int add_wch( const cchar_t *wch );
int wadd_wch( WINDOW *win, const cchar_t *wch );
int mvadd_wch( int y, int x, const cchar_t *wch );
int mvwadd_wch( WINDOW *win, int y, int x, const cchar_t *wch );
int echo_wchar( const cchar_t *wch );
int wecho_wchar( WINDOW *win, const cchar_t *wch );
functions put the complex character wch into the given
window at its current position,
which is then advanced.
These functions perform
wrapping and special-character processing as follows:
.IP • 4 If wch refers to a spacing character, then any previous character at that location is removed. A new character specified by wch is placed at that location with rendition specified by wch. The cursor then advances to the next spacing character on the screen.
.IP • 4 If wch refers to a non-spacing character, all previous characters at that location are preserved. The non-spacing characters of wch are added to the spacing complex character, and the rendition specified by wch is ignored.
.IP • 4 If the character part of wch is a tab, newline, backspace or other control character, the window is updated and the cursor moves as if addch were called.
The echo_wchar function is functionally equivalent to a call to add_wch followed by a call to refresh(3X). Similarly, the wecho_wchar is functionally equivalent to a call to wadd_wch followed by a call to wrefresh. The knowledge that only a single character is being output is taken into consideration and, for non-control characters, a considerable performance gain might be seen by using the *echo* functions instead of their equivalents.
|board of squares
|checker board (stipple)
|arrow pointing down
|arrow pointing left
|lower left-hand corner
|lower right-hand corner
|arrow pointing right
|scan line 1
|scan line 3
|scan line 7
|scan line 9
|arrow pointing up
|upper left-hand corner
|upper right-hand corner
The wide-character configuration of ncurses also defines symbols for thick lines (acsc ``J'' to ``V''):
|thick horizontal line
|thick lower left corner
|thick lower right corner
|thick tee pointing right
|thick large plus
|thick tee pointing left
|thick tee pointing down
|thick upper left corner
|thick upper right corner
|thick vertical line
and for double-lines (acsc ``A'' to ``I''):
|double horizontal line
|double lower left corner
|double lower right corner
|double tee pointing right
|double large plus
|double tee pointing left
|double tee pointing down
|double upper left corner
|double upper right corner
|double vertical line
Unicode's descriptions for these characters differs slightly from ncurses,
by introducing the term ``light'' (along with less important details).
Here are its descriptions for the normal, thick, and double horizontal lines:
.IP • 4 U+2500 BOX DRAWINGS LIGHT HORIZONTAL
.IP • 4 U+2501 BOX DRAWINGS HEAVY HORIZONTAL
.IP • 4 U+2550 BOX DRAWINGS DOUBLE HORIZONTAL
All routines return the integer ERR upon failure and OK on success.
All of these functions are described in the XSI Curses standard, Issue 4. The defaults specified for line-drawing characters apply in the POSIX locale.
X/Open Curses makes it clear that the WACS_ symbols should be defined as
a pointer to cchar_t data, e.g., in the discussion of border_set.
A few implementations are problematic:
.IP • 4 NetBSD curses defines the symbols as a wchar_t within a cchar_t.
.IP • 4 HPUX curses equates some of the ACS_ symbols to the analogous WACS_ symbols as if the ACS_ symbols were wide characters. The misdefined symbols are the arrows and other symbols which are not used for line-drawing.
X/Open Curses does not define symbols for thick- or double-lines. SVr4 curses implementations defined their line-drawing symbols in terms of intermediate symbols. This implementation extends those symbols, providing new definitions which are not in the SVr4 implementations.
Not all Unicode-capable terminals provide support for VT100-style alternate character sets (i.e., the acsc capability), with their corresponding line-drawing characters. X/Open Curses did not address the aspect of integrating Unicode with line-drawing characters. Existing implementations of Unix curses (AIX, HPUX, Solaris) use only the acsc character-mapping to provide this feature. As a result, those implementations can only use single-byte line-drawing characters. Ncurses 5.3 (2002) provided a table of Unicode values to solve these problems. NetBSD curses incorporated that table in 2010.
In this implementation, the Unicode values are used instead of the terminal description's acsc mapping as discussed in ncurses(3X) for the environment variable NCURSES_NO_UTF8_ACS. In contrast, for the same cases, the line-drawing characters described in curs_addch(3X) will use only the ASCII default values.
Having Unicode available does not solve all of the problems with
line-drawing for curses:
.IP • 4 The closest Unicode equivalents to the VT100 graphics S1, S3, S7 and S9 frequently are not displayed at the regular intervals which the terminal used.
.IP • 4 The lantern is a special case. It originated with the AT&T 4410 terminal in the early 1980s. There is no accessible documentation depicting the lantern symbol on the AT&T terminal.