Re: UPDATE: TeX Live 2019

[Jon Bernard]

* Edd Barrett  wrote:
> On Sun, Oct 20, 2019 at 10:37:48AM +0100, Edd Barrett wrote:
> > On Sun, Oct 20, 2019 at 10:35:05AM +0100, Edd Barrett wrote:
> > > Attached is a mega-diff to update TeX Live.
> > 
> > Forgot to say: I've tested this with a partial bulk on yesterday's
> > snapshot. No issues.
> 
> Has anyone had a chance to test this?

Works here, I haven't noticed any regressions.  Thanks Edd.

-- 
Jon

Re: NEW: devel/leiningen - Clojure project automation

[Jon Bernard]

* Klemens Nanni  wrote:
> This is a small port that merely fetches a bash script and a zip archive
> aka. jar file from GitHub and packages it:
> 
>   Information for inst:leiningen-2.9.1
> 
>   Comment:
>   Clojure project automation
> 
>   Description:
>   Leiningen is for automating Clojure projects without setting your hair 
> on fire.
> 
>   Maintainer: The OpenBSD ports mailing-list 
> 
>   WWW: https://leiningen.org
> 
> 
> I have no clue of Java or Clojure but with that package, I can do
> `lein repl' and hack away^W^Wfollow instructions on fancy clojure
> projects out there.
> 
> Portwise, I'm pretty much doing what FreeBSD[0] does execept
> 
> - using more recent Java
> - avoiding ugly sed(1) fiddling in post-patch
> - installing upstream documentation
> 
> Feedback? OK?

I submitted this some time back [1], and here's what I maintain
personally [2].  I think the shell completions should be named to match
the port to prevent possible collisions.  Otherwise it looks good to me.

[1] http://openbsd-archive.7691.n7.nabble.com/NEW-devel-leiningen-td321667.html
[2] https://git.jbernard.io/mystuff/tree/devel/leiningen

-- 
Jon

Re: UPDATE: Neovim-0.3.4

[Jon Bernard]

* Edd Barrett  wrote:
> Hi,
> 
> I tried to post this the other day, but it seems to have gotten lost.
> 
> Here's a diff to the latest neovim and Python bindings.
> 
>  * There's one small test failure. I don't think it's a blocker.
> 
>  * The Python bindings have been renamed upstream. If someone cares
>enough, in a later commit we can rename the port and use @pkgpath.
> 
> Comments? OK?

FWIW I've been using this for a week and all is well for me.  The
getpgid() patch also successfully silences the errors I see from time to
time.  I don't run anything that makes use of the python bindings, but
the main update appears solid.

-- 
Jon

> 
> 
> Index: neovim//Makefile
> ===
> RCS file: /cvs/ports/editors/neovim/Makefile,v
> retrieving revision 1.13
> diff -u -p -r1.13 Makefile
> --- neovim//Makefile  4 Mar 2019 14:22:25 -   1.13
> +++ neovim//Makefile  14 Mar 2019 11:04:10 -
> @@ -4,8 +4,7 @@ COMMENT = continuation and extension of 
>  
>  GH_ACCOUNT = neovim
>  GH_PROJECT = neovim
> -GH_TAGNAME = v0.3.1
> -REVISION =   2
> +GH_TAGNAME = v0.3.4
>  
>  CATEGORIES = editors devel
>  HOMEPAGE =   http://neovim.org
> @@ -54,6 +53,9 @@ TEST_DEPENDS =  shells/bash \
>  
>  # These are the "old tests". There is also a new suite, but we would need the
>  # "busted" test suite for Lua, which is not yet ported.
> +#
> +# There is currently one (minor) test failure:
> +# https://github.com/neovim/neovim/issues/9704
>  do-test:
>   cd ${WRKSRC} && env LC_CTYPE=en_US.UTF-8 ${MAKE_PROGRAM} \
>   -C src/nvim/testdir NVIM_PRG=${WRKBUILD}/bin/nvim ${MAKE_FLAGS}
> Index: neovim//distinfo
> ===
> RCS file: /cvs/ports/editors/neovim/distinfo,v
> retrieving revision 1.4
> diff -u -p -r1.4 distinfo
> --- neovim//distinfo  23 Aug 2018 09:11:17 -  1.4
> +++ neovim//distinfo  9 Mar 2019 12:28:56 -
> @@ -1,2 +1,2 @@
> -SHA256 (neovim-0.3.1.tar.gz) = vF45LUwHZAeQbM7LwoPhpEt4MsL0hsrYGqBMwplzrSI=
> -SIZE (neovim-0.3.1.tar.gz) = 8937900
> +SHA256 (neovim-0.3.4.tar.gz) = pkEQi96/rzGYRO1Gsb811vfDDvWurespugbhnDJ0vA4=
> +SIZE (neovim-0.3.4.tar.gz) = 9231575
> Index: neovim//patches/patch-src_nvim_CMakeLists_txt
> ===
> RCS file: /cvs/ports/editors/neovim/patches/patch-src_nvim_CMakeLists_txt,v
> retrieving revision 1.1
> diff -u -p -r1.1 patch-src_nvim_CMakeLists_txt
> --- neovim//patches/patch-src_nvim_CMakeLists_txt 4 Mar 2019 14:22:25 
> -   1.1
> +++ neovim//patches/patch-src_nvim_CMakeLists_txt 10 Mar 2019 17:33:18 
> -
> @@ -6,7 +6,7 @@ luajit (we use lua instead).
>  Index: src/nvim/CMakeLists.txt
>  --- src/nvim/CMakeLists.txt.orig
>  +++ src/nvim/CMakeLists.txt
> -@@ -359,10 +359,6 @@ endforeach()
> +@@ -362,10 +362,6 @@ endforeach()
>   
>   # Our dependencies come first.
>   
> Index: neovim//patches/patch-src_nvim_os_process_c
> ===
> RCS file: neovim//patches/patch-src_nvim_os_process_c
> diff -N neovim//patches/patch-src_nvim_os_process_c
> --- /dev/null 1 Jan 1970 00:00:00 -
> +++ neovim//patches/patch-src_nvim_os_process_c   10 Mar 2019 20:16:36 
> -
> @@ -0,0 +1,40 @@
> +$OpenBSD$
> +
> +Port job stopping bug fix to neovim:
> +
> +Original vim fix:
> +https://github.com/vim/vim/commit/76ab4fd61901090e6af3451ca6c5ca0fc370571f#diff-b68adb4fa34020d8d7f0ab40a2704335
> +
> +Index: src/nvim/os/process.c
> +--- src/nvim/os/process.c.orig
>  src/nvim/os/process.c
> +@@ -89,21 +89,16 @@ bool os_proc_tree_kill(int pid, int sig)
> + bool os_proc_tree_kill(int pid, int sig)
> + {
> +   assert(sig == SIGTERM || sig == SIGKILL);
> +-  int pgid = getpgid(pid);
> +-  if (pgid > 0) {  // Ignore error. Never kill self (pid=0).
> +-if (pgid == pid) {
> +-  ILOG("sending %s to process group: -%d",
> +-   sig == SIGTERM ? "SIGTERM" : "SIGKILL", pgid);
> +-  int rv = uv_kill(-pgid, sig);
> +-  return rv == 0;
> +-} else {
> +-  // Should never happen, because process_spawn() did setsid() in the 
> child.
> +-  ELOG("pgid %d != pid %d", pgid, pid);
> +-}
> ++  if (pid != 0) {  // Never kill self (pid=0).
> ++ILOG("sending %s to PIDs %d and %d",
> ++  sig == SIGTERM ? "SIGTERM" : "SIGKILL", -pid, pid);
> ++
> ++int rv1 = uv_kill(-pid, sig);
> ++int rv2 = uv_kill(pid, sig);
> ++return (rv1 == 0) && (rv2 == 0);
> +   } else {
> +-ELOG("getpgid(%d) returned %d", pid, pgid);
> ++return false;
> +   }
> +-  return false;
> + }
> + #endif
> + 
> Index: neovim//pkg/PLIST
> ===
> RCS file: /cvs/ports/editors/neovim/pkg/PLIST,v
> retrieving revision 1.6
> diff -u -p -r1.6 PLIST
> --- neovim//pkg/PLIST 4 Sep 2018 12:46:11 -   1.6
> +++ neovim//pkg/PLIST 9 Mar 2019 12:28:56 -

Re: UPDATE: TeX Live 2018

[Jon Bernard]

* Edd Barrett  wrote:
> Hi,
> 
> Here's a first go at TeX Live 2018.
> 
> I've tested it on only amd64 at this point, but all ports BUILD_DEPEND
> or LIB_DEPENDing on any of TeX Live have been build tested.
> 
> Please test your documents!

I tried several documents including an IEEEtran draft and everything
rendered as expected.  For my current use cases everything looks good
here.  Thanks for working on this.

-- 
Jon

> Stuart, is this still needed?
> https://cvsweb.openbsd.org/cgi-bin/cvsweb/ports/print/texlive/base/Makefile.diff?r1=1.103=1.104=h
> 
> If it is, can we make it a bit more verbose, because I'm not sure what
> it means :)
> 
> Cheers
> 
> -- 
> Best Regards
> Edd Barrett
> 
> http://www.theunixzoo.co.uk

Re: [update] editors/neovim 0.3.1

[Jon Bernard]

* Jon Bernard  wrote:
> Hello,
> 
> Here is an update to neovim for version 0.3.1.  The new release contains
> mostly bugfixes and nothing major.  As such, the diff is just a version
> bump and everything else remains the same.

If there's anything I can do to make this easier on reviewers please let
me know.  Here's the changelog in case it helps, I'll include this in
future updates moving forward.

FEATURES:
* man.vim: C highlighting for EXAMPLES section
* TUI: urxvt: also send xterm focus-reporting seqs
* API: emit nvim_buf_lines_event from :terminal
* fillchars: Add "eob" flag

FIXES:
* startup: avoid blank stdin buffer if other files were opened
* Only waitpid() for processes that we care about
* Check all child processes for exit in SIGCHLD handler
* channel.c: Prevent channel_destroy_early() from freeing uninitialized rpc 
stuff
* transstr_buf: fix length comparison
* TUI: Fix standout mode
* man.vim: fix for mandoc
* provider/node: npm --loglevel silent
* API: nvim_win_set_cursor: set curswant
* checkhealth: Python: fix VIRTUAL_ENV check
* checkhealth: node.js: also search yarn

CHANGES:
* defaults: shortmess+=F
* highlight: high-priority CursorLine if fg is set.
* terminal: handle  and :confirm on unloading
* screen: truncate showmode messages
* buffer: fix copying :setlocal options
* Ex mode: always "improved" (gQ)
* options: remove 'maxcombine` option (always 6)

INTERNAL:
* screen: use UTF-8 representation

-- 
Jon

> 
> OK?
> 
> -- 
> Jon

> Index: editors/neovim/Makefile
> ===
> RCS file: /cvs/ports/editors/neovim/Makefile,v
> retrieving revision 1.9
> diff -u -p -r1.9 Makefile
> --- editors/neovim/Makefile   1 Jul 2018 14:16:29 -   1.9
> +++ editors/neovim/Makefile   28 Jul 2018 18:21:09 -
> @@ -4,7 +4,7 @@ COMMENT = continuation and extension of 
>  
>  GH_ACCOUNT = neovim
>  GH_PROJECT = neovim
> -GH_TAGNAME = v0.3.0
> +GH_TAGNAME = v0.3.1
>  
>  CATEGORIES = editors devel
>  HOMEPAGE =   http://neovim.org
> Index: editors/neovim/distinfo
> ===
> RCS file: /cvs/ports/editors/neovim/distinfo,v
> retrieving revision 1.3
> diff -u -p -r1.3 distinfo
> --- editors/neovim/distinfo   1 Jul 2018 14:16:29 -   1.3
> +++ editors/neovim/distinfo   28 Jul 2018 18:21:22 -
> @@ -1,2 +1,2 @@
> -SHA256 (neovim-0.3.0.tar.gz) = 96y2GxbT9SGQfZnEhrep8eUF6LKhjJ72mm1/GPKfdLg=
> -SIZE (neovim-0.3.0.tar.gz) = 8903630
> +SHA256 (neovim-0.3.1.tar.gz) = vF45LUwHZAeQbM7LwoPhpEt4MsL0hsrYGqBMwplzrSI=
> +SIZE (neovim-0.3.1.tar.gz) = 8937900

[update] editors/neovim 0.3.1

[Jon Bernard]

Hello,

Here is an update to neovim for version 0.3.1.  The new release contains
mostly bugfixes and nothing major.  As such, the diff is just a version
bump and everything else remains the same.

OK?

-- 
Jon
Index: editors/neovim/Makefile
===
RCS file: /cvs/ports/editors/neovim/Makefile,v
retrieving revision 1.9
diff -u -p -r1.9 Makefile
--- editors/neovim/Makefile 1 Jul 2018 14:16:29 -   1.9
+++ editors/neovim/Makefile 28 Jul 2018 18:21:09 -
@@ -4,7 +4,7 @@ COMMENT =   continuation and extension of 
 
 GH_ACCOUNT =   neovim
 GH_PROJECT =   neovim
-GH_TAGNAME =   v0.3.0
+GH_TAGNAME =   v0.3.1
 
 CATEGORIES =   editors devel
 HOMEPAGE = http://neovim.org
Index: editors/neovim/distinfo
===
RCS file: /cvs/ports/editors/neovim/distinfo,v
retrieving revision 1.3
diff -u -p -r1.3 distinfo
--- editors/neovim/distinfo 1 Jul 2018 14:16:29 -   1.3
+++ editors/neovim/distinfo 28 Jul 2018 18:21:22 -
@@ -1,2 +1,2 @@
-SHA256 (neovim-0.3.0.tar.gz) = 96y2GxbT9SGQfZnEhrep8eUF6LKhjJ72mm1/GPKfdLg=
-SIZE (neovim-0.3.0.tar.gz) = 8903630
+SHA256 (neovim-0.3.1.tar.gz) = vF45LUwHZAeQbM7LwoPhpEt4MsL0hsrYGqBMwplzrSI=
+SIZE (neovim-0.3.1.tar.gz) = 8937900

Re: i3status: report wireless signal quality

[Jon Bernard]

* Jon Bernard  wrote:
> * joshua stein  wrote:
> > On Tue, 17 Jul 2018 at 21:42:46 -0400, Jon Bernard wrote:
> > > Hi ports,
> > > 
> > > Here's a small patch to i3status to report wireless signal quality.  If
> > > you're using the default configuration, you should see the '?' change to
> > > the current signal quality percentage.
> > > 
> > > If this looks good I'll submit it upstream.
> > > 
> > > OK?
> > 
> > I'm already getting wireless signal quality with i3status, is this 
> > not the same thing?  My i3status.conf has:
> > 
> > wireless iwm0 {
> > format_up = "up|%signal|%essid"
> > format_down = "down"
> > }
> > 
> > which reports:
> > 
> > [{"name":"wireless","instance":"iwm0","markup":"none","full_text":"up| 
> > 76%|xxx"},...
> 
> I think you're right.  I thought signal was the raw value but it appears
> that i3status computes the percentage for %signal based on the max.  The
> manpage doesn't even mention the signal field so that's fun.  Thanks for
> looking at this, sorry for the noise.

This patch might make more sense then, it updates the openbsd default
configuration to use signal (which we support) instead of quality which
will always yield a '?'.

-- 
Jon
Index: x11/i3status/patches/patch-i3status_conf
===
RCS file: /cvs/ports/x11/i3status/patches/patch-i3status_conf,v
retrieving revision 1.16
diff -u -p -r1.16 patch-i3status_conf
--- x11/i3status/patches/patch-i3status_conf31 Jan 2017 18:47:56 -  
1.16
+++ x11/i3status/patches/patch-i3status_conf18 Jul 2018 03:35:59 -
@@ -2,8 +2,9 @@ $OpenBSD: patch-i3status_conf,v 1.16 201
 
 Adjust config file to be reasonably useful on OpenBSD out of the box.
 
 i3status.conf.orig Tue Jan 31 19:44:46 2017
-+++ i3status.conf  Tue Jan 31 19:46:10 2017
+Index: i3status.conf
+--- i3status.conf.orig
 i3status.conf
 @@ -7,6 +7,7 @@
  # If the above line is not correctly displayed, fix your editor first!
  
@@ -30,7 +31,8 @@ Adjust config file to be reasonably usef
 +}
 +
  wireless _first_ {
- format_up = "W: (%quality at %essid) %ip"
+-format_up = "W: (%quality at %essid) %ip"
++format_up = "W: (%signal at %essid) %ip"
  format_down = "W: down"
  }
  

Re: i3status: report wireless signal quality

[Jon Bernard]

* joshua stein  wrote:
> On Tue, 17 Jul 2018 at 21:42:46 -0400, Jon Bernard wrote:
> > Hi ports,
> > 
> > Here's a small patch to i3status to report wireless signal quality.  If
> > you're using the default configuration, you should see the '?' change to
> > the current signal quality percentage.
> > 
> > If this looks good I'll submit it upstream.
> > 
> > OK?
> 
> I'm already getting wireless signal quality with i3status, is this 
> not the same thing?  My i3status.conf has:
> 
> wireless iwm0 {
> format_up = "up|%signal|%essid"
> format_down = "down"
> }
> 
> which reports:
> 
> [{"name":"wireless","instance":"iwm0","markup":"none","full_text":"up| 
> 76%|xxx"},...

I think you're right.  I thought signal was the raw value but it appears
that i3status computes the percentage for %signal based on the max.  The
manpage doesn't even mention the signal field so that's fun.  Thanks for
looking at this, sorry for the noise.

-- 
Jon

i3status: report wireless signal quality

[Jon Bernard]

Hi ports,

Here's a small patch to i3status to report wireless signal quality.  If
you're using the default configuration, you should see the '?' change to
the current signal quality percentage.

If this looks good I'll submit it upstream.

OK?

-- 
Jon
Index: x11/i3status/patches/patch-src_print_wireless_info_c
===
RCS file: /cvs/ports/x11/i3status/patches/patch-src_print_wireless_info_c,v
retrieving revision 1.13
diff -u -p -r1.13 patch-src_print_wireless_info_c
--- x11/i3status/patches/patch-src_print_wireless_info_c12 May 2018 
14:03:46 -  1.13
+++ x11/i3status/patches/patch-src_print_wireless_info_c18 Jul 2018 
01:28:42 -
@@ -1,6 +1,7 @@
 $OpenBSD: patch-src_print_wireless_info_c,v 1.13 2018/05/12 14:03:46 jasper 
Exp $
 
-warning: implicit declaration of function 'free' is invalid in C99
+* warning: implicit declaration of function 'free' is invalid in C99
+* report wireless signal quality
 
 Index: src/print_wireless_info.c
 --- src/print_wireless_info.c.orig
@@ -12,3 +13,15 @@ Index: src/print_wireless_info.c
  #include 
  #include 
  #include 
+@@ -454,6 +455,11 @@ error1:
+ info->signal_level_max = nr.nr_max_rssi;
+ 
+ info->flags |= WIRELESS_INFO_FLAG_HAS_SIGNAL;
++
++info->quality = IEEE80211_NODEREQ_RSSI();
++info->quality_max = 100;
++info->quality_average = 50;
++info->flags |= WIRELESS_INFO_FLAG_HAS_QUALITY;
+ }
+ }
+ }

Re: [update] editors/neovim 0.3.0

[Jon Bernard]

* Stuart Henderson  wrote:
> On 2018/06/22 15:32, Jon Bernard wrote:
> > Hello,
> > 
> > Here is an update to neovim.  The patch we carried in our tree was
> > merged upstream, so no longer needed.
> > 
> > OK?
> > 
> > -- 
> > Jon
> 
> > Index: Makefile
> > ===
> > RCS file: /cvs/ports/editors/neovim/Makefile,v
> > retrieving revision 1.7
> > diff -u -p -r1.7 Makefile
> > --- Makefile20 May 2018 08:30:56 -  1.7
> > +++ Makefile21 Jun 2018 18:45:07 -
> > @@ -4,8 +4,7 @@ COMMENT =   continuation and extension of 
> >  
> >  GH_ACCOUNT =   neovim
> >  GH_PROJECT =   neovim
> > -GH_TAGNAME =   v0.2.2
> > -REVISION = 0
> > +GH_TAGNAME =   v0.3.0
> >  
> >  CATEGORIES =   editors devel
> >  HOMEPAGE = http://neovim.org
> > @@ -16,6 +15,7 @@ PERMIT_PACKAGE_CDROM =Yes
> >  
> >  WANTLIB += c iconv intl ${MODLUA_WANTLIB} m msgpackc pthread termkey
> >  WANTLIB += unibilium util uv vterm
> > +WANTLIB += c++abi
> 
> c++abi is only on architectures with clang in base, so this will fail on other
> arches.

Here's a new version that uses COMPILER_LIBCXX and includes Edd's README
suggestions.

-- 
Jon
Index: Makefile
===
RCS file: /cvs/ports/editors/neovim/Makefile,v
retrieving revision 1.7
diff -u -p -r1.7 Makefile
--- Makefile20 May 2018 08:30:56 -  1.7
+++ Makefile26 Jun 2018 17:10:07 -
@@ -4,8 +4,7 @@ COMMENT =   continuation and extension of 
 
 GH_ACCOUNT =   neovim
 GH_PROJECT =   neovim
-GH_TAGNAME =   v0.2.2
-REVISION = 0
+GH_TAGNAME =   v0.3.0
 
 CATEGORIES =   editors devel
 HOMEPAGE = http://neovim.org
@@ -16,6 +15,7 @@ PERMIT_PACKAGE_CDROM =Yes
 
 WANTLIB += c iconv intl ${MODLUA_WANTLIB} m msgpackc pthread termkey
 WANTLIB += unibilium util uv vterm
+WANTLIB += ${COMPILER_LIBCXX}
 
 MODULES =  devel/cmake \
lang/lua \
Index: distinfo
===
RCS file: /cvs/ports/editors/neovim/distinfo,v
retrieving revision 1.2
diff -u -p -r1.2 distinfo
--- distinfo2 Dec 2017 12:28:05 -   1.2
+++ distinfo26 Jun 2018 17:06:21 -
@@ -1,2 +1,2 @@
-SHA256 (neovim-0.2.2.tar.gz) = qDjuB8yaLvit4bMaKk8tXpM54kSt5o5kVWwfS0DMxe0=
-SIZE (neovim-0.2.2.tar.gz) = 8325879
+SHA256 (neovim-0.3.0.tar.gz) = 96y2GxbT9SGQfZnEhrep8eUF6LKhjJ72mm1/GPKfdLg=
+SIZE (neovim-0.3.0.tar.gz) = 8903630
Index: patches/patch-src_nvim_vim_h
===
RCS file: patches/patch-src_nvim_vim_h
diff -N patches/patch-src_nvim_vim_h
--- patches/patch-src_nvim_vim_h2 Dec 2017 12:28:05 -   1.1
+++ /dev/null   1 Jan 1970 00:00:00 -
@@ -1,18 +0,0 @@
-$OpenBSD: patch-src_nvim_vim_h,v 1.1 2017/12/02 12:28:05 edd Exp $
-
-Fix opening character devices.
-https://github.com/neovim/neovim/issues/7542
-https://github.com/jamessan/neovim/commit/9c58b502d85e4ac08a358d29d5bf4bdae20e8ad4
-
-Index: src/nvim/vim.h
 src/nvim/vim.h.orig
-+++ src/nvim/vim.h
-@@ -314,7 +314,7 @@ enum { FOLD_TEXT_LEN = 51 };  //!< buffer size for get
- // Lowest number used for window ID. Cannot have this many windows per tab.
- #define LOWEST_WIN_ID 1000
- 
--#if (defined(__FreeBSD__) || defined(__FreeBSD_kernel__)) && defined(S_ISCHR)
-+#if (defined(__FreeBSD__) || defined(__FreeBSD_kernel__) || 
defined(__OpenBSD__)) && defined(S_ISCHR)
- # define OPEN_CHR_FILES
- #endif
- 
Index: pkg/PLIST
===
RCS file: /cvs/ports/editors/neovim/pkg/PLIST,v
retrieving revision 1.3
diff -u -p -r1.3 PLIST
--- pkg/PLIST   20 May 2018 08:30:56 -  1.3
+++ pkg/PLIST   26 Jun 2018 17:06:21 -
@@ -56,6 +56,8 @@ share/nvim/runtime/autoload/context.vim
 share/nvim/runtime/autoload/contextcomplete.vim
 share/nvim/runtime/autoload/csscomplete.vim
 share/nvim/runtime/autoload/decada.vim
+share/nvim/runtime/autoload/dist/
+share/nvim/runtime/autoload/dist/ft.vim
 share/nvim/runtime/autoload/gnat.vim
 share/nvim/runtime/autoload/gzip.vim
 share/nvim/runtime/autoload/health/
@@ -204,6 +206,7 @@ share/nvim/runtime/doc/api.txt
 share/nvim/runtime/doc/arabic.txt
 share/nvim/runtime/doc/autocmd.txt
 share/nvim/runtime/doc/change.txt
+share/nvim/runtime/doc/channel.txt
 share/nvim/runtime/doc/cmdline.txt
 share/nvim/runtime/doc/debug.txt
 share/nvim/runtime/doc/deprecated.txt
@@ -728,6 +731,8 @@ share/nvim/runtime/keymap/ukrainian-jcuk
 share/nvim/runtime/keymap/vietnamese-telex_utf-8.vim
 share/nvim/runtime/keymap/vietnamese-viqr_utf-8.vim
 share/nvim/runtime/keymap/vietnamese-vni_utf-8.vim
+share/nvim/runtime/lua/
+share/nvim/runti

[update] editors/py-neovim 0.2.6

[Jon Bernard]

Hello,

Here is an update to py-neovim.  Upstream switched from nose to pytest
for unit tests, but forgot to include the file defining their test
fixtures in the release tarball.  So that file is included as a patch
from the upstream source tree at the time of the tag.  The tests may
need increased file descriptor limits to succeed, depending on your
environment.

OK?

-- 
Jon
Index: Makefile
===
RCS file: /cvs/ports/editors/py-neovim/Makefile,v
retrieving revision 1.2
diff -u -p -r1.2 Makefile
--- Makefile2 Dec 2017 12:30:56 -   1.2
+++ Makefile21 Jun 2018 18:56:57 -
@@ -2,7 +2,7 @@
 
 COMMENT =  Python plugin support for Neovim
 
-MODPY_EGG_VERSION =0.2.0
+MODPY_EGG_VERSION =0.2.6
 DISTNAME = py-neovim-${MODPY_EGG_VERSION}
 
 CATEGORIES =   editors devel
@@ -26,11 +26,13 @@ RUN_DEPENDS =   net/py-msgpack${MODPY_FLA
devel/py-uv${MODPY_FLAVOR} \
devel/py-greenlet${MODPY_FLAVOR}
 TEST_DEPENDS = ${RUN_DEPENDS} \
-   devel/py-nose${MODPY_FLAVOR}
+   devel/py-test${MODPY_FLAVOR} \
+   ${BASE_PKGPATH}=${MODPY_EGG_VERSION}
 
 WRKDIST =  ${WRKDIR}/neovim-${MODPY_EGG_VERSION}
 
+# You may need to increase the file descriptor ulimits to run tests.
 do-test:
-   cd ${WRKSRC} && ${LOCALBASE}/bin/nosetests${MODPY_BIN_SUFFIX}
+   cd ${WRKSRC} && ${LOCALBASE}/bin/py.test${MODPY_BIN_SUFFIX}
 
 .include 
Index: distinfo
===
RCS file: /cvs/ports/editors/py-neovim/distinfo,v
retrieving revision 1.2
diff -u -p -r1.2 distinfo
--- distinfo2 Dec 2017 12:30:56 -   1.2
+++ distinfo21 Jun 2018 18:32:54 -
@@ -1,2 +1,2 @@
-SHA256 (py-neovim-0.2.0.tar.gz) = 1gvgS6ND9+qZxUqYYqJUognOnKvNw0gf5ayk1vN6k/s=
-SIZE (py-neovim-0.2.0.tar.gz) = 35957
+SHA256 (py-neovim-0.2.6.tar.gz) = bOWKdC4EJ0kcDhyBCFVu5yujOEQgm9niJrjalTgpknY=
+SIZE (py-neovim-0.2.6.tar.gz) = 36574
Index: patches/patch-test_conftest_py
===
RCS file: patches/patch-test_conftest_py
diff -N patches/patch-test_conftest_py
--- /dev/null   1 Jan 1970 00:00:00 -
+++ patches/patch-test_conftest_py  21 Jun 2018 18:32:54 -
@@ -0,0 +1,78 @@
+$OpenBSD$
+
+Although this file is present in the upstream source tree, it was not
+included when the release tarball was created.  This file is needed to
+run unit tests and was copied from the release tag matching this
+version.
+
+Index: test/conftest.py
+--- test/conftest.py.orig
 test/conftest.py
+@@ -0,0 +1,67 @@
++import json
++import os
++import textwrap
++
++import neovim
++import pytest
++
++neovim.setup_logging("test")
++
++
++@pytest.fixture(autouse=True)
++def cleanup_func(vim):
++fun = textwrap.dedent(''':function BeforeEachTest()
++set all&
++redir => groups
++silent augroup
++redir END
++for group in split(groups)
++exe 'augroup '.group
++autocmd!
++augroup END
++endfor
++autocmd!
++tabnew
++let curbufnum = eval(bufnr('%'))
++redir => buflist
++silent ls!
++redir END
++let bufnums = []
++for buf in split(buflist, '\\n')
++let bufnum = eval(split(buf, '[ u]')[0])
++if bufnum != curbufnum
++call add(bufnums, bufnum)
++endif
++endfor
++if len(bufnums) > 0
++exe 'silent bwipeout! '.join(bufnums, ' ')
++endif
++silent tabonly
++for k in keys(g:)
++exe 'unlet g:'.k
++endfor
++filetype plugin indent off
++mapclear
++mapclear!
++abclear
++comclear
++endfunction
++''')
++vim.input(fun)
++vim.command('call BeforeEachTest()')
++assert len(vim.tabpages) == len(vim.windows) == len(vim.buffers) == 1
++
++
++@pytest.fixture
++def vim():
++child_argv = os.environ.get('NVIM_CHILD_ARGV')
++listen_address = os.environ.get('NVIM_LISTEN_ADDRESS')
++if child_argv is None and listen_address is None:
++child_argv = '["nvim", "-u", "NONE", "--embed"]'
++
++if child_argv is not None:
++editor = neovim.attach('child', argv=json.loads(child_argv))
++else:
++editor = neovim.attach('socket', path=listen_address)
++
++return editor
Index: pkg/PLIST
===
RCS file: /cvs/ports/editors/py-neovim/pkg/PLIST,v
retrieving revision 1.2
diff -u -p -r1.2 PLIST
--- pkg/PLIST   2 Dec 2017 12:30:56 -   1.2
+++ pkg/PLIST   21 Jun 2018 18:58:25 -
@@ -12,7 +12,6 @@ ${MODPY_COMMENT}lib/python${MODPY_VERSIO
 
lib/python${MODPY_VERSION}/site-packages/neovim/${MODPY_PYCACHE}__init__.${MODPY_PYC_MAGIC_TAG}pyc
 

[update] editors/neovim 0.3.0

[Jon Bernard]

Hello,

Here is an update to neovim.  The patch we carried in our tree was
merged upstream, so no longer needed.

OK?

-- 
Jon
Index: Makefile
===
RCS file: /cvs/ports/editors/neovim/Makefile,v
retrieving revision 1.7
diff -u -p -r1.7 Makefile
--- Makefile20 May 2018 08:30:56 -  1.7
+++ Makefile21 Jun 2018 18:45:07 -
@@ -4,8 +4,7 @@ COMMENT =   continuation and extension of 
 
 GH_ACCOUNT =   neovim
 GH_PROJECT =   neovim
-GH_TAGNAME =   v0.2.2
-REVISION = 0
+GH_TAGNAME =   v0.3.0
 
 CATEGORIES =   editors devel
 HOMEPAGE = http://neovim.org
@@ -16,6 +15,7 @@ PERMIT_PACKAGE_CDROM =Yes
 
 WANTLIB += c iconv intl ${MODLUA_WANTLIB} m msgpackc pthread termkey
 WANTLIB += unibilium util uv vterm
+WANTLIB += c++abi
 
 MODULES =  devel/cmake \
lang/lua \
Index: distinfo
===
RCS file: /cvs/ports/editors/neovim/distinfo,v
retrieving revision 1.2
diff -u -p -r1.2 distinfo
--- distinfo2 Dec 2017 12:28:05 -   1.2
+++ distinfo19 Jun 2018 16:40:39 -
@@ -1,2 +1,2 @@
-SHA256 (neovim-0.2.2.tar.gz) = qDjuB8yaLvit4bMaKk8tXpM54kSt5o5kVWwfS0DMxe0=
-SIZE (neovim-0.2.2.tar.gz) = 8325879
+SHA256 (neovim-0.3.0.tar.gz) = 96y2GxbT9SGQfZnEhrep8eUF6LKhjJ72mm1/GPKfdLg=
+SIZE (neovim-0.3.0.tar.gz) = 8903630
Index: patches/patch-src_nvim_vim_h
===
RCS file: patches/patch-src_nvim_vim_h
diff -N patches/patch-src_nvim_vim_h
--- patches/patch-src_nvim_vim_h2 Dec 2017 12:28:05 -   1.1
+++ /dev/null   1 Jan 1970 00:00:00 -
@@ -1,18 +0,0 @@
-$OpenBSD: patch-src_nvim_vim_h,v 1.1 2017/12/02 12:28:05 edd Exp $
-
-Fix opening character devices.
-https://github.com/neovim/neovim/issues/7542
-https://github.com/jamessan/neovim/commit/9c58b502d85e4ac08a358d29d5bf4bdae20e8ad4
-
-Index: src/nvim/vim.h
 src/nvim/vim.h.orig
-+++ src/nvim/vim.h
-@@ -314,7 +314,7 @@ enum { FOLD_TEXT_LEN = 51 };  //!< buffer size for get
- // Lowest number used for window ID. Cannot have this many windows per tab.
- #define LOWEST_WIN_ID 1000
- 
--#if (defined(__FreeBSD__) || defined(__FreeBSD_kernel__)) && defined(S_ISCHR)
-+#if (defined(__FreeBSD__) || defined(__FreeBSD_kernel__) || 
defined(__OpenBSD__)) && defined(S_ISCHR)
- # define OPEN_CHR_FILES
- #endif
- 
Index: pkg/PLIST
===
RCS file: /cvs/ports/editors/neovim/pkg/PLIST,v
retrieving revision 1.3
diff -u -p -r1.3 PLIST
--- pkg/PLIST   20 May 2018 08:30:56 -  1.3
+++ pkg/PLIST   19 Jun 2018 16:40:39 -
@@ -56,6 +56,8 @@ share/nvim/runtime/autoload/context.vim
 share/nvim/runtime/autoload/contextcomplete.vim
 share/nvim/runtime/autoload/csscomplete.vim
 share/nvim/runtime/autoload/decada.vim
+share/nvim/runtime/autoload/dist/
+share/nvim/runtime/autoload/dist/ft.vim
 share/nvim/runtime/autoload/gnat.vim
 share/nvim/runtime/autoload/gzip.vim
 share/nvim/runtime/autoload/health/
@@ -204,6 +206,7 @@ share/nvim/runtime/doc/api.txt
 share/nvim/runtime/doc/arabic.txt
 share/nvim/runtime/doc/autocmd.txt
 share/nvim/runtime/doc/change.txt
+share/nvim/runtime/doc/channel.txt
 share/nvim/runtime/doc/cmdline.txt
 share/nvim/runtime/doc/debug.txt
 share/nvim/runtime/doc/deprecated.txt
@@ -728,6 +731,8 @@ share/nvim/runtime/keymap/ukrainian-jcuk
 share/nvim/runtime/keymap/vietnamese-telex_utf-8.vim
 share/nvim/runtime/keymap/vietnamese-viqr_utf-8.vim
 share/nvim/runtime/keymap/vietnamese-vni_utf-8.vim
+share/nvim/runtime/lua/
+share/nvim/runtime/lua/man.lua
 share/nvim/runtime/macmap.vim
 share/nvim/runtime/macros/
 share/nvim/runtime/macros/editexisting.vim
@@ -766,7 +771,6 @@ share/nvim/runtime/pack/dist/opt/vimball
 share/nvim/runtime/pack/dist/opt/vimball/plugin/
 share/nvim/runtime/pack/dist/opt/vimball/plugin/vimballPlugin.vim
 share/nvim/runtime/plugin/
-share/nvim/runtime/plugin/gui_shim.vim
 share/nvim/runtime/plugin/gzip.vim
 share/nvim/runtime/plugin/health.vim
 share/nvim/runtime/plugin/man.vim
@@ -815,6 +819,8 @@ share/nvim/runtime/print/mac-roman.ps
 share/nvim/runtime/print/prolog.ps
 share/nvim/runtime/rgb.txt
 share/nvim/runtime/scripts.vim
+share/nvim/runtime/spell/
+share/nvim/runtime/spell/en.utf-8.spl
 share/nvim/runtime/synmenu.vim
 share/nvim/runtime/syntax/
 share/nvim/runtime/syntax/2html.vim
@@ -1419,7 +1425,9 @@ share/nvim/runtime/syntax/zsh.vim
 share/nvim/runtime/tutor/
 share/nvim/runtime/tutor/en/
 share/nvim/runtime/tutor/en/vim-01-beginner.tutor
+share/nvim/runtime/tutor/en/vim-01-beginner.tutor.json
 share/nvim/runtime/tutor/tutor.tutor
+share/nvim/runtime/tutor/tutor.tutor.json
 share/pixmaps/
 share/pixmaps/nvim.png
 @exec %D/bin/update-desktop-database

Re: NEW: devel/leiningen

[Jon Bernard]

* Jasper Lievisse Adriaanse <jas...@openbsd.org> wrote:

On Sun, Jul 02, 2017 at 05:12:15PM +0100, Edd Barrett wrote:

Hi Jon,

On Thu, Jun 29, 2017 at 10:38:32AM -0400, Jon Bernard wrote:
> Attached is a port of leiningen, for automating clojure projects.

Comments:

 * There's a hard-coded path to bash in a patch. You will need to use a
   variable (LOCALBASE) in place of /usr/local and then use
   ${SUBST_CMD} in a custom make target, maybe post-install.

 * Did you try that script with /bin/sh? If it works that would be
   preferable. Otherwise, bash needs to be a RUN_DEPEND.

 * mandoc linter is unhappy with the man page:

   $ mandoc -Tlint /usr/local/man/man1/lein.1
   mandoc: /usr/local/man/man1/lein.1:2:17: WARNING: cannot parse date, using 
it verbatim: 2011 June 30

   Consider pathcing?

 * I was surprised to see that your port did not RUN_DEPEND on clojure.
   I see that upon first `lein repl` the tool will download a clojure
   using (the horror) maven (which must be bundled?). I *think*, but
   other porters can correct me, we would prefer to use the in-tree
   clojure instead of a moving target. Is this possible?

 * Even if I let it do its thing, the repl doesn't seem to work:

---8<---
$ lein repl
Retrieving org/clojure/clojure/1.8.0/clojure-1.8.0.pom from central
Retrieving org/sonatype/oss/oss-parent/7/oss-parent-7.pom from central
Retrieving org/clojure/tools.nrepl/0.2.12/tools.nrepl-0.2.12.pom from 
central
Retrieving org/clojure/pom.contrib/0.1.2/pom.contrib-0.1.2.pom from central
Retrieving 
clojure-complete/clojure-complete/0.2.4/clojure-complete-0.2.4.pom from clojars
Retrieving org/clojure/tools.nrepl/0.2.12/tools.nrepl-0.2.12.jar from 
central
Retrieving org/clojure/clojure/1.8.0/clojure-1.8.0.jar from central
Retrieving 
clojure-complete/clojure-complete/0.2.4/clojure-complete-0.2.4.jar from clojars
REPL server launch timed out.
--->8---

I know nothing about Clojure or lein, so I'm going to CC our clojure
MAINTAINER, Jasper.

Cheers

--
Best Regards
Edd Barrett

http://www.theunixzoo.co.uk


We used to have a port of leiningen, but I removed it more than 3 years ago:

8<
Revision 1.6, Sat Nov 9 10:40:20 2013 UTC (3 years, 7 months ago) by jasper
Branch: MAIN
CVS Tags: HEAD
Changes since 1.5: +1 -1 lines
FILE REMOVED

remove leiningen; the upstream script is not really intended to get packaged
anyway.
8<

Might be worth considering before proceeding to re-add the files if the
situation has improved since 2013.


I believe the upstream offering has changed since that time, there is 
now a 'lein-pkg' script which is intended for downstream packagers [1]. 

If you have time to review, I'd be very curious to know what you think. 
I suspect Edd's repl timeout is unrelated to the package itself, 
possibly a transient server error fetching the requisite jars. 


[1] https://github.com/technomancy/leiningen/blob/master/bin/lein-pkg#L3-L5

--
Jon

Re: NEW: x11/xcape

[Jon Bernard]

* Edd Barrett <e...@theunixzoo.co.uk> wrote:

Hi Jon,

On Tue, May 02, 2017 at 09:51:57AM -0400, Jon Bernard wrote:

This is my first port, I think everything is in order from what I've
read but do let me know if I've missed or misunderstood something.


Here are a few comments:

* You need to mark the port as not having tests (see NO_TEST).

* You are missing a WANTLIB line. There's a make target
  'port-lib-depends-check' that can help you.

* Funny that they have an install target which installs the manual, but
  not the binary! Doh! Might want to report that, so that in later
  versions you don't need a custom do-install target.

* You should set a HOMEPAGE, even if it's just the GitHub landing page.


All taken care of.  Thanks for looking at this.


The tool seems to work. With xcape running in the default config, left
control sends escape.

Fix up the above and send a new tarball and I will see if we can put
this in for you :)


New version with comments and suggestions applied.

--
Jon


xcape-1.2-2.tgz
Description: application/tar-gz

Re: NEW: x11/xcape

[Jon Bernard]

* Edd Barrett  wrote:

On Sun, Jul 02, 2017 at 04:38:49PM +0100, Edd Barrett wrote:

Fix up the above and send a new tarball and I will see if we can put
this in for you :)


Oh, and if I were nitpicking:

$ mandoc -Tlint /usr/local/man/man1/xcape.1
mandoc: /usr/local/man/man1/xcape.1:31:57: WARNING: whitespace at end of input 
line

But unless it breaks rendering, I wouldn't worry.


I patched it in the port and submitted upstream

 https://github.com/alols/xcape/pull/91

--
Jon

Re: NEW: x11/xcape

[Jon Bernard]

* Edd Barrett  wrote:

Hi,

On Mon, Jul 03, 2017 at 09:30:50AM +0200, Antoine Jacoutot wrote:

>  * Funny that they have an install target which installs the manual, but
>not the binary! Doh! Might want to report that, so that in later
>versions you don't need a custom do-install target.

They do install the binary.
You just need to specify proper variables:
MAKE_FLAGS =PREFIX=${PREFIX} \
MANDIR="/man/man1"


Antoine is right. If you look at the fake stage output:

---8<---
===>  Faking installation for xcape-1.2
install -d -m 755 /usr/ports/pobj/xcape-1.2/fake-amd64
install -d -m 0755 /usr/ports/pobj/xcape-1.2/fake-amd64/usr/bin
install -d -m 0755 /usr/ports/pobj/xcape-1.2/fake-amd64/usr/local/man/man1
install -m 0755 xcape /usr/ports/pobj/xcape-1.2/fake-amd64/usr/bin/xcape
install -m 0644 xcape.1 
/usr/ports/pobj/xcape-1.2/fake-amd64/usr/local/man/man1/xcape.1
--->8---

The binary goes under /usr/bin in the fake root, which is presumably why
`update-plist` doesn't find it.

In light of that, you shold be able to kill the custom do-install target.


Ok, will be included in the next version.


He's also right about HOMEPAGE :) Thanks Antoine!


That's nice, I didn't realize that either.

--
Jon

Re: NEW: x11/xcape

[Jon Bernard]

* Antoine Jacoutot <ajacou...@bsdfrog.org> wrote:

On Sun, Jul 02, 2017 at 04:38:49PM +0100, Edd Barrett wrote:

Hi Jon,

On Tue, May 02, 2017 at 09:51:57AM -0400, Jon Bernard wrote:
> This is my first port, I think everything is in order from what I've
> read but do let me know if I've missed or misunderstood something.

Here are a few comments:

 * You need to mark the port as not having tests (see NO_TEST).

 * You are missing a WANTLIB line. There's a make target
   'port-lib-depends-check' that can help you.

 * Funny that they have an install target which installs the manual, but
   not the binary! Doh! Might want to report that, so that in later
   versions you don't need a custom do-install target.


They do install the binary.
You just need to specify proper variables:
MAKE_FLAGS =PREFIX=${PREFIX} \
   MANDIR="/man/man1"


I'm not sure what I was thinking.. You're right, thanks for pointing this out.

--
Jon


 * You should set a HOMEPAGE, even if it's just the GitHub landing page.


There is already one, thanks to the GH_* veriables:
$ make show=HOMEPAGE
https://github.com/alols/xcape


The tool seems to work. With xcape running in the default config, left
control sends escape.

Fix up the above and send a new tarball and I will see if we can put
this in for you :)

Thanks

--
Best Regards
Edd Barrett

http://www.theunixzoo.co.uk



--
Antoine

Re: NEW: sysutils/myrepos

[Jon Bernard]

* Edd Barrett <e...@theunixzoo.co.uk> wrote:

On Mon, May 08, 2017 at 10:12:50AM -0400, Jon Bernard wrote:

Hi ports@,

Attached is a port of myrepos from Joey Hess.


Comments:

* COMMENT starts with a lower case letter unless the first word is a
  proper noun.

* Author moved the project away from github, so the port no longer
  fetches. If you can't find a tarball somewhere, you might have to
  roll your own and host it.

If you can fix the latter, I will complete the review.


No problem, attached a new version with the above changes.

Cheers,

--
Jon


myrepos-1.20170129-2.tgz
Description: application/tar-gz

NEW: devel/leiningen

[Jon Bernard]

Hi ports@,

Attached is a port of leiningen, for automating clojure projects.

DESCR:

Leiningen is the easiest way to use Clojure.  With a focus on project
automation and declarative configuration, it gets out of your way and
lets you focus on your code.

--
Jon


leiningen-2.7.1.tgz
Description: application/tar-gz

NEW: sysutils/myrepos

[Jon Bernard]

Hi ports@,

Attached is a port of myrepos from Joey Hess.

DESCR:

You have a lot of version control repositories.  Sometimes you want to
update them all at once.  Or push out all your local changes.  You use
special command lines in some repositories to implement specific
workflows.  Myrepos provides a mr command, which is a tool to manage all
your version control repositories.

-- 
Jon


myrepos-1.20170129.tgz
Description: application/tar-gz

NEW: x11/xcape

[Jon Bernard]

Hi ports@,

Attached is a port of xcape from Albin Olsson.

DESCR:

xcape allows you to use a modifier key as another key when pressed and
released on its own. Note that it is slightly slower than pressing the
original key, because the pressed event does not occur until the key is
released. The default behaviour is to generate the Escape key when Left
Control is pressed and released on its own. (If you don't understand why
anybody would want this, I'm guessing that Vim is not your favourite
text editor ;)

This is my first port, I think everything is in order from what I've
read but do let me know if I've missed or misunderstood something.

Cheers,

-- 
Jon


xcape-1.2.tgz
Description: application/tar-gz