically. Most
  scrolling is triggered by editing and intra-field movement (the
  library scrolls the field to keep the cursor visible). It is
  possible to explicitly request scrolling with the following
  requests:</p>

  <dl>
    <dt><code>REQ_SCR_FLINE</code>
    </dt>

    <dd>Scroll vertically forward a line.</dd>

    <dt><code>REQ_SCR_BLINE</code>
    </dt>

    <dd>Scroll vertically backward a line.</dd>

    <dt><code>REQ_SCR_FPAGE</code>
    </dt>

    <dd>Scroll vertically forward a page.</dd>

    <dt><code>REQ_SCR_BPAGE</code>
    </dt>

    <dd>Scroll vertically backward a page.</dd>

    <dt><code>REQ_SCR_FHPAGE</code>
    </dt>

    <dd>Scroll vertically forward half a page.</dd>

    <dt><code>REQ_SCR_BHPAGE</code>
    </dt>

    <dd>Scroll vertically backward half a page.</dd>

    <dt><code>REQ_SCR_FCHAR</code>
    </dt>

    <dd>Scroll horizontally forward a character.</dd>

    <dt><code>REQ_SCR_BCHAR</code>
    </dt>

    <dd>Scroll horizontally backward a character.</dd>

    <dt><code>REQ_SCR_HFLINE</code>
    </dt>

    <dd>Scroll horizontally one field width forward.</dd>

    <dt><code>REQ_SCR_HBLINE</code>
    </dt>

    <dd>Scroll horizontally one field width backward.</dd>

    <dt><code>REQ_SCR_HFHALF</code>
    </dt>

    <dd>Scroll horizontally one half field width forward.</dd>

    <dt><code>REQ_SCR_HBHALF</code>
    </dt>

    <dd>Scroll horizontally one half field width backward.</dd>
  </dl>

  <p>For scrolling purposes, a <em>page</em> of a field is the
  height of its visible part.</p>

  <h3><a name="fedit" id="fedit">Editing Requests</a></h3>

  <p>When you pass the forms driver an ASCII character, it is
  treated as a request to add the character to the field's data
  buffer. Whether this is an insertion or a replacement depends on
  the field's edit mode (insertion is the default.</p>

  <p>The following requests support editing the field and changing
  the edit mode:</p>

  <dl>
    <dt><code>REQ_INS_MODE</code>
    </dt>

    <dd>Set insertion mode.</dd>

    <dt><code>REQ_OVL_MODE</code>
    </dt>

    <dd>Set overlay mode.</dd>

    <dt><code>REQ_NEW_LINE</code>
    </dt>

    <dd>New line request (see below for explanation).</dd>

    <dt><code>REQ_INS_CHAR</code>
    </dt>

    <dd>Insert space at character location.</dd>

    <dt><code>REQ_INS_LINE</code>
    </dt>

    <dd>Insert blank line at character location.</dd>

    <dt><code>REQ_DEL_CHAR</code>
    </dt>

    <dd>Delete character at cursor.</dd>

    <dt><code>REQ_DEL_PREV</code>
    </dt>

    <dd>Delete previous word at cursor.</dd>

    <dt><code>REQ_DEL_LINE</code>
    </dt>

    <dd>Delete line at cursor.</dd>

    <dt><code>REQ_DEL_WORD</code>
    </dt>

    <dd>Delete word at cursor.</dd>

    <dt><code>REQ_CLR_EOL</code>
    </dt>

    <dd>Clear to end of line.</dd>

    <dt><code>REQ_CLR_EOF</code>
    </dt>

    <dd>Clear to end of field.</dd>

    <dt><code>REQ_CLEAR_FIELD</code>
    </dt>

    <dd>Clear entire field.</dd>
  </dl>

  <p>The behavior of the <code>REQ_NEW_LINE</code> and
  <code>REQ_DEL_PREV</code> requests is complicated and partly
  controlled by a pair of forms options. The special cases are
  triggered when the cursor is at the beginning of a field, or on
  the last line of the field.</p>

  <p>First, we consider <code>REQ_NEW_LINE</code>:</p>

  <p>The normal behavior of <code>REQ_NEW_LINE</code> in insert
  mode is to break the current line at the position of the edit
  cursor, inserting the portion of the current line after the
  cursor as a new line following the current and moving the cursor
  to the beginning of that new line (you may think of this as
  inserting a newline in the field buffer).</p>

  <p>The normal behavior of <code>REQ_NEW_LINE</code> in overlay
  mode is to clear the current line from the position of the edit
  cursor to end of line. The cursor is then moved to the beginning
  of the next line.</p>

  <p>However, <code>REQ_NEW_LINE</code> at the beginning of a
  field, or on the last line of a field, instead does a
  <code>REQ_NEXT_FIELD</code>. <code>O_NL_OVERLOAD</code> option is
  off, this special action is disabled.</p>

  <p>Now, let us consider <code>REQ_DEL_PREV</code>:</p>

  <p>The normal behavior of <code>REQ_DEL_PREV</code> is to delete
  the previous character. If insert mode is on, and the cursor is
  at the start of a line, and the text on that line will fit on the
  previous one, it instead appends the contents of the current line
  to the previous one and deletes the current line (you may think
  of this as deleting a newline from the field buffer).</p>

  <p>However, <code>REQ_DEL_PREV</code> at the beginning of a field
  is instead treated as a <code>REQ_PREV_FIELD</code>.</p>

  <p>If the <code>O_BS_OVERLOAD</code> option is off, this special
  action is disabled and the forms driver just returns
  <code>E_REQUEST_DENIED</code>.</p>

  <p>See <a href="#frmoptions">Form Options</a> for discussion of
  how to set and clear the overload options.</p>

  <h3><a name="forder" id="forder">Order Requests</a></h3>

  <p>If the type of your field is ordered, and has associated
  functions for getting the next and previous values of the type
  from a given value, there are requests that can fetch that value
  into the field buffer:</p>

  <dl>
    <dt><code>REQ_NEXT_CHOICE</code>
    </dt>

    <dd>Place the successor value of the current value in the
    buffer.</dd>

    <dt><code>REQ_PREV_CHOICE</code>
    </dt>

    <dd>Place the predecessor value of the current value in the
    buffer.</dd>
  </dl>

  <p>Of the built-in field types, only <code>TYPE_ENUM</code> has
  built-in successor and predecessor functions. When you define a
  field type of your own (see <a href="#fcustom">Custom Validation
  Types</a>), you can associate our own ordering functions.</p>

  <h3><a name="fappcmds" id="fappcmds">Application Commands</a></h3>

  <p>Form requests are represented as integers above the
  <code>curses</code> value greater than <code>KEY_MAX</code> and
  less than or equal to the constant <code>MAX_COMMAND</code>. If
  your input-virtualization routine returns a value above
  <code>MAX_COMMAND</code>, the forms driver will ignore it.</p>

  <h2><a name="fhooks" id="fhooks">Field Change Hooks</a></h2>

  <p>It is possible to set function hooks to be executed whenever
  the current field or form changes. Here are the functions that
  support this:</p>

  <pre>
typedef void    (*HOOK)();       /* pointer to function returning void */

int set_form_init(FORM *form,    /* form to alter */
                  HOOK hook);    /* initialization hook */

HOOK form_init(FORM *form);      /* form to query */

int set_form_term(FORM *form,    /* form to alter */
                  HOOK hook);    /* termination hook */

HOOK form_term(FORM *form);      /* form to query */

int set_field_init(FORM *form,   /* form to alter */
                  HOOK hook);    /* initialization hook */

HOOK field_init(FORM *form);     /* form to query */

int set_field_term(FORM *form,   /* form to alter */
                  HOOK hook);    /* termination hook */

HOOK field_term(FORM *form);     /* form to query */
</pre>

  <p>These functions allow you to either set or query four
  different hooks. In each of the set functions, the second
  argument should be the address of a hook function. These
  functions differ only in the timing of the hook call.</p>

  <dl>
    <dt>form_init</dt>

    <dd>This hook is called when the form is posted; also, just
    after each page change operation.</dd>

    <dt>field_init</dt>

    <dd>This hook is called when the form is posted; also, just
    after each field change</dd>

    <dt>field_term</dt>

    <dd>This hook is called just after field validation; that is,
    just before the field is altered. It is also called when the
    form is unposted.</dd>

    <dt>form_term</dt>

    <dd>This hook is called when the form is unposted; also, just
    before each page change operation.</dd>
  </dl>

  <p>Calls to these hooks may be triggered</p>

  <ol>
    <li>When user editing requests are processed by the forms
    driver</li>

    <li>When the current page is changed by
    <code>set_current_field()</code> call</li>

    <li>When the current field is changed by a
    <code>set_form_page()</code> call</li>
  </ol>

  <p>See <a name="ffocus" id="ffocus">Field Change Commands</a> for
  discussion of the latter two cases.</p>

  <p>You can set a default hook for all fields by passing one of
  the set functions a NULL first argument.</p>

  <p>You can disable any of these hooks by (re)setting them to
  NULL, the default value.</p>

  <h2><a href="#ffocus">Field Change Commands</a></h2>

  <p>Normally, navigation through the form will be driven by the
  user's input requests. But sometimes it is useful to be able to
  move the focus for editing and viewing under control of your
  application, or ask which field it currently is in. The following
  functions help you accomplish this:</p>

  <pre>
int set_current_field(FORM *form,         /* form to alter */
                      FIELD *field);      /* field to shift to */

FIELD *current_field(FORM *form);         /* form to query */

int field_index(FORM *form,               /* form to query */
                FIELD *field);            /* field to get index of */
</pre>

  <p>The function <code>field_index()</code> returns the index of
  the given field in the given form's field array (the array passed
  to <code>new_form()</code> or
  <code>set_form_fields()</code>).</p>

  <p>The initial current field of a form is the first active field
  on the first page. The function <code>set_form_fields()</code>
  resets this.</p>

  <p>It is also possible to move around by pages.</p>

  <pre>
int set_form_page(FORM *form,             /* form to alter */
                  int page);              /* page to go to (0-origin) */

int form_page(FORM *form);                /* return form's current page */
</pre>

  <p>The initial page of a newly-created form is 0. The function
  <code>set_form_fields()</code> resets this.</p>

  <h2><a name="frmoptions" id="frmoptions">Form Options</a></h2>

  <p>Like fields, forms may have control option bits. They can be
  changed or queried with these functions:</p>

  <pre>
int set_form_opts(FORM *form,             /* form to alter */
                  int attr);              /* attribute to set */

int form_opts_on(FORM *form,              /* form to alter */
                 int attr);               /* attributes to turn on */

int form_opts_off(FORM *form,             /* form to alter */
                  int attr);              /* attributes to turn off */

int form_opts(FORM *form);                /* form to query */
</pre>

  <p>By default, all options are on. Here are the available option
  bits:</p>

  <dl>
    <dt>O_NL_OVERLOAD</dt>

    <dd>Enable overloading of <code>REQ_NEW_LINE</code> as
    described in <a href="#fedit">Editing Requests</a>. The value
    of this option is ignored on dynamic fields that have not
    reached their size limit; these have no last line, so the
    circumstances for triggering a <code>REQ_NEXT_FIELD</code>
    never arise.</dd>

    <dt>O_BS_OVERLOAD</dt>

    <dd>Enable overloading of <code>REQ_DEL_PREV</code> as
    described in <a href="#fedit">Editing Requests</a>.</dd>
  </dl>

  <p>The option values are bit-masks and can be composed with
  logical-or in the obvious way.</p>

  <h2><a name="fcustom" id="fcustom">Custom Validation Types</a></h2>

  <p>The <code>form</code> library gives you the capability to
  define custom validation types of your own. Further, the optional
  additional arguments of <code>set_field_type</code> effectively
  allow you to parameterize validation types. Most of the
  complications in the validation-type interface have to do with
  the handling of the additional arguments within custom validation
  functions.</p>

  <h3><a name="flinktypes" id="flinktypes">Union Types</a></h3>

  <p>The simplest way to create a custom data type is to compose it
  from two preexisting ones:</p>

  <pre>
FIELD *link_fieldtype(FIELDTYPE *type1,
                      FIELDTYPE *type2);
</pre>

  <p>This function creates a field type that will accept any of the
  values legal for either of its argument field types (which may be
  either predefined or programmer-defined). If a
  <code>set_field_type()</code> call later requires arguments, the
  new composite type expects all arguments for the first type, than
  all arguments for the second. Order functions (see <a href=
  "#forder">Order Requests</a>) associated with the component types
  will work on the composite; what it does is check the validation
  function for the first type, then for the second, to figure what
  type the buffer contents should be treated as.</p>

  <h3><a name="fnewtypes" id="fnewtypes">New Field Types</a></h3>

  <p>To create a field type from scratch, you need to specify one
  or both of the following things:</p>

  <ul>
    <li>A character-validation function, to check each character as
    it is entered.</li>

    <li>A field-validation function to be applied on exit from the
    field.</li>
  </ul>

  <p>Here is how you do that:</p>

  <pre>
typedef int     (*HOOK)();       /* pointer to function returning int */

FIELDTYPE *new_fieldtype(HOOK f_validate, /* field validator */
                         HOOK c_validate) /* character validator */

int free_fieldtype(FIELDTYPE *ftype);     /* type to free */
</pre>

  <p>At least one of the arguments of <code>new_fieldtype()</code>
  must be non-NULL. The forms driver will automatically call the
  new type's validation functions at appropriate points in
  processing a field of the new type.</p>

  <p>The function <code>free_fieldtype()</code> deallocates the
  argument fieldtype, freeing all storage associated with it.</p>

  <p>Normally, a field validator is called when the user attempts
  to leave the field. Its first argument is a field pointer, from
  which it can get to field buffer 0 and test it. If the function
  returns TRUE, the operation succeeds; if it returns FALSE, the
  edit cursor stays in the field.</p>

  <p>A character validator gets the character passed in as a first
  argument. It too should return TRUE if the character is valid,
  FALSE otherwise.</p>

  <h3><a name="fcheckargs" id="fcheckargs">Validation Function
  Arguments</a></h3>

  <p>Your field- and character- validation functions will be passed
  a second argument as well. This second argument is the address of
  a structure (which we will call a <em>pile</em>) built from any
  of the field-type-specific arguments passed to
  <code>set_field_type()</code>. If no such arguments are defined
  for the field type, this pile pointer argument will be NULL.</p>

  <p>In order to arrange for such arguments to be passed to your
  validation functions, you must associate a small set of
  storage-management functions with the type. The forms driver will
  use these to synthesize a pile from the trailing arguments of
  each <code>set_field_type()</code> argument, and a pointer to the
  pile will be passed to the validation functions.</p>

  <p>Here is how you make the association:</p>

  <pre>
typedef char    *(*PTRHOOK)();    /* pointer to function returning (char *) */
typedef void    (*VOIDHOOK)();    /* pointer to function returning void */

int set_fieldtype_arg(FIELDTYPE *type,    /* type to alter */
                      PTRHOOK make_str,   /* make structure from args */
                      PTRHOOK copy_str,   /* make copy of structure */
                      VOIDHOOK free_str); /* free structure storage */
</pre>

  <p>Here is how the storage-management hooks are used:</p>

  <dl>
    <dt><code>make_str</code>
    </dt>

    <dd>This function is called by <code>set_field_type()</code>.
    It gets one argument, a <code>va_list</code> of the
    type-specific arguments passed to
    <code>set_field_type()</code>. It is expected to return a pile
    pointer to a data structure that encapsulates those
    arguments.</dd>

    <dt><code>copy_str</code>
    </dt>

    <dd>This function is called by form library functions that
    allocate new field instances. It is expected to take a pile
    pointer, copy the pile to allocated storage, and return the
    address of the pile copy.</dd>

    <dt><code>free_str</code>
    </dt>

    <dd>This function is called by field- and type-deallocation
    routines in the library. It takes a pile pointer argument, and
    is expected to free the storage of that pile.</dd>
  </dl>

  <p>The <code>make_str</code> and <code>copy_str</code> functions
  may return NULL to signal allocation failure. The library
  routines will that call them will return error indication when
  this happens. Thus, your validation functions should never see a
  NULL file pointer and need not check specially for it.</p>

  <h3><a name="fcustorder" id="fcustorder">Order Functions For
  Custom Types</a></h3>

  <p>Some custom field types are simply ordered in the same
  well-defined way that <code>TYPE_ENUM</code> is. For such types,
  it is possible to define successor and predecessor functions to
  support the <code>REQ_NEXT_CHOICE</code> and
  <code>REQ_PREV_CHOICE</code> requests. Here is how:</p>

  <pre>
typedef int     (*INTHOOK)();     /* pointer to function returning int */

int set_fieldtype_arg(FIELDTYPE *type,    /* type to alter */
                      INTHOOK succ,       /* get successor value */
                      INTHOOK pred);      /* get predecessor value */
</pre>

  <p>The successor and predecessor arguments will each be passed
  two arguments; a field pointer, and a pile pointer (as for the
  validation functions). They are expected to use the function
  <code>field_buffer()</code> to read the current value, and
  <code>set_field_buffer()</code> on buffer 0 to set the next or
  previous value. Either hook may return TRUE to indicate success
  (a legal next or previous value was set) or FALSE to indicate
  failure.</p>

  <h3><a name="fcustprobs" id="fcustprobs">Avoiding Problems</a></h3>

  <p>The interface for defining custom types is complicated and
  tricky. Rather than attempting to create a custom type entirely
  from scratch, you should start by studying the library source
  code for whichever of the pre-defined types seems to be closest
  to what you want.</p>

  <p>Use that code as a model, and evolve it towards what you
  really want. You will avoid many problems and annoyances that
  way. The code in the <code>ncurses</code> library has been
  specifically exempted from the package copyright to support
  this.</p>

  <p>If your custom type defines order functions, have do something
  intuitive with a blank field. A useful convention is to make the
  successor of a blank field the types minimum value, and its
  predecessor the maximum.</p>
</body>
</html>