Mercurial > vim
view src/testdir/test_terminal_fail.vim @ 15054:2d6e930c7613 v8.1.0538
patch 8.1.0538: evaluating a modeline might invoke using a shell command
commit https://github.com/vim/vim/commit/5958f95a40a4a44bd9e7f3b7ec6554a6ef3e42ca
Author: Bram Moolenaar <Bram@vim.org>
Date: Tue Nov 20 04:25:21 2018 +0100
patch 8.1.0538: evaluating a modeline might invoke using a shell command
Problem: Evaluating a modeline might invoke using a shell command. (Paul
Huber)
Solution: Set the sandbox flag when setting options from a modeline.
author | Bram Moolenaar <Bram@vim.org> |
---|---|
date | Tue, 20 Nov 2018 04:30:07 +0100 |
parents | e9dbdc4d8279 |
children | f38fcbf343ce |
line wrap: on
line source
" This test is in a separate file, because it usually causes reports for memory " leaks under valgrind. That is because when fork/exec fails memory is not " freed. Since the process exists right away it's not a real leak. if !has('terminal') finish endif source shared.vim func Test_terminal_redir_fails() if has('unix') let buf = term_start('xyzabc', {'err_io': 'file', 'err_name': 'Xfile'}) call term_wait(buf) call WaitFor('len(readfile("Xfile")) > 0') call assert_match('executing job failed', readfile('Xfile')[0]) call WaitFor('!&modified') call delete('Xfile') bwipe endif endfunc