zsh-workers
 help / color / mirror / code / Atom feed
* [PATCH] Improve _htop
@ 2022-05-07  2:25 dana
  2022-05-08  0:07 ` Bart Schaefer
  0 siblings, 1 reply; 3+ messages in thread
From: dana @ 2022-05-07  2:25 UTC (permalink / raw)
  To: Zsh hackers list

[PATCH] Improve _htop

Someone submitted a PR on GitHub (#89) to fix htop's --version short-hand
amongst other things. There were a few trivial issues with the change as
well as some existing bugs/omissions that it didn't account for, so i said
i'd use the PR as a basis for a more elaborate improvement. Here's that
finally. With their permission i'll probably commit it in the original
author's name

Are we holding back completion improvements like this until after 5.9 is
released? (See also Marlon's patch from workers/49954)

Changes:

* Corrected -v to -V
* Made -u argument optional
* Enabled option stacking
* Improved descriptions (added ranges+defaults, &c.)
* Fixed broken sort keys with htop 3.x and added new key descriptions
* Removed _sequence limit on -p (it was incorrect on non-Linux systems,
  and making it correct seems like too much work for a limit that probably
  nobody will ever hit)
* Added several missing options

dana


diff --git a/Completion/Linux/Command/_htop b/Completion/Linux/Command/_htop
index 28c7512bf..6635181bc 100644
--- a/Completion/Linux/Command/_htop
+++ b/Completion/Linux/Command/_htop
@@ -1,11 +1,56 @@
-#compdef htop
-
-_arguments -S : \
-  '(-d --delay)'{-d+,--delay=}'[update frequency]:duration (tenths of seconds)' \
-  '(-C --no-color --no-colour)'{-C,--no-colo{,u}r}'[monochrome mode]' \
-  '(-)'{-h,--help}'[display usage information]' \
-  \*{-p+,--pid=}'[show given pids]: : _sequence -n ${$(</proc/sys/kernel/pid_max)\:-32768} _pids' \
-  '(-s --sort-key)'{-s+,--sort-key=}'[sort by key]:key:( ${(f)"$(_call_program sort-keys $words[1] --sort-key help)"} )' \
-  '(-t --tree)'{-t,--tree}'[show tree view of processes]' \
-  '(-u --user)'{-u+,--user=}'[show processes of user]: : _users' \
-  '(-)'{-v,--version}'[display version information]'
+#compdef htop pcp-htop
+
+# Notes:
+# - htop allows long options to be passed with a single dash; we don't account
+#   for this
+# - htop parses optional arguments to -H and -u 'cleverly' by allowing the next
+#   word to be the optarg if it doesn't begin with a '-'; we don't fully account
+#   for this
+# - There is a special version of htop designed to be used with PCP (Performance
+#   CoPilot); we don't fully account for this
+# - Some of the ranges and defaults listed here had to be found in the source
+
+local MATCH MBEGIN MEND ret=1
+local -a context line state state_descr args tmp
+
+args=(
+  '(-d --delay)'{-d+,--delay=}'[specify update frequency]:delay (tenths of seconds) (1-100) [15]'
+  '(-C --no-color --no-colour)'{-C,--no-colo{,u}r}'[use monochrome colour scheme]'
+  '(-F --filter)'{-F+,--filter=}'[show only commands matching specified filter]:case-insensitive command-line sub-string:_process_names -a'
+  '(-)'{-h,--help}'[display usage information]'
+  '(-H --highlight-changes)'{-H+,--highlight-changes=}'[highlight new and old processes (optionally specify delay)]::delay (seconds) (1-86400) [5]'
+  '(-M --no-mouse)'{-M,--no-mouse}'[disable mouse]'
+  \*{-p+,--pid=}'[show only specified PIDs]: : _sequence _pids'
+  '--readonly[disable all system and process changing features]'
+  '(-s --sort-key)'{-s+,--sort-key=}'[sort by specified column]: :->sort-keys'
+  '(-t --tree)'{-t,--tree}'[show tree view of processes]'
+  '(-u --user)'{-u+,--user=}'[show only processes of current or specified user]:: : _users'
+  '(-U --no-unicode)'{-U,--no-unicode}'[disable Unicode]'
+  '(-)'{-V,--version}'[display version information]'
+)
+
+[[ $OSTYPE == linux* ]] &&
+(( ! EUID || $+_comp_priv_prefix )) &&
+_pick_variant libcap=drop-capabilities $OSTYPE --help &&
+args+=(
+  '--drop-capabilities=-[drop specified capabilties]::mode [basic]:((
+    off\:"do not drop capabilities"
+    basic\:"drop capabilities not needed for standard functionality (retains kill, renice, etc.)"
+    strict\:"drop capabilities not needed for core functionality"
+  ))'
+)
+
+_arguments -s -S : $args && ret=0
+
+case $state in
+  sort-keys)
+    tmp=( ${(f)"$(_call_program sort-keys $words[1] --sort-key help)"} )
+    tmp=( ${tmp/#[[:space:]]##} )
+    tmp=( ${tmp//:/\\:} )
+    tmp=( ${tmp/[[:space:]]##/:} )
+    tmp=( ${tmp/(#m):[A-Z]/${(L)MATCH}} )
+    _describe -t sort-keys 'column (key)' tmp && ret=0
+    ;;
+esac
+
+return ret


^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [PATCH] Improve _htop
  2022-05-07  2:25 [PATCH] Improve _htop dana
@ 2022-05-08  0:07 ` Bart Schaefer
  2022-05-08  3:54   ` dana
  0 siblings, 1 reply; 3+ messages in thread
From: Bart Schaefer @ 2022-05-08  0:07 UTC (permalink / raw)
  To: dana; +Cc: Zsh hackers list

On Fri, May 6, 2022 at 7:26 PM dana <dana@dana.is> wrote:
>
> [PATCH] Improve _htop
>
> Are we holding back completion improvements like this until after 5.9 is
> released? (See also Marlon's patch from workers/49954)

I didn't hear any objections to including that, just your review of
possible further improvements.

There's also that tiny fix to _values that I just posted in 50184.

Since you're the person who will be re-rolling everything, and none of
these things are critical path, I think you can make the call.  I have
a commit ready to push if OK.

And then we should try to do the release before there's too much more
incentive to fiddle with it!


^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [PATCH] Improve _htop
  2022-05-08  0:07 ` Bart Schaefer
@ 2022-05-08  3:54   ` dana
  0 siblings, 0 replies; 3+ messages in thread
From: dana @ 2022-05-08  3:54 UTC (permalink / raw)
  To: Bart Schaefer; +Cc: Zsh hackers list

On Sat 7 May 2022, at 19:07, Bart Schaefer wrote:
> Since you're the person who will be re-rolling everything, and none of
> these things are critical path, I think you can make the call.  I have
> a commit ready to push if OK.
>
> And then we should try to do the release before there's too much more
> incentive to fiddle with it!

I doubt _gradle or _htop would have got much testing anyway, and _values
is riskier but it seems obviously wrong the way it is. So yeah let's put
those in i guess

If there are no other concerns i'll probably release next week end

dana


^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2022-05-08  3:56 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-05-07  2:25 [PATCH] Improve _htop dana
2022-05-08  0:07 ` Bart Schaefer
2022-05-08  3:54   ` dana

Code repositories for project(s) associated with this 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).