* [PATCH] rewrite _killall
@ 2018-06-28 7:39 Jun T
2018-06-28 13:10 ` Jun T.
2018-06-28 14:52 ` dana
0 siblings, 2 replies; 6+ messages in thread
From: Jun T @ 2018-06-28 7:39 UTC (permalink / raw)
To: zsh-workers
_killall has many options which need be completed.
The previous patch to _ttys is required.
diff --git a/Completion/Unix/Command/_killall b/Completion/Unix/Command/_killall
index 375b1bf74..989772c5d 100644
--- a/Completion/Unix/Command/_killall
+++ b/Completion/Unix/Command/_killall
@@ -1,9 +1,77 @@
#compdef killall killall5
-if [[ $service = killall && "$OSTYPE" = (linux*|*bsd*|darwin*) ]]; then
- _alternative \
- 'signals:: _signals -p' \
- 'processes-names:process:{ compadd "$expl[@]" ${${${${(f)"$(_call_program processes-names ps ${${EUID/(#s)0(#e)/xa}//[0-9]#/}ho command 2> /dev/null)"//[][\(\)]/}:#(ps|COMMAND|-*)}%%\ *}:t} }'
-else
+if [[ $service = killall5 || $OSTYPE != (linux|*bsd|dragonfly|darwin)* ]]; then
_signals -p
+ return
+fi
+
+typeset -a opts args
+[[ $EUID = 0 || $_comp_priv_prefix[1] = sudo ]] && opts=( -a )
+
+if _pick_variant psmisc=PSmisc unix --version; then
+
+ local curcontext=$curcontext state state_descr line ret=1
+ typeset -A opt_args
+ args=(
+ '(-e --exact)'{-e,--exact}'[require exact match for names longer than 15 chars]'
+ '(-I --ignore-case)'{-I,--ignore-case}'[do case insensitive process name match]'
+ '(-g --process-group)'{-g,--process-group}'[kill the process group to which the process blongs]'
+ '(-i --interactive)'{-i,--interactive}'[interactively ask for confirmation before killing]'
+ '(- : *)'{-l,--list}'[list all known signal names]'
+ '(-n --ns)'{-n,--ns}'[match against the PID namespace of the given PID]:pid: _pids'
+ '(-o --older-than)'{-o,--older-than}'[match only processes older than the specified time]:time:->time'
+ '(-q --quiet)'{-q,--quiet}'[do not complain if no processes were killed]'
+ '(-r --regexp)'{-r,--regexp}'[interpret process name as extended regular expression]'
+ '(-s --signal)'{-s,--signal}'[send the speified signal]:signal:_signals'
+ '(-u --user)'{-u,--user}'[kill only processes owned by specified user]: : _users'
+ '(-v --verbose)'{-v,--verbose}'[report if the signal was successfully sent]'
+ '(-w --wait)'{-w,--wait}'[wait for all killed processes to die]'
+ '(- : *)'{-V,--version}'[display version information]'
+ '(-y --younger-than)'{-y,--younger-than}'[match only processes younger than the specified time]:time:->time'
+ "1: : _alternative 'signals: :_signals -p' 'processes-names: :_process_names $opts'"
+ "*: :_process_names $opts"
+ )
+ [[ $CURRENT = 2 || ( $CURRENT = 3 && $words[2] = (-Z|--context) ) ]] && \
+ args+=( '(-Z --context)'{-Z,--context}'[specify SELinux security context]:regex pattern: ' )
+
+ _arguments -s -S -C : $args && ret=0
+
+ case $state in
+ (time)
+ local -a units=( 's:seconds' 'm:minuts' 'h:hours' 'd:days'
+ 'w:weeks' 'M:months' 'y:years' )
+ if compset -P '[0-9]##(|.[0-9]#)'; then
+ _alternative 'float-number:: _message "float number"' \
+ 'unit:: _describe unit units' && ret=0
+ else
+ _message 'float number and unit' && ret=0
+ fi
+ ;;
+ esac
+
+ return ret
+
+else # bsd and darwin
+
+ if [[ $OSTYPE == (freebsd|dragonfly)* ]]; then
+ args=( '-j[kill processes in specified jail]: :_jails -0' )
+ opts+=( -t ) # long process names must be truncated
+ fi
+ args+=(
+ '-v[be more verbose about what will be done]'
+ '-e[use effective user ID for -u option]'
+ '(- : *)-help[give a help on command usage and exit]'
+ '(- : *)-l[list names of available signals and exit]'
+ '-m[interprit specified name as a regular expression]'
+ '-s[show only what will be done, but do notsend any signal]'
+ '-d[print info about processes matched, but do not send any signal]'
+ '-u[limit to processes belonging to specified user]: :_users'
+ '-t[limit to processes running on specified tty]: :_ttys -od'
+ "-c[with -u or -t, limit to processes matching specified name]: :_process_names $opts"
+ '-z[do not skip zombies]'
+ "1: : _alternative 'signals::_signals -p' 'processes-names::_process_names $opts'"
+ "*: :_process_names $opts"
+ )
+ _arguments -s -S : $args
+
fi
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH] rewrite _killall
2018-06-28 7:39 [PATCH] rewrite _killall Jun T
@ 2018-06-28 13:10 ` Jun T.
2018-06-28 14:52 ` dana
1 sibling, 0 replies; 6+ messages in thread
From: Jun T. @ 2018-06-28 13:10 UTC (permalink / raw)
To: zsh-workers
> 2018/06/28 16:39、Jun T <takimoto-j@kba.biglobe.ne.jp>のメール:
>
> + '(-s --signal)'{-s,--signal}'[send the speified signal]:signal:_signals'
This line must be
'(1 -s --signal)'{-s,--signal}'[send the speified signal]:signal:_signals'
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH] rewrite _killall
2018-06-28 7:39 [PATCH] rewrite _killall Jun T
2018-06-28 13:10 ` Jun T.
@ 2018-06-28 14:52 ` dana
2018-06-29 11:45 ` Jun T
2018-07-01 16:17 ` Jun T.
1 sibling, 2 replies; 6+ messages in thread
From: dana @ 2018-06-28 14:52 UTC (permalink / raw)
To: Jun T; +Cc: zsh-workers
On 28 Jun 2018, at 02:39, Jun T <takimoto-j@kba.biglobe.ne.jp> wrote:
>+if _pick_variant psmisc=PSmisc unix --version; then
Is this OK? I know it's checking for killall5/Solaris early on, but is there any
possibility that a SysV-style killall could slip past that check? Maybe that's
overly paranoid, idk.
On 28 Jun 2018, at 02:39, Jun T <takimoto-j@kba.biglobe.ne.jp> wrote:
>+ '(-n --ns)'{-n,--ns}'[match against the PID namespace of the given PID]:pid: _pids'
Trivial thing, but it seems preferable to let _pids, _signals, &c., handle their
descriptions unless there's a specific need to override them, yeah?
On 28 Jun 2018, at 02:39, Jun T <takimoto-j@kba.biglobe.ne.jp> wrote:
>+ '(-g --process-group)'{-g,--process-group}'[kill the process group to which the process blongs]'
>+ '(-s --signal)'{-s,--signal}'[send the speified signal]:signal:_signals'
>+ '-m[interprit specified name as a regular expression]'
>+ '-s[show only what will be done, but do notsend any signal]'
Some typos here:
* blongs -> belongs
* speified -> specified
* interprit -> interpret
* notsend -> not send
On 28 Jun 2018, at 02:39, Jun T <takimoto-j@kba.biglobe.ne.jp> wrote:
>'(-u --user)'{-u,--user}'[kill only processes owned by specified user]: : _users'
>+ '-u[limit to processes belonging to specified user]: :_users'
These should all be in -o+ and --opt= form, i'd think.
On 28 Jun 2018, at 02:39, Jun T <takimoto-j@kba.biglobe.ne.jp> wrote:
>+ _arguments -s -S : $args
`-A '-*'` would be appropriate for this one.
dana
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH] rewrite _killall
2018-06-28 14:52 ` dana
@ 2018-06-29 11:45 ` Jun T
2018-07-01 16:17 ` Jun T.
1 sibling, 0 replies; 6+ messages in thread
From: Jun T @ 2018-06-29 11:45 UTC (permalink / raw)
To: zsh-workers
Thank you for the comments.
> 2018/06/28 23:52、dana <dana@dana.is>のメール:
>
> On 28 Jun 2018, at 02:39, Jun T <takimoto-j@kba.biglobe.ne.jp> wrote:
>> +if _pick_variant psmisc=PSmisc unix --version; then
>
> Is this OK? I know it's checking for killall5/Solaris early on, but is there any
> possibility that a SysV-style killall could slip past that check? Maybe that's
> overly paranoid, idk.
I guess on virtually all Linux distros killall comes from psmisc, and if it is
ever not psmics, I think the chances are very low that it is a SysV-variant.
> On 28 Jun 2018, at 02:39, Jun T <takimoto-j@kba.biglobe.ne.jp> wrote:
>> + '(-n --ns)'{-n,--ns}'[match against the PID namespace of the given PID]:pid: _pids'
>
> Trivial thing, but it seems preferable to let _pids, _signals, &c., handle their
> descriptions unless there's a specific need to override them, yeah?
The :pid: was doing nothing due to the space before _pids.
I will remove it.
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH] rewrite _killall
2018-06-28 14:52 ` dana
2018-06-29 11:45 ` Jun T
@ 2018-07-01 16:17 ` Jun T.
2018-07-01 19:14 ` dana
1 sibling, 1 reply; 6+ messages in thread
From: Jun T. @ 2018-07-01 16:17 UTC (permalink / raw)
To: zsh-workers
> 2018/06/28 23:52、dana <dana@dana.is>のメール:
> On 28 Jun 2018, at 02:39, Jun T <takimoto-j@kba.biglobe.ne.jp> wrote:
>> + _arguments -s -S : $args
>
> `-A '-*'` would be appropriate for this one.
I didn't add this because "killall -SIGNAL -<TAB>" does not complete
options if -A '-*' is included.
If we do not want to complete options after process names, something
like the following (patch to my previous patch) may be necessary.
A problem of this is now users can't customize by tag-order.
diff --git a/Completion/Unix/Command/_killall b/Completion/Unix/Command/_killall
index a62181c25..d603e6e4a 100644
--- a/Completion/Unix/Command/_killall
+++ b/Completion/Unix/Command/_killall
@@ -69,9 +69,9 @@ else # bsd and darwin
'-t[limit to processes running on specified tty]: :_ttys -od'
"-c[with -u or -t, limit to processes matching specified name]: :_process_names $opts"
'-z[do not skip zombies]'
- "1: : _alternative 'signals::_signals -p' 'processes-names::_process_names $opts'"
"*: :_process_names $opts"
+ + '(signal)' -$signals[2,-3]
)
- _arguments -s -S : $args
+ _arguments -s -S -A '-*' : $args
fi
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH] rewrite _killall
2018-07-01 16:17 ` Jun T.
@ 2018-07-01 19:14 ` dana
0 siblings, 0 replies; 6+ messages in thread
From: dana @ 2018-07-01 19:14 UTC (permalink / raw)
To: Jun T.; +Cc: zsh-workers
On 1 Jul 2018, at 11:17, Jun T. <takimoto-j@kba.biglobe.ne.jp> wrote:
>I didn't add this because "killall -SIGNAL -<TAB>" does not complete
>options if -A '-*' is included.
Oh, yeah. That function was problematic for me when i was messing with _pgrep as
well; i ended up just not using it. I don't know, if hacking around it
introduces other problems i'm sure it's fine to leave it until a nicer solution
is found. I was just being pedantic anyway
dana
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2018-07-01 19:14 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2018-06-28 7:39 [PATCH] rewrite _killall Jun T
2018-06-28 13:10 ` Jun T.
2018-06-28 14:52 ` dana
2018-06-29 11:45 ` Jun T
2018-07-01 16:17 ` Jun T.
2018-07-01 19:14 ` dana
Code repositories for project(s) associated with this public inbox
https://git.vuxu.org/mirror/zsh/
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).