Re: Getting X selection reliably (Re: idea for capture anywhere in x)

[Ihor Radchenko <yantar92@posteo.net>]

Max Nikulin <manikulin@gmail.com> writes:

>> I am unsure what can be done here. AFAIU, Org just follows the %x
>> description. Nothing wrong on Org side.
>> 
>> The best we can do is alternative %?? that will do something more
>> reliable. But I am unsure what it can be.
>
> Depending of particular capture template (even with :immediate-finish t) 
> empty selection may or may not be an error. In Samuel's case it is an 
> error. Unfortunately, while the template is processed, signaled errors 
> are caught and their text is inserted into the text. That is why in my 
> example I put checks outside of `org-capture'.

Are you sure? `org-get-x-clipboard' returns nil in case of error and
`org-capture-fill-template' assigns empty string to v-x when no
clipboard value can be obtained.

> We may:
> - Add a capture template option that allows errors to pass through to 
> the caller. Then a user may add %(...) expression with all necessary checks.

I am not very sure what kind of errors you are talking about. AFAIU, we
are talking about scenarios when user explicitly wants one of
PRIMARY/CLIPBOARD/SECONDARY, don't we?

>>> 2. Jean pointed out that Emacs ignores X selection if it does not have
>>> an X frame. It was almost buried in the haystack of "SQL instead of Org"
>>> noise. It happens if Emacs is running as a daemon. I am unsure which way
>>> Samuel starts Emacs (xinit, systemd user session, systemd socket
>>> activation, manually, etc.)
>> 
>> May we modify org-get-x-clipboard adding
>> (server-select-display (getenv \"DISPLAY\")) ?
>> Will it be enough?
>
> It will work for most of users, however sometimes it is incorrect. There 
> are a number of options that should have higher priority than emacs 
> server DISPLAY environment
> - emacsclient -display command line option,
> - emacsclient process DISPLAY environment,
> - emacs -display server argument.
>
> I have no idea how to get "emacs -display" argument and I need to look 
> into emacsclient source and server.el to figure out which way client's 
> display argument is passed to the server process.

Maybe it can be considered an Emacs bug or something to improve on Emacs
side?

-- 
Ihor Radchenko // yantar92,
Org mode contributor,
Learn more about Org mode at <https://orgmode.org/>.
Support Org development at <https://liberapay.com/org-mode>,
or support my work at <https://liberapay.com/yantar92>