| Age | Commit message (Collapse) | Author |
|---|
|
|
|
|
|
|
|
|
|
|
|
|
|
In other words, warn when a function is missing static. I don't see
why this isn't in -Wextra.
|
|
|
|
I know, it feels wrong.
|
|
Losing tab complete and text macros, for now.
This new implementation works on an instance of a struct and does
not interact with the rest of catgirl, making it possible to copy
into another project. Unlike existing line editing libraries, this
one is entirely abstract and can be rendered externally.
My goal with this library is to be able to implement vi mode. Since
it operates on struct instances rather than globals, it might also
be possible to give catgirl separate line editing buffers for each
window, which would be a nice UX improvement.
|
|
Do some extra work by adding the portion before the cursor to the
input window twice, but simplify the interaction with the split
point. This fixes the awkward behaviour when moving the cursor
across colour codes where the code would be partially interpreted
up to the cursor.
|
|
|
|
|
|
|
|
We don't use any flags, so just define them in the compat source
file.
|
|
At 5d08070de6aadd979e3752263d2ec1e709bf037e.
|
|
|
|
This matches behaviour of, e.g. zsh -o emacs.
|
|
|
|
|
|
Woops! BASE64_SIZE is the size of the string buffer. Somehow ergo
is the only server software (that I know of) to reject the accidental
null byte.
Reported by smlavine.
|
|
So that they can be hidden with M-+.
|
|
|
|
When cross-compiling, it's common to have executables prefixed with
the name of the architecture you're building for,
e.g. aarch64-unknown-linux-musl-cc or x86_64-unknown-freebsd-pkg-config.
Lots of build tools support a PKG_CONFIG environment variable to
enable this use case.
With this change, I was able to successfully cross-compile and run
catgirl.
|
|
Allows completing a nick at the beginning of a message without a
colon by continuing to press tab, as well as after another nick
already followed by a colon without turning it into a comma-separated
list of nicks all followed by a colon. For example, tab can be used
to cycle between the following pairs:
nick1: |
nick1 |
nick1, nick2: |
nick1: nick2 |
|
|
How did this normal sounding format string get in there!
|
|
<https://modern.ircdocs.horse/index.html#rplwhoisspecial-320>
|
|
Strip formatting when calculating the timestamp width to avoid
moving a bunch of code around. Use styleAdd (now with an initial
style parameter) to show timestamps.
This allows changing the style of the timestamps from the default
gray using literal IRC formatting codes in the string. Not ideal,
but no new options needed.
Suggested by Hoël Bézier and Sebastian LaVine.
|
|
POSIX does not define a %F for strptime[1], but does define %F for
strftime[2]. Afaik most libc's implement %F for both, but musl is very
standards-compliant and does not have %F on strptime, leading to
unparsed message tag times, which causes all backlog sent from bouncer
on startup to have a timestamp of the current time, instead of the
actual timestamp sent.
[1] https://pubs.opengroup.org/onlinepubs/9699919799/functions/strptime.html
[2] https://pubs.opengroup.org/onlinepubs/9699919799/functions/strftime.html
|
|
Otherwise only /ban can be used to list bans and /mode b or /mode
+b won't show the listing.
|
|
|
|
|
|
catgirl correctly handles receiving "self-messages". pounce always
sends them.
[1]: https://wiki.znc.in/Query_buffers
[2]: https://defs.ircdocs.horse/info/selfmessages.html
|
|
Only the first colon should be replaced with a null byte.
Ported from pounce.
|
|
This fixes the case where an IRCd does not normalize channel names,
e.g. PRIVMSG #TEST is relayed as-is, rather than as #test or whatever
the canonical casing of the channel name is. It also fixes the case
of opening a query window with incorrect case, e.g. /query nickserv.
However, this solution is only completely correct when
CASEMAPPING=ascii.[1] I do not think the extra mappings of
CASEMAPPING=rfc1459 are relevant enough to justify adding the code
to handle it.
[1]: https://modern.ircdocs.horse/#casemapping-parameter
|
|
How embarrassing.
|
|
Otherwise a tag with no value would cause a segfault trying to
unescape the NULL tag pointer. This shouldn't happen for the server
tags we parse, but clients could send @+draft/reply with no value.
|
|
On OpenBSD it's in <fcntl.h>, and it gets declared anyway on FreeBSD
and macOS. Curiously, on GNU/Linux, LOCK_* are defined, but flock(2)
isn't declared.
|
|
Wouldn't happen anyway since configPath will always return at least
one path.
|
|
|
|
|
|
caph_enter(3) is the same as cap_enter(2) except that it returns
success even if the kernel does not support capability mode. Since
we only enter capability mode when explicitly requested by the
restrict option, it should fail loudly if it is not supported. On
the other hand, we make calls to caph_rights_limit(3) and friends
in some places regardless of whether we actually enter capability
mode (to keep the code simple), so those should continue to succeed
even if capability mode is not supported.
|
|
Messages don't really need to be hidden from <network> and I think
it could be confusing. Debug messages are all Cold so everything
would be hidden, and I want to keep them that way so that <debug>
doesn't clutter the status line needlessly.
|
|
Silencing all windows with `M-+' (across multiple catgirl instances)
can be cumbersome, so provide an option to hide events, JOIN/PART noise,
etc. by default (each window's threshold will persist across load/save
cycles, i.e. when using the `-s/save' option).
Started out as `-v | visibility = threshold' to set a specific level,
the idea of a simpler toggle comes from june, who also squashed other
bugs (as usual).
|
|
"base-index" expects integer values, tmux prints a warning at load-time
but otherwise ignores the configuration line.
|
|
Follow a79a3fc "Use NS and CS server aliases".
|
|
To keep the "main" sequence of events on one screen, while emphasizing
that sandboxing happens either side of ircConnect().
|
|
|
|
ircConnect() yields a connected TCP socket after which "inet dns" is
no longer needed.
Possibly having loaded private key material, it seems a tad more
comforting to speak TLS *after* dropping any network capabilities
(except for socket read/write to the IRC host, of course).
Instead of moving the final pledge into irc.c:ircConnect() and thus
complicating the code around pledge across two C modules, simply
stub out an mnemonic ircHandshake() and call that explicitly.
This restores behaviour gained with
981ebc4 "Remove explicit tls_handshake(3) from ircConnect" which
was reverted for other reasons.
|
|
No need to keep them at runtime; do so unconditionally for the sake of
simplicity.
Declare TLS config globally so ircConnect() can clear it and declare
both client and config statically as they are not used outside the irc.c
module.
|
