Commit Graph

351 Commits

Author SHA1 Message Date
nicm
115bb33257 Ignore default-shell (and use /bin/sh) if it invalid not just if it is
tmux itself, also refuse to set the option to something invalid in the
first place. GitHub issue 2120.
2020-03-17 11:10:12 +00:00
nicm
d162ff48f3 Send mouse down event immediately rather than waiting for double click
to finish which would now mean it was out of order. Reported by Mark
Kelly.
2020-03-16 06:12:42 +00:00
nicm
f7bc753442 Change how double and triple clicks works so that one or the other is
fired - a double click is no longer triggered on the way to a triple
click.
2020-03-12 13:16:16 +00:00
nicm
b20753f2a3 A few fixes to make modifier keys and dragging work - need to remove the
modifiers before checking for the dragging marker key, and apply them
before looking up the end key. Also fix key-to-string with modifiers for
special keys.
2020-02-19 14:25:00 +00:00
nicm
dc882adb2e Remove unused variables from Ben Boeckel, and a Pp from jmc. 2020-02-11 07:01:08 +00:00
nicm
24350879cd Add a define for flags meaning a client is not attached, and fix
unattached counter, reported by Thomas Sattler.
2020-01-28 08:06:11 +00:00
nicm
1bdd4828bd If /dev/fd/X is a symlink and realpath() expands symlinks, /dev/fd/X
ends up pointing to the wrong place before it is passed to the client.
The path is only used internally so there is no real need for
realpath(), remove it and move the get_path function to file.c where all
the callers are.
2019-12-16 16:39:03 +00:00
nicm
eaa58d28dc Instead of using large buffers in imsgs, add the data or path onto the end. 2019-12-16 15:48:50 +00:00
nicm
c284ebe0ad Rewrite the code for reading and writing files. Now, if the client is
not attached, the server process asks it to open the file, similar to
how works for stdin, stdout, stderr. This makes special files like
/dev/fd/X work (used by some shells). stdin, stdout and stderr and
control mode are now just special cases of the same mechanism. This will
also make it easier to use for other commands that read files such as
source-file.
2019-12-12 11:39:56 +00:00
nicm
7826d40ff9 Style nits in function arguments. 2019-12-03 10:47:22 +00:00
nicm
bc5881c4d2 Long lines and spacing fixes. 2019-11-28 09:56:25 +00:00
nicm
2349b1dbef Make a best effort to set xpixel and ypixel for each pane and add
formats for them.
2019-11-28 09:45:15 +00:00
nicm
bad95db878 Limit lazy resize to panes in attached sessions only - those in
unattached are likely to have been resized by something like
split-window where the user probably wants the resize to happen
immediately. GitHub issue 1963.
2019-11-01 20:26:21 +00:00
nicm
647887b794 Add a "latest" window-size option which tries to size windows based on
the most recently used client. From Tommie Gannert in GitHub issue 1869
based on earlier changes from me.
2019-09-19 09:02:30 +00:00
nicm
df0334d3b3 The resize event was never deciding to actually resize the pane if there
was output in the pane faster than the timer would fire, so change how
it works to only defer the timer again if the pane was actually resized
within the last timer period. Reported by James Tai in GitHub issue
1880.
2019-08-28 07:34:32 +00:00
nicm
9e7774bb96 Clear overlay on normal key press. 2019-07-17 17:46:51 +00:00
nicm
ddf53d6e4e Correctly adjust mouse position if the status line is at the top and
more than one line. GitHub issue 1822.
2019-07-06 20:56:34 +00:00
nicm
6a489fa7f6 Command prompt key presses need to avoid the command queue, GitHub issue
1817. Also a tmux.1 fix from jmc.
2019-07-02 20:09:19 +00:00
nicm
87ea14328c Pass keys that aren't 0-9 on to normal key processing when display-panes
is active (restores previous behaviour).
2019-06-26 18:28:31 +00:00
nicm
97a317a656 Need to always check focus even if not current window. 2019-06-20 19:29:38 +00:00
nicm
ae541287d3 Expand command formats in %if and move the config file loading later (to
when the first client has identified) so all the client formats are
available, fixes problems reported by Thomas Sattler.
2019-06-20 06:51:36 +00:00
nicm
1a9f9c09b4 Do not resize panes unless they are in an attached, active window. From
Morten M Neergaard in GitHub issue 1782.
2019-06-11 13:09:00 +00:00
nicm
e37f34facc Do not load the config file if the server is exiting because it failed
to start, otherwise commands like lsk which start the server again can
end up looping infinitely. Also make the first client exit
correctly. Problem reported by Wael M Nasreddine.
2019-06-07 20:09:17 +00:00
nicm
f8d3d247d8 Merge cmd_list_parse into cmd-parse.y so it can use the new alias
processing code.
2019-05-25 07:18:20 +00:00
nicm
e128c7fcd8 Replace the various identical error callbacks with a single one in cmd-queue.c. 2019-05-20 11:46:06 +00:00
nicm
00f19b7f91 Fix some indentation and dead assignments. 2019-05-12 18:16:33 +00:00
nicm
f9682d2e55 Add a flag to redraw only the overlay, and remove the overlay on resize. 2019-05-08 18:07:12 +00:00
nicm
89db309e10 Move around the display-panes identify code to make it a bit more
generic and hide the display-panes specific bits into
cmd-display-panes.c.
2019-05-07 20:01:41 +00:00
nicm
85a9c2f52b Treat keys in identify mode (display-panes) specially and handle them
immediately rather than queuing them (the command can block the queue
which means they were not being seen until it finished which was too
late). Reported by denis@ and solene@, ok solene@.
2019-05-07 11:24:03 +00:00
nicm
9f75635596 Allow panes to be empty (no command), output can be piped to them with
split-window or display-message -I.
2019-05-03 20:44:24 +00:00
nicm
4097257bef Do not store the mouse position we calculate as the start of a drag back
into the mouse event that later code uses, it has been adjusted and they
should use the original position. GitHub issue 1710.
2019-05-03 18:42:40 +00:00
nicm
fc3d85e34b Fix mouse positioning when the pane is not entirely visible. 2019-05-03 16:51:29 +00:00
nicm
33298d6df6 Instead of processing keys all together, put them up on the client
command queue so they are ordered correctly with the commands that they
execute.
2019-05-03 14:51:30 +00:00
nicm
f3ab05e7cd Update session activity on focus event, from tafryn at gmail dot com. 2019-04-18 10:11:52 +00:00
nicm
d21f8ecc13 Add StatusDefault binding for the mouse on any otherwise unassigned
parts of the status line, from Avi Halachmi.
2019-03-25 18:59:55 +00:00
nicm
517d673dbe Ignore mouse on status line which are not part of a range, GitHub issue 1649. 2019-03-25 09:22:09 +00:00
nicm
979313832c Extend the #[] style syntax and use that together with previous format
changes to allow the status line to be entirely configured with a single
option.

Now that it is possible to configure their content, enable the existing
code that lets the status line be multiple lines in height. The status
option can now take a value of 2, 3, 4 or 5 (as well as the previous on
or off) to configure more than one line. The new status-format array
option configures the format of each line, the default just references
the existing status-* options, although some of the more obscure status
options may be eliminated in time.

Additions to the #[] syntax are: "align" to specify alignment (left,
centre, right), "list" for the window list and "range" to configure
ranges of text for the mouse bindings.

The "align" keyword can also be used to specify alignment of entries in
tree mode and the pane status lines.
2019-03-18 20:53:33 +00:00
nicm
b4f5b99e4b Tidy and rename some bits of status line code. 2019-03-16 17:14:07 +00:00
nicm
85044a634b Move status line free into its own function. 2019-03-15 14:46:58 +00:00
nicm
10d60faba5 Store the time in the format tree rather than passing it around. 2019-03-14 23:14:27 +00:00
nicm
1416ceb575 Accept 0 time as a shorthand for now to format_expand_time. 2019-03-14 21:27:26 +00:00
nicm
3f6bfbaf2b Allow multiple modes to be open in a pane. A stack of open modes is kept
and the previous restored when the top is exited. If a mode that is
already on the stack is entered, the existing instance is moved to the
top as the active mode rather than being opened new.
2019-03-12 11:16:49 +00:00
nicm
f98c66ece8 Add a separate mode struct for the active window mode if any. 2019-03-07 20:24:21 +00:00
nicm
82f0c859a2 Use starting client cwd in config file, GitHub issue 1606. 2019-02-16 11:42:08 +00:00
nicm
bde0224deb Pass window into mode functions. 2018-12-18 13:20:44 +00:00
nicm
749f67b7d8 evbuffer_new and bufferevent_new can both fail (when malloc fails) and
return NULL. GitHub issue 1547.
2018-11-19 13:35:40 +00:00
nicm
0c7f64458f If a non-repeating key is used when repeating, it should be treated as
an entirely new key press, not checked in root table and ignored if not
found. GitHub issue 1513.
2018-11-07 08:06:28 +00:00
nicm
646995384d Support for windows larger than visible on the attached client. This has
been a limitation for a long time.

There are two new options, window-size and default-size, and a new
command, resize-window. The force-width and force-height options and the
session_width and session_height formats have been removed.

The new window-size option tells tmux how to work out the size of
windows: largest means it picks the size of the largest session,
smallest the smallest session (similar to the old behaviour) and manual
means that it does not automatically resize windows. The default is
currently largest but this may change. aggressive-resize modifies the
choice of session for largest and smallest as it did before.

If a window is in a session attached to a client that is too small, only
part of the window is shown. tmux attempts to keep the cursor visible,
so the part of the window displayed is changed as the cursor moves (with
a small delay, to try and avoid excess redrawing when applications
redraw status lines or similar that are not currently visible). The
offset of the visible portion of the window is shown in status-right.

Drawing windows which are larger than the client is not as efficient as
those which fit, particularly when the cursor moves, so it is
recommended to avoid using this on slow machines or networks (set
window-size to smallest or manual).

The resize-window command can be used to resize a window manually. If it
is used, the window-size option is automatically set to manual for the
window (undo this with "setw -u window-size"). resize-window works in a
similar way to resize-pane (-U -D -L -R -x -y flags) but also has -a and
-A flags. -a sets the window to the size of the smallest client (what it
would be if window-size was smallest) and -A the largest.

For the same behaviour as force-width or force-height, use resize-window
-x or -y, and "setw -u window-size" to revert to automatic sizing..

If the global window-size option is set to manual, the default-size
option is used for new windows. If -x or -y is used with new-session,
that sets the default-size option for the new session.

The maximum size of a window is 10000x10000. But expect applications to
complain and much higher memory use if making a window excessively
big. The minimum size is the size required for the current layout
including borders.

The refresh-client command can be used to pan around a window, -U -D -L
-R moves up, down, left or right and -c returns to automatic cursor
tracking. The position is reset when the current window is changed.
2018-10-18 08:38:01 +00:00
nicm
bd9133b31d Do not check for mouse events on pane borders when zoomed, based on a
fix from Avi Halachmi.
2018-09-11 06:37:54 +00:00
nicm
f57aa143c1 Keep any text killed in the command prompt with C-w and yank it with
C-y, only use the top buffer if no text has previously been killed. This
and previous change promped by discussion with kn@.
2018-08-29 09:50:32 +00:00
nicm
55db3623bf Add StatusLeft and StatusRight mouse key modifiers for the left and
right parts of the status line.
2018-08-22 20:06:14 +00:00
nicm
cac4eadca0 Add a flag to force redrawing of the status line even if the content
hasn't changed, needed for resizing.
2018-08-19 20:13:07 +00:00
nicm
88327c7698 Add a client redraw-window flag instead of the redraw-all flag and for
all just use the three flags together (window, borders, status).
2018-08-19 16:45:03 +00:00
nicm
bd2896b65e SESSION_UNATTACHED flag is no longer necessary now we have an attached
count instead.
2018-08-18 20:08:52 +00:00
nicm
6048b0f483 Make key trees and some other bits static. 2018-08-02 11:44:07 +00:00
nicm
969af935f3 When a key isn't in the first table, we need to try the same key again
not the any key. Also rename some labels. Fixes GitHub issue 1406
reeported by Mark Kelly.
2018-07-17 18:02:40 +00:00
nicm
0d88f8a78b Add an "Any" key to run a command if a key is pressed that is not bound
in the current key table. GitHub issue 1404.
2018-07-16 08:48:22 +00:00
nicm
d254293a6d Add pane focus hooks. 2018-07-04 12:25:26 +00:00
nicm
b9a6162d2f Make server_client_get_cwd used (almost) everywhere we need to work out
the cwd, and do not fall back to "." as it is pretty useless. GitHub
issue 1331.
2018-05-24 09:42:49 +00:00
nicm
19f3a5c612 Add a missing client-detached hook when the server shuts down, and do
not exit until jobs started from run-shell/if-shell have finished (add a
job flags member and a flag to indicate other jobs). GitHub issue 1245.
2018-03-08 08:09:10 +00:00
nicm
e97daead43 Check prefix when retrying so it is checked while repeat flag is
set. GitHub issue 1239.
2018-02-22 10:58:12 +00:00
nicm
7f4513ec34 Add struct status_line to hold status line members of struct client, not
used yet but will be soon. From Thomas Adam.
2018-02-05 08:21:54 +00:00
nicm
b20a00f93e Report better error from server when socket create fails, GitHub issue
1201.
2017-12-19 15:00:39 +00:00
nicm
2f6935a630 Infrastructure for drawing status lines of more than one line in height,
still only one is allowed but this lets tmux draw bigger ones.
2017-10-16 19:30:53 +00:00
nicm
466066c3a1 Do not attempt to use TIOCSWINSZ on a -1 file descriptor (possible if
the pane has already died).
2017-09-06 07:12:41 +00:00
nicm
17cf1b21c6 Pass flags into cmd_find_from_* to fix prefer-unattached, reported by
Thomas Sattler.
2017-08-30 10:33:57 +00:00
nicm
c6a8ad23a1 Add -d flag to display-panes to specify timeout, and make 0 mean no
timeout. From Laurens Post.
2017-08-16 12:12:54 +00:00
nicm
2678fe53f5 Fix redraw defer code in the presence of multiple clients - the timer
may be needed for all of them, so don't delete it on the first; and
don't skip setting the redraw flag if the timer is already running.

Reported by Pol Van Aubel in GitHub issue 1003.
2017-07-14 08:04:23 +00:00
nicm
d0d42dc4cb proc_send_s now seems unnecessary. 2017-07-12 09:07:52 +00:00
nicm
ac7080b31b Remove xterm flag from key before checking prefix, reported by Peter
Fern in GitHub issue 974.
2017-06-13 07:12:33 +00:00
nicm
50b27c8c0d Continue and pass keys through if they are repeated keys, so that the
first key after a repeated key doesn't get lost.
2017-06-06 15:49:35 +00:00
nicm
d60663ea86 Some applications like vi(1) and tmux until 10 minutes or so ago, do not
redraw on SIGWINCH if the size returns to the original size between the
original SIGWINCH and when they get around to calling TIOCGWINSZ. So use
the existing resize timer to introduce a small delay between the two
resizes.
2017-05-31 11:00:00 +00:00
nicm
ea6428a5d2 It is not OK to ignore SIGWINCH if SIOCGWINSZ reports the size has
unchanged, because it may have changed and changed back in the time
between us getting the signal and calling ioctl(). Always redraw when we
see SIGWINCH.
2017-05-31 10:29:15 +00:00
nicm
80c6b487dc Because we defer actually resizing applications (calling TIOCSWINSZ)
until the end of the server loop, tmux may have gone through several
internal resizes in between. This can be a problem if the final size is
the same as the initial size (what the application things it currently
is), because the application may choose not to redraw, assuming the
screen state is unchanged, when in fact tmux has thrown away parts of
the screen, assuming the application will redraw them.

To avoid this, do an extra resize if the new size is the same size as
the initial size. This should force the application to redraw when tmux
needs it to, while retaining the benefits of deferring (so we now resize
at most two times instead of at most one - and only two very rarely).

Fixes a problem with break-pane and zoomed panes reported by Michal
Mazurek.
2017-05-31 10:15:51 +00:00
nicm
b95e5827c1 Store a copy of the old status line, will be needed soon for new choose mode. 2017-05-29 20:41:29 +00:00
nicm
8a214b2f8e Function to count clients. 2017-05-29 20:37:30 +00:00
nicm
31625c2d17 Line length and spaces to tabs. 2017-05-16 12:57:26 +00:00
nicm
0e3c5ebe1a Insert copy mode bindings at the right place in the command queue. 2017-05-10 10:46:59 +00:00
nicm
3b35daacf7 If the current screen was complex enough, it was possible to make redraw
itself hit the "terminal can't keep up" check. To avoid this, record how
much data we send during redraw (we know we will be starting with 0) and
skip the check until it has been flushed. GitHub issue 912.
2017-05-09 13:04:36 +00:00
nicm
d52f579fd5 Up to now, tmux sees \033\033[OA as M-Up and since we turned on
xterm-keys by default, generates \033[1;3A instead of
\033\033[OA. Unfortunately this confuses vi, which doesn't understand
xterm keys and now sees Escape+Up pressed within escape-time as Escape
followed by A.

The issue doesn't happen in xterm itself because it gets the keys from X
and can distinguish between a genuine M-Up and Escape+Up.

Because xterm can, tmux can too: xterm will give us \033[1;3A (that is,
kUP3) for a real M-Up and \033\033OA for Escape+Up - in fact, we can be
sure any \033 preceding an xterm key is a real Escape key press because
Meta would be part of the xterm key instead of a separate \033.

So change tmux to recognise both sequences as M-Up for its own purposes,
but generate the xterm version of M-Up only if it originally received
the xterm version from the terminal.

This means we will return to sending \033\033OA instead of the xterm key
for terminals that do not support xterm keys themselves, but there is no
practical way around this because they do not allow us to distinguish
between Escape+Up and M-Up. xterm style escape sequences are now the de
facto standard for these keys in any case.

Problem reported by jsing@ and subsequently by Cecile Tonglet in GitHub
issue 907.
2017-05-07 21:25:59 +00:00
nicm
0ccfb61bb0 In order that people can use formats like #D in #() in the status line
and not have to wait for an update when they change pane, we allow
commands to run more than once a second if the expanded form
changes. Unfortunately this can mean them being run far too often
(pretty much continually) when multiple clients exist, because some
formats (including #D) will always differ between clients.

To avoid this, give each client its own tree of jobs which means that
the same command will be different instances for each client - similar
to how we have the tag to separate commands for different panes.

GitHub issue 889; test case reported by Paul Johnson.
2017-05-01 12:20:55 +00:00
nicm
2c0f826c36 Mouse bindings and hooks set up an initial current state when running a
command. This is used for the session, window and pane for all commands
in the command sequence if there is no -t or -s.

However, using it for all commands in the command sequence means that if
the active pane or current session is changed, subsequent commands still
use the previous state. So make commands which explicitly change the
current state (such as neww and selectp) update it themselves for later
commands. Commands which may invalidate the state (like killp) are
already OK because an invalid state will be ignored.

Also fill in the current state for all key bindings rather than just the
mouse, so that any omissions are easier to spot.
2017-04-22 08:56:24 +00:00
nicm
c8ecbf38ab Log error properly when no current state, and some other minor tweaks. 2017-04-21 22:23:24 +00:00
nicm
2ad09ab5af Key needs to be initialized to zero now it has flags in it. 2017-04-21 19:33:07 +00:00
nicm
afa4e3ed9c Add cmd_find_from_winlink_pane and use it in a couple of places, and
make functions that can't fail void.
2017-04-21 17:22:20 +00:00
nicm
92a77e7654 It is annoying that the copy mode key table (or any other key table)
will suppress root key table bindings. So change to always check the
root table if no binding is found in the current table (whether it be
the prefix table from pressing the prefix or the copy mode table from a
pane).

A root key binding can be blocked by binding the key to a command that
does nothing (like send-keys with no arguments).

Problem reported by Thomas Sattler.
2017-04-21 16:04:18 +00:00
nicm
bba588752f Store state shared between multiple commands in the queue in a shared
structure.
2017-04-21 14:01:19 +00:00
nicm
51a0dbb172 Only set up a current target for mouse key bindings. Fixes:
bind q select-pane -U \; resize-pane -Z

(There is still some possible weirdness with the way we do current
targets, it should probably be done in a different way at some point.)
2017-04-20 15:16:20 +00:00
nicm
53fde21bb8 Add a suspend helper function, and do not allow detaching or suspending
while already doing so.
2017-04-19 14:00:28 +00:00
nicm
fa6deb5866 When the data we have buffered to write to a terminal grows beyond a
reasonable amount (currently width * height * 8 bytes), discard all
output to the terminal and start trying to redraw periodically
instead. Continue with this until the amount of data we are trying to
write falls to a low level again.

This helps to prevent tmux sitting on a huge buffer of data when there
are processes with fast output running inside tmux but the outside
terminal is slow.

A new client_discarded format holds the amount of data that has been
discarded due to this mechanism.

The three variables (when to start this, when to stop, and how often to
redraw) are basically "works for me" at the moment, this is going in to
see how it goes and if it causes problems for anyone else.
2017-04-19 06:52:27 +00:00
nicm
8c8ce08d79 On terminals without DECSLRM, when a pane that is less than the full
with of the terminal scrolls, tmux needs to redraw the entire pane. This
results in a large amount of output data which can cause slow terminals
to struggle, particularly when many lines are scrolled together quickly.

This can be reduced by only redrawing when tmux doesn't hold any
buffered data for the terminal. If a redraw is required and data is
buffered, the redraw is deferred until all that data is consumed (it is
checked after every event loop, a timer is used to ensure this happens
at some point). While a redraw is pending, no additional data will be
written to the terminal.

The redraw still happens, now it is just pushed back if it is possible
it would just add more data on top of a terminal that is already
behind. This both gives the terminal a chance to catch up, and allows
tmux to process more scrolling (that would require additional redraws)
in the meantime.

Helps with a problem reported by Greg Hurrell.
2017-04-18 20:37:49 +00:00
nicm
7461c165b5 Remove a couple of redraw flags that no longer have any effect. 2017-04-17 06:40:32 +00:00
nicm
9b28200578 Give each client a name. This defaults to the tty name as before but
falls back to an alternative if the tty name is not available. This is
clearer than overloading the client ttyname member and allows us to
remove the path stored in the tty struct, it should always be the same
as the client.
2017-04-05 10:49:46 +00:00
nicm
55e73e3612 Clear the bracket paste mode when in the command prompt. 2017-03-09 22:00:46 +00:00
nicm
bce1dee034 Move the client identify (display-panes) code into server-client.c. 2017-03-09 17:06:35 +00:00
nicm
e340df2034 Make source-file look for files relative to the client working directory
(like load-buffer and save-buffer), from Chris Pickel. Also break the
where-is-this-file code out into its own function for loadb and saveb.
2017-02-14 18:13:05 +00:00
nicm
b1fa3e25e4 Break the message storage function into its own function, useful for
debugging.
2017-02-09 12:09:33 +00:00
nicm
7475165cd8 Some other tidying bits. 2017-02-08 15:49:29 +00:00
nicm
7d23d019c0 Add a window or pane id "tag" to each format tree and use it to separate
jobs, this means that if the same job is used for different windows or
panes (for example in pane-border-format), it will be run separately for
each pane.
2017-02-03 11:57:27 +00:00
nicm
dd0c814779 Implement "all event" (1003) mouse mode but in a way that works. The
main issue is that if we have two panes, A with 1002 and B with 1003, we
need to set 1003 outside tmux in order to get all the mouse events, but
then we need to suppress the ones that pane A doesn't want. This is easy
in SGR mouse mode, because buttons == 3 is only used for movement events
(for other events the trailing m/M marks a release instead), but in
normal mouse mode we can't tell so easily. So for that, look at the
previous event instead - if it is drag+release as well, then the current
event is a movement event.
2017-02-01 09:55:07 +00:00