[PATCH v4 5/5] notmuch-emacs-mua: do not create a frame by default with --client

[Jani Nikula]

On Wed, 26 Aug 2015, Tomi Ollila <tomi.ollila at iki.fi> wrote:
> On Sat, Aug 15 2015, Jani Nikula <jani at nikula.org> wrote:
>
>> Make the default behaviour for --client the same as emacsclient
>> default: do not create a new frame. Add a new option --create-frame,
>> passing the same option to emacsclient to create a frame.
>>
>> ---
>>
>> v2: fix killing frame with --create-frame
>>
>> v3: fix --create-frame w/o --client parameter check
>>
>> v4: fix kill terminal logic, document behaviour with daemon mode
>> ---
>>  doc/man1/notmuch-emacs-mua.rst |  6 ++++++
>>  notmuch-emacs-mua              | 14 ++++++++++----
>>  2 files changed, 16 insertions(+), 4 deletions(-)
>>
>> diff --git a/doc/man1/notmuch-emacs-mua.rst b/doc/man1/notmuch-emacs-mua.rst
>> index 89a2fead17b1..c3689eb82c8a 100644
>> --- a/doc/man1/notmuch-emacs-mua.rst
>> +++ b/doc/man1/notmuch-emacs-mua.rst
>> @@ -45,6 +45,12 @@ Supported options for **notmuch-emacs-mua** include
>>          Automatically start Emacs in daemon mode, if the Emacs server
>>          is not running. Applicable with ``--client``.
>>  
>> +    ``--create-frame``
>> +        Create a new frame instead of trying to use the current Emacs
>> +        frame. Applicable with ``--client``. This will be required
>> +        when Emacs is running (or automatically started with
>> +        ``--auto-daemon``) in daemon mode.
>> +
>>      ``--print``
>>          Output the resulting elisp to stdout instead of evaluating it.
>>  
>> diff --git a/notmuch-emacs-mua b/notmuch-emacs-mua
>> index ac03a4a60d56..b51d8d0e2c5b 100755
>> --- a/notmuch-emacs-mua
>> +++ b/notmuch-emacs-mua
>> @@ -37,11 +37,10 @@ PRINT_ONLY=
>>  NO_WINDOW=
>>  USE_EMACSCLIENT=
>>  AUTO_DAEMON=
>> -CREATE_FRAME="-c"
>> +CREATE_FRAME=
>>  
>>  # The crux of it all: construct an elisp progn and eval it.
>>  ELISP="(prog1 'done (require 'notmuch) (notmuch-mua-new-mail)"
>> -ELISP="${ELISP} (setq message-exit-actions (list #'save-buffers-kill-terminal))"
>>  
>>  # Short options compatible with mutt(1).
>>  while getopts :s:c:b:i:h opt; do
>> @@ -65,7 +64,7 @@ while getopts :s:c:b:i:h opt; do
>>  		    opt=${opt%%=*}
>>  		    ;;
>>  		# Long options without arguments.
>> -		--help|--print|--no-window-system|--client|--auto-daemon)
>> +		--help|--print|--no-window-system|--client|--auto-daemon|--create-frame)
>>  		    ;;
>>  		*)
>>  		    echo "$0: unknown long option ${opt}, or argument mismatch." >&2
>> @@ -103,7 +102,6 @@ while getopts :s:c:b:i:h opt; do
>>  	    ;;
>>  	--no-window-system)
>>  	    NO_WINDOW="-nw"
>> -	    CREATE_FRAME=
>>  	    ;;
>>  	--client)
>>  	    USE_EMACSCLIENT="yes"
>> @@ -111,6 +109,9 @@ while getopts :s:c:b:i:h opt; do
>>  	--auto-daemon)
>>  	    AUTO_DAEMON="--alternate-editor="
>>  	    ;;
>> +	--create-frame)
>> +	    CREATE_FRAME="-c"
>> +	    ;;
>>  	*)
>>  	    # We should never end up here.
>>  	    echo "$0: internal error (option ${opt})." >&2
>
> This looks pretty good now; I player quite a bit with --client,
> --create-frame and --auto-daemon 
> (with and without setting EMACSCLIENT='echo emacsclient')
>
> Based on that I think this check should be added here (after option
> parsing):
>
> if [ -n "$AUTO_DAEMON" -a -z "$CREATE_FRAME" ]; then
>     echo "$0: --auto-daemon is only applicable with --create-frame." >&2
>     exit 1       
> fi
>
> without this one may execute ./notmuch-emacs-mua --client --auto-daemon
> which yields starting emacs in daemon mode (in this example it is expected
> emacs is not running; otherwise --auto-daemon has no use in this example)
> -- but no ui to that newly-running emacs is provided. Similar behaviour
> can be observed by the following

This problem is present with the current code already, right now. I
don't see the need to fix everything in my series.

> $ emacsclient --eval '(kill emacs) ;; but beware! ' 
> $ emacsclient --alternate-editor= --eval '(find-file "/etc/passwd")'
>
> (side note: using this --alternate-editor= is pretty nifty way to get it done)
>
> emacs --daemon is executed which loaded /etc/passwd but w/o ui
>
> Note the difference in:
>
> emacsclient --alternate-editor= /etc/passwd
>
> now emacs --daemon is executed *and* ui where /etc/passwd is available is
> provided!
>
> otoh,
> emacsclient --alternate-editor= -c [-nw] --eval '(find-file "/etc/passwd")'
>
> works fine, albeit a bit different than the one where /etc/passwd was given
> as a simple filename argument.
>
> comparative to ./notmuch-emacs-mua --client --create-frame --auto-daemon 
>
>> @@ -128,6 +129,11 @@ for arg; do
>>      ELISP="${ELISP} (message-goto-to) (insert \"${arg}, \")"
>>  done
>>  
>> +# Kill the terminal/frame if we're creating one.
>> +if [ -z "$USE_EMACSCLIENT" -o -n "$CREATE_FRAME" -o -n "$NO_WINDOW" ]; then
>> +    ELISP="${ELISP} (setq message-exit-actions (list #'save-buffers-kill-terminal))"
>> +fi
>
> I am not very happy that message-exit-actions was added to $ELISP when
> not using emacsclient; when emacs is started its sole (initial) purpose is
> to serve mail sending (and not lending a frame in some other emacs) -- in
> this case it would be nice to be able to retrieve the sent mail buffer.

And I am not very happy with being expected to modify these patches ad
infinitum while trying to get back the *one* use case I wrote this
script for in the first place, and that got broken somewhere down the
line.

Sincerely,
Jani.


>
> perhaps the above could be changed to just 
>
> if [ -n "$CREATE_FRAME" ]; then
>     ELISP="${ELISP} (setq message-exit-actions (list #'save-buffers-kill-terminal))"
> fi
>
> another option is --yet-another-option ;/
>
> No other issues (relevant to this series).
>
> Tomi
>
>
>> +
>>  # End progn.
>>  ELISP="${ELISP})"
>>  
>> -- 
>> 2.1.4