Commit Graph

1956 Commits

Author SHA1 Message Date
nicm
661d0dfac9 Make key_code unsigned long long not uint64_t which is more portable for
printf formats, and move UTF8_SIZE define down to near the rest of the
UTF-8 bits.
2015-11-15 22:50:38 +00:00
nicm
a582b62287 Accidentally turned off pledge, turn it back on. 2015-11-15 14:32:48 +00:00
nicm
14d90e4901 The character is an int so use %x not %hhx. 2015-11-14 12:20:19 +00:00
nicm
3db0d50df4 The private use area at U+E000 to U+F8FF is not very useful if it is
width 0, make it width 1 instead.
2015-11-14 12:03:23 +00:00
nicm
205d15e82d All these return values from utf8_* are confusing, use an enum. 2015-11-14 11:45:43 +00:00
nicm
f401791a56 Rename a variable in utf8_combine for consistency and use 0xfffd for
unknown Unicode.
2015-11-14 11:13:44 +00:00
nicm
dab63b029e Couple of assignments to remove compiler warnings. 2015-11-14 10:57:59 +00:00
nicm
64333e3ef8 Be more strict about invalid UTF-8. 2015-11-14 10:56:31 +00:00
nicm
c56b81a2ce Push stdout and stderr to clients more aggressively, and add an event to
continue if the send fails.
2015-11-14 09:41:06 +00:00
nicm
908e6bb68f Log more of UTF-8 input. 2015-11-14 08:25:12 +00:00
nicm
c4893d8efd Log option names in fatal() for missing option. 2015-11-13 16:06:43 +00:00
nicm
88aa1c8dc3 Two spacing and spelling nits. 2015-11-13 12:18:52 +00:00
nicm
531869bd92 Add window_visible_layout which ignores zoomed panes and use it for
control mode (which needs to know all panes), from George Nachman.
2015-11-13 10:00:26 +00:00
nicm
c5689a5a40 Long overdue change to the way we store cells in the grid: now, instead
of storing a full grid_cell with UTF-8 data and everything, store a new
type grid_cell_entry. This can either be the cell itself (for ASCII
cells), or an offset into an extended array (per line) for UTF-8
data.

This avoid a large (8 byte) overhead on non-UTF-8 cells (by far the
majority for most users) without the complexity of the shadow array we
had before. Grid memory without any UTF-8 is about half.

The disadvantage that cells can no longer be modified in place and need
to be copied out of the grid and back but it turned out to be lot less
complicated than I expected.
2015-11-13 08:09:28 +00:00
nicm
e71a915412 Rename overly-long utf8data to ud throughout. 2015-11-12 22:04:37 +00:00
nicm
f2d03f4fdd grid_put_utf8 is unused, remove it. 2015-11-12 14:50:57 +00:00
nicm
a209ea3953 Add utf8_padcstr and use it to align columns in list-keys. 2015-11-12 12:43:36 +00:00
jmc
1da7475d0e tweak previous; ok nicm 2015-11-12 12:36:34 +00:00
nicm
d6daf37df4 Tidy utf8.c a little: build table on first use, and make utf8_width take
a u_int rather than splitting and then combining again in utf8_split.
2015-11-12 12:19:57 +00:00
nicm
0cc812ae34 tmux is UTF-8, so if $TMUX is set (tmux running in tmux), the client is
UTF-8. Also try to make the existing checks more readable.
2015-11-12 11:24:08 +00:00
nicm
c41673f3fa If we know the terminal outside tmux is not UTF-8, replace UTF-8 in
error messages and whatnot with underscores the same as we do when we
draw UTF-8 characters as part of the screen.
2015-11-12 11:10:50 +00:00
nicm
1b86f520ea Nuke the utf8 and status-utf8 options and make tmux only a UTF-8
terminal. We still support non-UTF-8 terminals outside tmux, but inside
it is always UTF-8 (as when the utf8 and status-utf8 options were on).
2015-11-12 11:09:11 +00:00
nicm
a0f3999ce7 Remove the mouse_utf8_flag format as well. 2015-11-12 11:07:10 +00:00
nicm
69e0b8326a Support UTF-8 key bindings by expanding the key type from int to
uint64_t and converting UTF-8 to Unicode on input and the reverse on
output. (This allows key bindings, there are still omissions - the
largest being that the various prompts do not accept UTF-8.)
2015-11-12 11:05:34 +00:00
nicm
7062b0e65d Default history-file should be "" not NULL, from Greg Onufe. 2015-11-12 08:19:18 +00:00
nicm
00c34df186 Drop mouse-utf8 option and always turn on UTF-8 mouse if the client says
it supports UTF-8.
2015-11-11 23:23:33 +00:00
nicm
6f3475c6c7 If realpath() fails just try the original path. 2015-11-10 22:33:47 +00:00
nicm
005e462e01 Handle absolute paths properly, and don't use resolved path in
realpath() fails.
2015-11-10 22:29:33 +00:00
nicm
dcdccf8333 Same bug as last commit, but in the other copy of the loop in this file... 2015-11-05 23:32:21 +00:00
schwarze
e9b58d9de4 Update the internal wcwidth(3) table of tmux(1) to match the data
in /usr/src/share/locale/ctype/en_US.UTF-8.src, with one single
exception:  Keep U+00AD SOFT HYPHEN at width 1 rather than moving
it to width 0, a tradition already observed in the old
https://www.cl.cam.ac.uk/~mgk25/ucs/wcwidth.c .
While here, manually rebalance the btree for optimal lookup speed.
OK nicm@
2015-11-05 16:44:25 +00:00
nicm
5577535891 Pass through right click if mouse is on, from Patrick Palka. 2015-11-05 11:05:30 +00:00
nicm
455284f1c0 Detach the client we are looping over, from Thomas Adam. 2015-11-03 15:07:36 +00:00
nicm
ba7fb49fb9 Fall back silently to ~ or / rather than checking -c with access(), this
was the old behaviour.
2015-10-31 14:51:15 +00:00
nicm
b0a99e85b6 Don't shift version out of peerid, it is needed later. 2015-10-31 13:43:38 +00:00
nicm
abb4e9e2fa The output log is only useful once and it means creating a file, so open
it once at startup instead of in every call to tty_open.
2015-10-31 13:12:03 +00:00
nicm
01defc9f49 Because pledge(2) does not allow us to pass directory file descriptors
around, we can't use file descriptors for the working directory because
we will be unable to pass it to a privileged process to tell it where to
read or write files or spawn children. So move tmux back to using
strings for the current working directory. We try to check it exists
with access() when it is set but ultimately fall back to ~ if it fails
at time of use (or / if that fails too).
2015-10-31 08:13:58 +00:00
nicm
45f3cea263 Break version check into a separate function, and limit version to 8 bits. 2015-10-29 09:35:31 +00:00
nicm
bf9c933cae Like options, move the environ struct into environ.c. 2015-10-28 09:51:55 +00:00
nicm
44657bf932 Move struct options into options.c. 2015-10-27 15:58:42 +00:00
nicm
67c3a014b9 No more TMPDIR. 2015-10-27 14:51:35 +00:00
nicm
07b0ea03c3 Break the common process set up, event loop and imsg dispatch code
between server and client out into a separate internal API. This will
make it easier to add another process.
2015-10-27 13:23:24 +00:00
nicm
9952201ca7 Count brackets in #{?...} so that nested conditional formats work, from
Daniel De Graaf.
2015-10-27 09:28:31 +00:00
nicm
17c2c4219d The format callback may not always succeed, so we need to check for
NULL. From Patrick Palka.
2015-10-27 09:18:06 +00:00
nicm
3fc001d0a2 Use copy-mode -et= in WheelUpPane binding, from Patrick Palka. 2015-10-27 09:15:21 +00:00
nicm
640c6fdd5f If a mouse event has no key binding, pass it through to the pane it
happened in, not the active pane like normal key presses. Fixes problems
seen by Enrico Ghirardi.
2015-10-26 23:16:18 +00:00
nicm
380a1ea8ef Default bindings for mouse wheel on status line to change window (like
we had before), from Patrick Palka.
2015-10-26 23:06:18 +00:00
nicm
b85be36d1c Handle unknown keys more gracefully, return a string instead of NULL. 2015-10-26 22:03:04 +00:00
nicm
a22fe33aa0 Some extra logging of where keys are actually going. 2015-10-26 17:17:06 +00:00
jmc
c582f7d177 space before punctuation; 2015-10-26 00:15:37 +00:00
nicm
e65306d8e7 Extend the modifiers allowed before formats: as well as the existing
#{=10:...}  length limit, add #{t:...} to convert a time_t format to a
string, #{b:...} for basename and #{d:...} for dirname. Remove all the
foo_string time formats as they can now be replaced by "t:", for example
#{window_activity_string} becomes #{t:window_activity}.
2015-10-25 22:29:17 +00:00