When this option is used, the child process in the new terminal
instance will inherit its environment from the footclient process,
instead of the foot server’s.
Implemented by sending (yet another) dynamic string list as part of
the client -> server setup packet. When the new option is *not* used,
the setup packet is now 2 bytes larger than before.
On the server side, the slave process now uses execvpe() instead of
execvp(). There’s plumbing to propagate a new ‘envp’ argument from
term_init() all the way down to slave_exec(). If ‘envp’ is NULL, we
use ‘environ’ instead (thus matching the old behavior of execvp()).
Closes#1004
Before this, foot always emitted a CSI if _any_ modifier was
active. XTerm doesn’t behave quite like this. Instead, it appears to
special-case shift somewhat:
If the generated symbol (e.g ‘A’) is the upper case version of the
pressed key’s base symbol (‘a’), and is in the range a-z, emit a CSI.
If not emit a plain symbol:
Examples (Swedish layout):
* shift-a (generated symbol is ‘A’) emits a CSI
* shift-, (generated symbol is ‘;’) emits ‘;’
* shift-alt-, (generated symbol is ‘;’) emits a CSI
Closes#1009
Checking for specific strings of length 0 or 1 can be done with e.g.:
if (str[0] == '\0') {...}
if (str[0] == '?' && str[1] == '\0') {...}
Doing it this way instead of using strlen(3) means we avoid the
function call overhead and also avoid scanning through more of the
string than is neceessary.
A compiler could perhaps optimize away calls to strlen(3) when the
result is compared to a small constant, but GCC 11.2 only seems to
actually do this[1] for lengths of 0.
[1]: https://godbolt.org/z/qxoW8qqW6
The underline cursor is positioned just below regular underlines. A
bug in the positioning logic related to this, sometimes resulted in
the cursor being thinner than what it should be, or even invisible.
Fixes#1005
This makes the example config `foot.ini` and its man page slightly more
coherent regarding the specification of default values.
Note that the cursor color is not hardcoded like e.g. foreground or
background, thus in the example config, `<inverse foreground/background>`
makes more sense.
If a request handler doesn’t define its own put() handler, simply drop
all DCS parameter data. This way, we won’t end up allocating
potentially large buffers for unsupported/unimplemented DCS requests.
Closes#959
When matching “untranslated” bindings (by matching the base symbol of
the key, e.g. ctrl+shift+2 in US layout), require that no
non-significant modifiers are active.
This fixes an issue where AltGr was “ignored”, and would cause certain
combinations to match a key binding.
Example: ctrl+altgr+0, on many European layouts matched against the
default ctrl+0 (reset the font size), instead of emitting ^]
To make this work, we now need to filter out “locked”
modifiers (e.g. NumLock and CapsLock). Otherwise having e.g. NumLock
active would prevent *all* untranslated matching to fail.
Closes#983
Use the (relatively new) macro __has_include() to check if
stdc-predef.h exists, and only include it if it does.
If stdc-predef.h does not exist, or if the compiler does not implement
__has_include(), stdc-predef.h is *not* included.
In non-bracketed paste mode, we translate \n to \r, and \r\n to
\r. The latter matches XTerm, urxvt, alacritty and kitty. The former
matches alacritty and kitty (xterm and urxvt just blindly replaces all
\n occurrences with \r, meaning \r\n is translated to \r\r.
For some reason, we then unconditionally translated all \r back to \n,
regardless of whether bracketed paste was enabled or not. Unsure
why/where this comes from, but it doesn't match any of the other
terminal emulators I tested.
One example where this caused issues is in older versions of nano (at
least up to 2.9).
Closes#980
This could be triggered by running, for example:
printf '\033P$qrxyz\033\\'
...which would cause a memcmp() of 4 bytes on a 2 byte string literal.
Fixes: #960
With fixed increments of 128 bytes, an OSC 52 copy operation could end
up doing thousands or tens of thousands of realloc(3) calls just to
copy a few MB.
- Support for multiples concurrent instances, templated on
WAYLAND_DISPLAY
- Use standard input for the socket file descriptor (inetd style)
- Always use the socket provided by systemd with the systemd user
service
- wayland-instance@.target is intended to be a special target a bit like
graphical-session.target.
If the argument to --server is parsed as a number, consider it to be a
file descriptor, and use that as a socket.
This is necessary to be able to use socket activation with the server
mode of foot.
