NAME
gl_get_line, new_GetLine, del_GetLine,
gl_customize_completion, gl_change_terminal - allow the user
to compose an input line
SYNOPSIS
#include <stdio.h>
#include <libtecla.h>
GetLine *new_GetLine(size_t linelen, size_t histlen);
GetLine *del_GetLine(GetLine *gl);
char *gl_get_line(GetLine *gl, const char *prompt,
const char *start_line, int start_pos);
int gl_customize_completion(GetLine *gl, void *data,
CplMatchFn *match_fn);
int gl_change_terminal(GetLine *gl, FILE *input_fp,
FILE *output_fp, const char *term);
DESCRIPTION
The gl_get_line() function is part of the tecla library (see
the libtecla(3) man page). If the user is typing at a termi-
nal, it prompts them for an line of input, then provides
interactive editing facilities, similar to those of the unix
tcsh shell. In addition to simple command-line editing, it
supports recall of previously entered command lines, TAB
completion of file names, and in-line wild-card expansion of
filenames.
AN EXAMPLE
The following shows a complete example of how to use the
gl_get_line() function to get input from the user:
#include <stdio.h>
#include <libtecla.h>
int main(int argc, char *argv[])
{
char *line; /* The line that the user typed */
Getline *gl; /* The gl_get_line() resource object */
gl = new_GetLine(1024, 2048);
if(!gl)
return 1;
while((line=gl_get_line(gl, "$ ", NULL, -1)) != NULL &&
strcmp(line, "exit\n") != 0)
printf("You typed: %s\n", line);
gl = del_GetLine(gl);
return 0;
}
In the example, first the resources needed by the
gl_get_line() function are created by calling new_GetLine().
This allocates the memory used in subsequent calls to the
gl_get_line() function, including the history buffer for
recording previously entered lines. Then one or more lines
are read from the user, until either an error occurs, or the
user types exit. Then finally the resources that were allo-
cated by new_GetLine(), are returned to the system by cal-
ling del_GetLine(). Note the use of the NULL return value of
del_GetLine() to make gl NULL. This is a safety precaution.
If the program subsequently attempts to pass gl to
gl_get_line(), said function will complain, and return an
error, instead of attempting to use the deleted resource
object.
THE FUNCTIONS USED IN THE EXAMPLE
The descriptions of the functions used in the example are as
follows:
GetLine *new_GetLine(size_t linelen, size_t histlen)
This function creates the resources used by the
gl_get_line() function and returns an opaque pointer to the
object that contains them. The maximum length of an input
line is specified via the linelen argument, and the number
of bytes to allocate for storing history lines is set by the
histlen argument. History lines are stored back-to-back in a
single buffer of this size. On error, a message is printed
to stderr and NULL is returned.
GetLine *del_GetLine(GetLine *gl)
This function deletes the resources that were returned by a
previous call to new_GetLine(). It always returns NULL (ie a
deleted object). It does nothing if the gl argument is NULL.
char *gl_get_line(GetLine *gl, const char *prompt,
const char *start_line, int start_pos);
The gl_get_line() function can be called any number of times
to read input from the user. The gl argument must have been
previously returned by a call to new_GetLine(). The prompt
argument should be a normal NUL terminated string,
specifying the prompt to present the user with. If you want
to specify the initial contents of the line, for the user to
edit, pass the desired string via the start_line argument.
You can then specify which character of this line the cursor
is initially positioned over, using the start_pos argument.
This should be -1 if you want the cursor to follow the last
character of the start line. If you don't want to preload
the line in this manner, send start_line as NULL, and set
start_pos to -1.
The gl_get_line() function returns a pointer to the line
entered by the user, or NULL on error or at the end of the
input. The returned pointer is part of the specified gl
resource object, and thus should not be free'd by the
caller, or assumed to be unchanging from one call to the
next. When reading from a user at a terminal, there will
always be a newline character at the end of the returned
line. When standard input is being taken from a pipe or a
file, there will similarly be a newline unless the input
line was too long to store in the internal buffer. In the
latter case you should call gl_get_line() again to read the
rest of the line. Note that this behavior makes
gl_get_line() similar to fgets(). In fact when stdin isn't
connected to a terminal,gl_get_line() just calls fgets().
THE AVAILABLE KEY BINDING FUNCTIONS
The gl_get_line() function provides a number of functions
which can be bound to key sequences. The names of these
functions, and what they do, are given below.
user-interrupt - Send a SIGINT signal to the
parent process.
abort - Send a SIGABRT signal to the
parent process.
suspend - Suspend the parent process.
stop-output - Pause terminal output.
start-output - Resume paused terminal output.
literal-next - Arrange for the next character
to be treated as a normal
character. This allows control
characters to be entered.
cursor-right - Move the cursor one character
right.
cursor-left - Move the cursor one character
left.
insert-mode - Toggle between insert mode and
overwrite mode.
beginning-of-line - Move the cursor to the
beginning of the line.
end-of-line - Move the cursor to the end of
the line.
delete-line - Delete the contents of the
current line.
kill-line - Delete everything that follows
the cursor.
backward-kill-line - Delete all characters between
the cursor and the start of the
line.
forward-word - Move to the end of the word
which follows the cursor.
forward-to-word - Move the cursor to the start of
the word that follows the
cursor.
backward-word - Move to the start of the word
which precedes the cursor.
goto-column - Move the cursor to the
1-relative column in the line
specified by any preceding
digit-argument sequences (see
ENTERING REPEAT COUNTS below).
find-parenthesis - If the cursor is currently
over a parenthesis character,
move it to the matching
parenthesis character. If not
over a parenthesis character
move right to the next close
parenthesis.
forward-delete-char - Delete the character under the
cursor.
backward-delete-char - Delete the character which
precedes the cursor.
list-or-eof - This is intended for binding
to ^D. When invoked when the
cursor is within the line it
displays all possible
completions then redisplays
the line unchanged. When
invoked on an empty line, it
signals end-of-input (EOF) to
the caller of gl_get_line().
del-char-or-list-or-eof - This is intended for binding
to ^D. When invoked when the
cursor is within the line it
invokes forward-delete-char.
When invoked at the end of the
line it displays all possible
completions then redisplays
the line unchanged. When
invoked on an empty line, it
signals end-of-input (EOF) to
the caller of gl_get_line().
forward-delete-word - Delete the word which follows
the cursor.
backward-delete-word - Delete the word which precedes
the cursor.
upcase-word - Convert all of the characters
of the word which follows the
cursor, to upper case.
downcase-word - Convert all of the characters
of the word which follows the
cursor, to lower case.
capitalize-word - Capitalize the word which
follows the cursor.
change-case - If the next character is upper
case, toggle it to lower case
and vice versa.
redisplay - Redisplay the line.
clear-screen - Clear the terminal, then
redisplay the current line.
transpose-chars - Swap the character under the
cursor with the character just
before the cursor.
set-mark - Set a mark at the position of
the cursor.
exchange-point-and-mark - Move the cursor to the last
mark that was set, and move
the mark to where the cursor
used to be.
kill-region - Delete the characters that lie
between the last mark that was
set, and the cursor.
copy-region-as-kill - Copy the text between the mark
and the cursor to the cut
buffer, without deleting the
original text.
yank - Insert the text that was last
deleted, just before the
current position of the cursor.
append-yank - Paste the current contents of
the cut buffer, after the
cursor.
up-history - Recall the next oldest line
that was entered. Note that
in vi mode you are left in
command mode.
down-history - Recall the next most recent
line that was entered. If no
history recall session is
currently active, the next
line from a previous recall
session is recalled. Note that
in vi mode you are left in
command mode.
history-search-backward - Recall the next oldest line
who's prefix matches the string
which currently precedes the
cursor (in vi command-mode the
character under the cursor is
also included in the search
string). Note that in vi mode
you are left in command mode.
history-search-forward - Recall the next newest line
who's prefix matches the string
which currently precedes the
cursor (in vi command-mode the
character under the cursor is
also included in the search
string). Note that in vi mode
you are left in command mode.
history-re-search-backward -Recall the next oldest line
who's prefix matches that
established by the last
invocation of either
history-search-forward or
history-search-backward.
history-re-search-forward - Recall the next newest line
who's prefix matches that
established by the last
invocation of either
history-search-forward or
history-search-backward.
complete-word - Attempt to complete the
incomplete word which
precedes the cursor. Unless
the host program has customized
word completion, filename
completion is attempted. In vi
commmand mode the character
under the cursor is also
included in the word being
completed, and you are left in
vi insert mode.
expand-filename - Within the command line, expand
wild cards, tilde expressions
and dollar expressions in the
filename which immediately
precedes the cursor. In vi
commmand mode the character
under the cursor is also
included in the filename being
expanded, and you are left in
vi insert mode.
list-glob - List any filenames which match
the wild-card, tilde and dollar
expressions in the filename
which immediately precedes the
cursor, then redraw the input
line unchanged.
read-from-file - Temporarily switch to reading
input from the file who's
name precedes the cursor.
beginning-of-history - Move to the oldest line in the
history list. Note that in vi
mode you are left in command
mode.
end-of-history - Move to the newest line in the
history list (ie. the current
line). Note that in vi mode
this leaves you in command
mode.
digit-argument - Enter a repeat count for the
next key-binding function.
For details, see the ENTERING
REPEAT COUNTS section.
newline - Terminate and return the
current contents of the line.
repeat-history - Return the line that is being
edited, then arrange for the
next most recent entry in the
history buffer to be recalled
when gl_get_line() is
next called. Repeatedly
invoking this action causes
successive historical input
lines to be re-executed. Note
that this action is equivalent
to the 'Operate' action in
ksh.
ring-bell - Ring the terminal bell, unless
the bell has been silenced via
the nobeep configuration
option (see the THE TECLA
CONFIGURATION FILE section).
forward-copy-char - Copy the next character into
the cut buffer (NB. use repeat
counts to copy more than one).
backward-copy-char - Copy the previous character
into the cut buffer.
forward-copy-word - Copy the next word into the cut
buffer.
backward-copy-word - Copy the previous word into the
cut buffer.
forward-find-char - Move the cursor to the next
occurrence of the next
character that you type.
backward-find-char - Move the cursor to the last
occurrence of the next
character that you type.
forward-to-char - Move the cursor to the
character just before the next
occurrence of the next
character that the user types.
backward-to-char - Move the cursor to the
character just after the last
occurrence before the cursor
of the next character that the
user types.
repeat-find-char - Repeat the last
backward-find-char,
forward-find-char,
backward-to-char or
forward-to-char.
invert-refind-char - Repeat the last
backward-find-char,
forward-find-char,
backward-to-char, or
forward-to-char in the
opposite direction.
delete-to-column - Delete the characters from the
cursor up to the column that
is specified by the repeat
count.
delete-to-parenthesis - Delete the characters from the
cursor up to and including
the matching parenthesis, or
next close parenthesis.
forward-delete-find - Delete the characters from the
cursor up to and including the
following occurence of the
next character typed.
backward-delete-find - Delete the characters from the
cursor up to and including the
preceding occurence of the
next character typed.
forward-delete-to - Delete the characters from the
cursor up to, but not
including, the following
occurence of the next
character typed.
backward-delete-to - Delete the characters from the
cursor up to, but not
including, the preceding
occurence of the next
character typed.
delete-refind - Repeat the last *-delete-find
or *-delete-to action.
delete-invert-refind - Repeat the last *-delete-find
or *-delete-to action, in the
opposite direction.
copy-to-column - Copy the characters from the
cursor up to the column that
is specified by the repeat
count, into the cut buffer.
copy-to-parenthesis - Copy the characters from the
cursor up to and including
the matching parenthesis, or
next close parenthesis, into
the cut buffer.
forward-copy-find - Copy the characters from the
cursor up to and including the
following occurence of the
next character typed, into the
cut buffer.
backward-copy-find - Copy the characters from the
cursor up to and including the
preceding occurence of the
next character typed, into the
cut buffer.
forward-copy-to - Copy the characters from the
cursor up to, but not
including, the following
occurence of the next
character typed, into the cut
buffer.
backward-copy-to - Copy the characters from the
cursor up to, but not
including, the preceding
occurence of the next
character typed, into the cut
buffer.
copy-refind - Repeat the last *-copy-find
or *-copy-to action.
copy-invert-refind - Repeat the last *-copy-find
or *-copy-to action, in the
opposite direction.
vi-mode - Switch to vi mode from emacs
mode.
emacs-mode - Switch to emacs mode from vi
mode.
vi-insert - From vi command mode, switch to
insert mode.
vi-overwrite - From vi command mode, switch to
overwrite mode.
vi-insert-at-bol - From vi command mode, move the
cursor to the start of the line
and switch to insert mode.
vi-append-at-eol - From vi command mode, move the
cursor to the end of the line
and switch to append mode.
vi-append - From vi command mode, move the
cursor one position right, and
switch to insert mode.
vi-replace-char - From vi command mode, replace
the character under the cursor
with the the next character
entered.
vi-forward-change-char - From vi command mode, delete
the next character then enter
insert mode.
vi-backward-change-char - From vi command mode, delete
the preceding character then
enter insert mode.
vi-forward-change-word - From vi command mode, delete
the next word then enter
insert mode.
vi-backward-change-word - From vi command mode, delete
the preceding word then
enter insert mode.
vi-change-rest-of-line - From vi command mode, delete
from the cursor to the end of
the line, then enter insert
mode.
vi-change-line - From vi command mode, delete
the current line, then enter
insert mode.
vi-change-to-bol - From vi command mode, delete
all characters between the
cursor and the beginning of
the line, then enter insert
mode.
vi-change-to-column - From vi command mode, delete
the characters from the cursor
up to the column that is
specified by the repeat count,
then enter insert mode.
vi-change-to-parenthesis - Delete the characters from the
cursor up to and including
the matching parenthesis, or
next close parenthesis, then
enter vi insert mode.
vi-forward-change-find - From vi command mode, delete
the characters from the
cursor up to and including the
following occurence of the
next character typed, then
enter insert mode.
vi-backward-change-find - From vi command mode, delete
the characters from the
cursor up to and including the
preceding occurence of the
next character typed, then
enter insert mode.
vi-forward-change-to - From vi command mode, delete
the characters from the
cursor up to, but not
including, the following
occurence of the next
character typed, then enter
insert mode.
vi-backward-change-to - From vi command mode, delete
the characters from the
cursor up to, but not
including, the preceding
occurence of the next
character typed, then enter
insert mode.
vi-change-refind - Repeat the last
vi-*-change-find or
vi-*-change-to action.
vi-change-invert-refind - Repeat the last
vi-*-change-find or
vi-*-change-to action, in the
opposite direction.
vi-undo - In vi mode, undo the last
editing operation.
vi-repeat-change - In vi command mode, repeat the
last command that modified the
line.
DEFAULT KEY BINDINGS IN EMACS MODE
The following default key bindings, which can be overriden
by the tecla configuration file, are designed to mimic most
of the bindings of the unix tcsh shell, when it is in emacs
editing mode.
This is the default editing mode of the tecla library.
Note that a key sequence like ^A or C-a means hold the
control-key down while pressing the letter A, and that where
you see \E or M- in a binding, this represents the escape
key or the Meta modifier key. Also note that to
gl_get_line(), pressing the escape key before a key is
equivalent to pressing the meta key at the same time as that
key. Thus the key sequence M-p can be typed in two ways, by
pressing the escape key, followed by pressing p, or by
pressing the Meta key at the same time as p.
Under UNIX the terminal driver sets a number of special keys
for certain functions. The tecla library attempts to use the
same keybindings to maintain consistency. The key sequences
shown for the following 6 bindings are thus just examples of
what they will probably be set to. If you have used the stty
command to change these keys, then the default bindings
should match.
^C -> user-interrupt
^\ -> abort
^Z -> suspend
^Q -> start-output
^S -> stop-output
^V -> literal-next
The cursor keys are refered to by name, as follows. This is
necessary because different types of terminals generate dif-
ferent key sequences when their cursor keys are pressed.
right -> cursor-right
left -> cursor-left
up -> up-history
down -> down-history
The remaining bindings don't depend on the terminal sett-
tings.
^F -> cursor-right
^B -> cursor-left
M-i -> insert-mode
^A -> beginning-of-line
^E -> end-of-line
^U -> delete-line
^K -> kill-line
M-f -> forward-word
M-b -> backward-word
^D -> del-char-or-list-or-eof
^H -> backward-delete-char
^? -> backward-delete-char
M-d -> forward-delete-word
M-^H -> backward-delete-word
M-^? -> backward-delete-word
M-u -> upcase-word
M-l -> downcase-word
M-c -> capitalize-word
^R -> redisplay
^L -> clear-screen
^T -> transpose-chars
^@ -> set-mark
^X^X -> exchange-point-and-mark
^W -> kill-region
M-w -> copy-region-as-kill
^Y -> yank
^P -> up-history
^N -> down-history
M-p -> history-search-backward
M-n -> history-search-forward
^I -> complete-word
^X* -> expand-filename
^X^F -> read-from-file
M-< -> beginning-of-history
M-> -> end-of-history
\n -> newline
\r -> newline
M-o -> repeat-history
M-^V -> vi-mode
M-0, M-1, ... M-9 -> digit-argument (see below)
Note that ^I is what the TAB key generates, and that ^@ can
be generated not only by pressing the control key and the @
key simultaneously, but also by pressing the control key and
the space bar at the same time.
DEFAULT KEY BINDINGS IN VI MODE
The following default key bindings are designed to mimic the
vi style of editing as closely as possible. This means that
very few editing functions are provided in the initial char-
acter input mode, editing functions instead being provided
by the vi command mode. Vi command mode is entered whenever
the escape character is pressed, or whenever a key-sequence
that starts with a meta character is entered. In addition to
mimicing vi, libtecla provides bindings for tab completion,
wild-card expansion of file names, and historical line
recall.
To learn how to tell the tecla library to use vi mode
instead of the default emacs editing mode, see the section
entitled THE TECLA CONFIGURATION FILE.
As already mentioned above in the emacs section, Note that a
key sequence like ^A or C-a means hold the control-key down
while pressing the letter A, and that where you see \E or M-
in a binding, this represents the escape key or the Meta
modifier key. Also note that to gl_get_line(), pressing the
escape key before a key is equivalent to pressing the meta
key at the same time as that key. Thus the key sequence M-p
can be typed in two ways, by pressing the escape key, fol-
lowed by pressing p, or by pressing the Meta key at the same
time as p.
Under UNIX the terminal driver sets a number of special keys
for certain functions. The tecla library attempts to use the
same keybindings to maintain consistency, binding them both
in input mode and in command mode. The key sequences shown
for the following 6 bindings are thus just examples of what
they will probably be set to. If you have used the stty com-
mand to change these keys, then the default bindings should
match.
^C -> user-interrupt
^\ -> abort
^Z -> suspend
^Q -> start-output
^S -> stop-output
^V -> literal-next
M-^C -> user-interrupt
M-^\ -> abort
M-^Z -> suspend
M-^Q -> start-output
M-^S -> stop-output
Note that above, most of the bindings are defined twice,
once as a raw control code like ^C and then a second time as
a meta character like M-^C. The former is the binding for vi
input mode, whereas the latter is the binding for vi command
mode. Once in command mode all key-sequences that the user
types that they don't explicitly start with an escape or a
meta key, have their first key secretly converted to a meta
character before the key sequence is looked up in the key
binding table. Thus, once in command mode, when you type the
letter i, for example, the tecla library actually looks up
the binding for M-i.
The cursor keys are refered to by name, as follows. This is
necessary because different types of terminals generate dif-
ferent key sequences when their cursor keys are pressed.
right -> cursor-right
left -> cursor-left
up -> up-history
down -> down-history
The cursor keys normally generate a keysequence that start
with an escape character, so beware that using the arrow
keys will put you into command mode (if you aren't already
in command mode).
The following are the terminal-independent key bindings for
vi input mode.
^D -> list-or-eof
^G -> list-glob
^H -> backward-delete-char
^I -> complete-word
\r -> newline
\n -> newline
^L -> clear-screen
^N -> down-history
^P -> up-history
^R -> redisplay
^U -> backward-kill-line
^W -> backward-delete-word
^X* -> expand-filename
^X^F -> read-from-file
^? -> backward-delete-char
The following are the key bindings that are defined in vi
command mode, this being specified by them all starting with
a meta character. As mentioned above, once in command mode
the initial meta character is optional. For example, you
might enter command mode by typing Esc, and then press h
twice to move the cursor two positions to the left. Both h
characters get quietly converted to M-h before being com-
pared to the key-binding table, the first one because Escape
followed by a character is always converted to the
equivalent meta character, and the second because command
mode was already active.
M-\ -> cursor-right (Meta-space)
M-$ -> end-of-line
M-* -> expand-filename
M-+ -> down-history
M-- -> up-history
M-< -> beginning-of-history
M-> -> end-of-history
M-^ -> beginning-of-line
M-; -> repeat-find-char
M-, -> invert-refind-char
M-| -> goto-column
M-~ -> change-case
M-. -> vi-repeat-change
M-% -> find-parenthesis
M-a -> vi-append
M-A -> vi-append-at-eol
M-b -> backward-word
M-B -> backward-word
M-C -> vi-change-rest-of-line
M-cb -> vi-backward-change-word
M-cB -> vi-backward-change-word
M-cc -> vi-change-line
M-ce -> vi-forward-change-word
M-cE -> vi-forward-change-word
M-cw -> vi-forward-change-word
M-cW -> vi-forward-change-word
M-cF -> vi-backward-change-find
M-cf -> vi-forward-change-find
M-cT -> vi-backward-change-to
M-ct -> vi-forward-change-to
M-c; -> vi-change-refind
M-c, -> vi-change-invert-refind
M-ch -> vi-backward-change-char
M-c^H -> vi-backward-change-char
M-c^? -> vi-backward-change-char
M-cl -> vi-forward-change-char
M-c\ -> vi-forward-change-char (Meta-c-space)
M-c^ -> vi-change-to-bol
M-c0 -> vi-change-to-bol
M-c$ -> vi-change-rest-of-line
M-c| -> vi-change-to-column
M-c% -> vi-change-to-parenthesis
M-dh -> backward-delete-char
M-d^H -> backward-delete-char
M-d^? -> backward-delete-char
M-dl -> forward-delete-char
M-d -> forward-delete-char (Meta-d-space)
M-dd -> delete-line
M-db -> backward-delete-word
M-dB -> backward-delete-word
M-de -> forward-delete-word
M-dE -> forward-delete-word
M-dw -> forward-delete-word
M-dW -> forward-delete-word
M-dF -> backward-delete-find
M-df -> forward-delete-find
M-dT -> backward-delete-to
M-dt -> forward-delete-to
M-d; -> delete-refind
M-d, -> delete-invert-refind
M-d^ -> backward-kill-line
M-d0 -> backward-kill-line
M-d$ -> kill-line
M-D -> kill-line
M-d| -> delete-to-column
M-d% -> delete-to-parenthesis
M-e -> forward-word
M-E -> forward-word
M-f -> forward-find-char
M-F -> backward-find-char
M-- -> up-history
M-h -> cursor-left
M-H -> beginning-of-history
M-i -> vi-insert
M-I -> vi-insert-at-bol
M-j -> down-history
M-J -> history-search-forward
M-k -> up-history
M-K -> history-search-backward
M-l -> cursor-right
M-L -> end-of-history
M-n -> history-re-search-forward
M-N -> history-re-search-backward
M-p -> append-yank
M-P -> yank
M-r -> vi-replace-char
M-R -> vi-overwrite
M-s -> vi-forward-change-char
M-S -> vi-change-line
M-t -> forward-to-char
M-T -> backward-to-char
M-u -> vi-undo
M-w -> forward-to-word
M-W -> forward-to-word
M-x -> forward-delete-char
M-X -> backward-delete-char
M-yh -> backward-copy-char
M-y^H -> backward-copy-char
M-y^? -> backward-copy-char
M-yl -> forward-copy-char
M-y\ -> forward-copy-char (Meta-y-space)
M-ye -> forward-copy-word
M-yE -> forward-copy-word
M-yw -> forward-copy-word
M-yW -> forward-copy-word
M-yb -> backward-copy-word
M-yB -> backward-copy-word
M-yf -> forward-copy-find
M-yF -> backward-copy-find
M-yt -> forward-copy-to
M-yT -> backward-copy-to
M-y; -> copy-refind
M-y, -> copy-invert-refind
M-y^ -> copy-to-bol
M-y0 -> copy-to-bol
M-y$ -> copy-rest-of-line
M-yy -> copy-line
M-Y -> copy-line
M-y| -> copy-to-column
M-y% -> copy-to-parenthesis
M-^E -> emacs-mode
M-^H -> cursor-left
M-^? -> cursor-left
M-^L -> clear-screen
M-^N -> down-history
M-^P -> up-history
M-^R -> redisplay
M-^D -> list-or-eof
M-^I -> complete-word
M-\r -> newline
M-\n -> newline
M-0, M-1, ... M-9 -> digit-argument (see below)
Note that ^I is what the TAB key generates.
ENTERING REPEAT COUNTS
Many of the key binding functions described previously, take
an optional count, typed in before the target keysequence.
This is interpretted as a repeat count by most bindings. A
notable exception is the goto-column binding, which inter-
prets the count as a column number.
By default you can specify this count argument by pressing
the meta key while typing in the numeric count. This relies
on the digit-argument action being bound to Meta-0, Meta-1
etc. Once any one of these bindings has been activated, you
can optionally take your finger off the meta key to type in
the rest of the number, since every numeric digit thereafter
is treated as part of the number, unless it is preceded by
the literal-next binding. As soon as a non-digit, or literal
digit key is pressed the repeat count is terminated and
either causes the just typed character to be added to the
line that many times, or causes the next key-binding func-
tion to be given that argument.
For example, in emacs mode, typing:
M-12a
causes the letter 'a' to be added to the line 12 times,
whereas
M-4M-c
Capitalizes the next 4 words.
In vi command mode the Meta modifier is automatically added
to all characters typed in, so to enter a count in vi
command-mode, just involves typing in the number, just as at
it does in the vi editor itself. So for example, in vi com-
mand mode, typing:
4w2x
moves the cursor four words to the right, then deletes two
characters.
You can also bind digit-argument to other key sequences. If
these end in a numeric digit, that digit gets appended to
the current repeat count. If it doesn't end in a numeric
digit, a new repeat count is started with a value of zero,
and can be completed by typing in the number, after letting
go of the key which triggered the digit-argument action.
THE TECLA CONFIGURATION FILE
When new_GetLine() is called, it first defines the default
emacs key bindings, as listed above, then looks for a file
called .teclarc in your home directory (ie. ~/.teclarc). If
it finds this file, it reads it, interpreting each line as
defining a new key binding or an editing configuration
option. For those wanting to use the non-default vi editing
mode, the most important item to go in this file is the fol-
lowing line:
edit-mode vi
This will re-configure the default bindings for vi-mode.
To prevent the terminal bell from being rung, such as when
an unrecognized control-sequence is typed, place the follow-
ing line in the configuration file:
nobeep
An example of a key binding line in the configuration file
is the following.
bind M-[2~ insert-mode
On a Sun keyboard, the above key sequence is generated when
one presses the insert key, so with this keybinding, one can
toggle between emacs insert and overwrite mode by hitting
one key. One could also do it by typing out the above
sequence of characters one by one. As explained above, the
M- part of this sequence can be typed either by pressing the
escape key before the following key, or by pressing the Meta
key at the same time as the following key. Thus if you had
set the above key binding, and the insert key on your key-
board didn't generate the above key sequence, you could
still type it in either of the following 2 ways.
1. Hit the escape key momentarily, then press '[', then '2', then
finally '~'.
2. Press the meta key at the same time as pressing the '[' key,
then press '2', then '~'.
If you set a keybinding for a key-sequence that is already
bound to a function, the new binding overrides the old one.
If in the new binding you omit the name of the new function
to bind to the key-sequence, the original binding becomes
undefined.
Note that all key-sequences are required to start with a
control or meta character. The only exception to this rule
is for symbolically named characters such as the arrow keys.
For example, to rebind the up and down arrow keys to use the
history search mechanism instead of the simple recall
method, you would need to place the following in your confi-
guration file:
bind up history-search-backwards
bind down history-search-backwards
FILENAME AND TILDE COMPLETION
With the default key bindings, pressing the TAB key (aka.
^I) results in gl_get_line() attempting to complete the
incomplete filename that precedes the cursor. gl_get_line()
searches backwards from the cursor, looking for the start of
the filename, stopping when it hits either a space or the
start of the line. If more than one file has the specified
prefix, gl_get_line() completes the filename up to the point
at which the ambiguous matches start to differ, then lists
the possible matches.
In addition to literally written filenames, gl_get_line()
can complete files that start with ~/ and ~user/ expressions
and that contain $envvar expressions. In particular, if you
hit TAB within an incomplete ~user, expression,
gl_get_line() will attempt to complete the username, listing
any ambiguous matches.
The completion binding is implemented using the
cpl_word_completions() function, which is also available
separately to users of this library. See the
cpl_word_completions(3) man page for more details.
CUSTOMIZED WORD COMPLETION
If in your application, you would like to have TAB comple-
tion complete other things in addition or instead of
filenames, you can arrange this by registering an alternate
completion callback function, via a call to the
gl_customize_completion() function.
int gl_customize_completion(GetLine *gl, void *data,
CplMatchFn *match_fn);
The data argument provides a way for your application to
pass arbitrary, application-specific information to the
callback function. This is passed to the callback every time
that it is called. It might for example, point to the symbol
table from which possible completions are to be sought. The
match_fn argument specifies the callback function to be
called. The CplMatchFn function type is defined in
libtecla.h, as is a CPL_MATCH_FN() macro that you can use to
declare and prototype callback functions. The declaration
and responsibilities of callback functions are described in
depth in the cpl_complete_word(3) man page.
In brief, the callback function is responsible for looking
backwards in the input line, back from the point at which
the user pressed TAB, to find the start of the word being
completed. It then must lookup possible completions of this
word, and record them one by one in the WordCompletion
object that is passed to it as an argument, by calling the
cpl_add_completion() function. If the callback function
wishes to provide filename completion in addition to its own
specific completions, it has the option of itself calling
the builtin file-name completion callback. This also, is
documented in the cpl_complete_word(3) man page.
FILENAME EXPANSION
With the default key bindings, pressing ^X* causes
gl_get_line() to expand the filename that precedes the cur-
sor, replacing ~/ and ~user/ expressions with the
corresponding home directories, and replacing $envvar
expressions with the value of the specified environment
variable, then if there are any wildcards, replacing the so
far expanded filename with a space-separated list of the
files which match the wild cards.
The expansion binding is implemented using the
ef_expand_file() function. See the ef_expand_file(3) man
page for more details.
RECALLING PREVIOUSLY TYPED LINES
Every time that a new line is entered by the user, it is
appended to a list of historical input lines maintained
within the GetLine resource object. You can traverse up and
down this list using the up and down arrow keys. Alterna-
tively, you can do the same with the ^P, and ^N keys, and in
vi command mode you can alternatively use the k and j char-
acters. Thus pressing up-arrow once, replaces the current
input line with the previously entered line. Pressing up-
arrow again, replaces this with the line that was entered
before it, etc.. Having gone back one or more lines into the
history list, one can return to newer lines by pressing
down-arrow one or more times. If you do this sufficient
times, you will return to the original line that you were
entering when you first hit up-arrow.
Note that in vi mode, all of the history recall functions
switch the library into command mode.
In emacs mode the M-p and M-n keys work just like the ^P and
^N keys, except that they skip all but those historical
lines which share the prefix that precedes the cursor. In vi
command mode the upper case K and J characters do the same
thing, except that the string that they search for includes
the character under the cursor as well as what precedes it.
Thus for example, suppose that you were in emacs mode, and
you had just entered the following list of commands in the
order shown:
ls ~/tecla/
cd ~/tecla
ls -l getline.c
emacs ~/tecla/getline.c
If you next typed:
ls
and then hit M-p, then rather than returning the previously
typed emacs line, which doesn't start with "ls",
gl_get_line() would recall the "ls -l getline.c" line.
Pressing M-p again would recall the "ls ~/tecla/" line.
CHANGING TERMINALS
The new_GetLine() constructor function assumes that input is
to be read from stdin, and output written to stdout. The
following function allows you to switch to different input
and output streams.
int gl_change_terminal(GetLine *gl, FILE *input_fp,
FILE *output_fp, const char *term);
The gl argument is the object that was returned by
new_GetLine(). The input_fp argument specifies the stream
to read from, and output_fp specifies the stream to be writ-
ten to. Only if both of these refer to a terminal, will
interactive terminal input be enabled. Otherwise
gl_get_line() will simply call fgets() to read command
input. If both streams refer to a terminal, then they must
refer to the same terminal, and the type of this terminal
must be specified via the term argument. The value of the
term argument is looked up in the terminal information data-
base (terminfo or termcap), in order to determine which spe-
cial control sequences are needed to control various aspects
of the terminal. new_GetLine() for example, passes the
return value of getenv("TERM") in this argument. Note that
if one or both of input_fp and output_fp don't refer to a
terminal, then it is legal to pass NULL instead of a termi-
nal type.
Note that if you want to pass file descriptors to
gl_change_terminal(), you can do this by creating stdio
stream wrappers using the POSIX fdopen() function.
THREAD SAFETY
In a multi-threaded program, you should use the libtecla_r.a
version of the library. This uses reentrant versions of sys-
tem functions, where available. Unfortunately neither ter-
minfo nor termcap were designed to be reentrant, so you
can't safely use the functions of the getline module in mul-
tiple threads (you can use the separate file-expansion and
word-completion modules in multiple threads, see the
corresponding man pages for details). However due to the use
of POSIX reentrant functions for looking up home directories
etc, it is safe to use this module from a single thread of a
multi-threaded program, provided that your other threads
don't use any termcap or terminfo functions.
FILES
libtecla.a - The tecla library
libtecla.h - The tecla header file.
~/.teclarc - The personal tecla customization file.
SEE ALSO
libtecla(3), ef_expand_file(3), cpl_complete_word(3),
pca_lookup_file(3)
AUTHOR
Martin Shepherd (mcs@astro.caltech.edu)