Chr. von Stuckrad <[email protected]> wrote:
> Am 01.09.2018 um 15:30 schrieb Bram Moolenaar:> > ... the timeouts
> longer?
> >
> > Which test is that? Ideally we would have some way of measuring how
> > slow the machine is and adjusting to that. Or we can just increase
> > the
> > timeout. That is not a good idea when it's supposed to timeout, so we
> > need to know what to change.
> The test is 'test_quotestar.vim' and seems to ask for the X-Buffers.
> I should have mentioned, all happened on a virtually small debian
> 'workstation' running in an vmware-cluster for tests.
> ------------------------------- output was -----------------------------
> From test_quotestar.vim:
> Found errors in Test_quotestar():
> First run:
> Caught exception in Test_quotestar(): WaitFor() timed out after 5000
> msec @ function
> RunTheTest[40]..Test_quotestar[11]..Do_test_quotestar_for_x11[56]..WaitFor,
> line 4
> Second run:
> Caught exception in Test_quotestar(): WaitFor() timed out after 5000
> msec @ function
> RunTheTest[40]..Test_quotestar[11]..Do_test_quotestar_for_x11[56]..WaitFor,
> line 4
> TEST FAILURE
> ------------------------------------------------------------------------
>
> Testing the buffers from remote to my display with 'xclip' works
> perfectly without undue delays.
> BUT as I found out today, my problem seems to be, that X11 may be broken
> by missing some dependencies(?). The GUI (auto) was ignored but the test
> still is tried. May be the test is not skipped, because I configured
> with --with-x(?) or is there 'with x but without GUI'?
X can be used without any gui i.e. when configuring
with --enable-gui=none as X can be used for copy/paste.
Now I don't know why the test failed.
> >> Another test fails, seemingly because there is no
> >> textual representation of my passwd entry. The machine
> >> ...
> >> 'getent passwd $USER' still works, but with the old
> >> 'grep $USER /etc/passwd' nobody will be seen.
> >> ...
> > What is the log output for this (in src/testdir/messages)?
> ------------------------------- output was
> -----------------------------From test_cmdline.vim:
> Found errors in Test_cmdline_complete_user_names():
> function RunTheTest[40]..Test_cmdline_complete_user_names line 8:
> Pattern '^"e \\~.*\\<stucki\\>' does not match '"e ~salt saned sddm snmp
> speech-dispatcher sshd sssd statd sync sys systemd-bus-proxy
> systemd-network systemd-resolve systemd-timesync'
> TEST FAILURE
> Makefile:41: recipe for target 'report' failed
> make[1]: *** [report] Error 1
> make[1]: Leaving directory '/home/stucki/GIT/vim/src/testdir'
> Makefile:2083: recipe for target 'scripttests' failed
> make: *** [scripttests] Error 2
> ------------------------------------------------------------------------
>
> If I then add a line "stucki:*:####:####:c v. Stuckrad:/home/stucki"
> (with #### the correct numbers of course) to '/etc/passwd' and rerun
> with 'make test' all tests pass. As I mentioned; 'getent passwd stucki'
> shows the correct line in all cases. 'sssd' seems to be some modern
> expansion of the name service switch(ing) services, which allows e.g.
> laptops to cache logins from the last connection for a while to be
> abroad without network and continue to login locally.
The test looks fine here. It's Vim which does not
complete to the valid user name 'stucki' when
doing :e ~s<CTRL-A> if I understand well.
Function init_users() in misc1.c:4724 finds all
user names. It uses getpwent(). Apparently that
does not work in your setup. You say that
"getent passwd stucki" works. "man getent'
says:
=== BEGIN QUOTE "man getent" ===
passwd When no key is provided, use setpwent(3), getpwent(3),
and endpwent(3) to enumerate the passwd database.
When one or more key arguments are provided, pass each
numeric key to getpwuid(3) and each nonnumeric key to
getpwnam(3) and display the result.
=== END QUOTE ===
So the getent command also uses getpwent()
just like Vim when doing "getenv passwd" but
uses getpwuid() when doing "getent passwd stucki".
So I suppose that user name "stucki" is not shown
eithe when doing this from the shell:
$ getent passwd
But it is shown when doing:
$ getenv passwd stucki
Can you confirm?
It looks like we could improve function init_users() in
misc1.c by adding an entry for getpwnam(name)
where name is the value of the env variable $USER,
in case it's not already found by getpwent().
Or is there a better way to find all user names for
completion in init_users()?
Dominique
--
--
You received this message from the "vim_dev" maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php
---
You received this message because you are subscribed to the Google Groups
"vim_dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/d/optout.
