Problem: When using complete() and typing a character undo is saved after
the character was inserted. (Shougo)
Solution: Save for undo before inserting the character.
d56a79d339
The "technically correct" interpretation is to execute the first line
that is seen (and this is what happens on middle-click paste in Vim).
^M is only intended to "defuse" the newline, so the user can review it.
The parent commit changed the behavior to insert <Space> between lines,
but that's a higher-risk change: it is arguably possible that some user
*wants* the literal ^M chars when e.g. assigning to a register:
:let @a='<C-R>b'
To avoid that risk, keep the old behavior and only omit the last ^M.
This makes `yy:<C-R>0` nicer at no cost.
^M isn't any more "correct" than space: the "technically correct"
interpretation is to execute the first line that is seen (and this is
what happens on middle-click paste in Vim). ^M is only intended to
defuse the newline, so that the user can review the command. We can do
that with a space instead, and then the command can be executed without
having to fix it up first.
These are just blobs that we jammed into the package. find_program() and
WindowsDllCopy.cmake do not make sense here, they search include paths
and try to determine DLL dependencies (GetPrerequisites).
The exists('g:loaded_foo') convention is rather common, and may be
relied upon in some cases. It's also very unlikely that a user or plugin
has any reason to set g:loaded_foo to zero, so the principle of least
surprise can be brushed aside here.
https://github.com/neovim/neovim/issues/6107#issuecomment-279532143
- Do not exclude any directories from `find` search, remove dumps before tests
instead.
- Install `apport` on travis so that linux tests should produce core dumps
(based on information from travis-ci/travis-ci#3754, not sure whether it still
applies).
- Check cores in lua so that one has an idea which test is failing exactly. Do
this only 10% of time on linux because traversing the file system is slow.
Unit tests are still not touched, though it is what `app` argument in
`check_cores` is for.
TODO? consider using `find`, it may be faster. Consider retiring `os.execute`,
dealing with escaping is bad.
We must invoke src/nvim/testdir/Makefile directly.
Explained in 3d1084f264:
> Running tests from the top-level Makefile will use the third-party
> dependencies from .deps instead of the ones from the Travis cache.
If we could run `oldtest` with CMake, we would not need to do this. Need
USES_TERMINAL feature (CMake 3.2+) for that.
- Add support for TEST_FILE to the `oldtest` target, for consistency
with the busted/lua tests.
Caveat: with the busted/lua tests TEST_FILE takes a full path, whereas
for `oldtest` it must be "test_foo.res".
- Add support for NVIM_PRG, again so that all test-related targets are
consistent.
- Use consistent name for NVIM_PRG. But still need to support NVIM_PROG
for QuickBuild CI.
Note: The `oldtest` target is driven by the top-level Makefile, because
it requires a TTY. CMake 3.2 added a USES_TERMINAL flag to
add_custom_target(). But we support CMake 2.8...
add_custom_target(oldtest
COMMAND make clean
COMMAND make NVIM_PRG=$<TARGET_FILE:nvim> $ENV{MAKEOVERRIDES}
DEPENDS nvim
WORKING_DIRECTORY "${CMAKE_CURRENT_SOURCE_DIR}/src/nvim/testdir"
USES_TERMINAL true
)
