Unsure what those situations are, but when it happens it breakes ein's undo
facility. Calling `length' on a dotted list will generate wrong-argument-type
errors. To help soften the consequences of those situations we call `safe-length'
instead of `length'.
* Better consistency with kernel connections in anonymous ob-ein blocks.
And some tests to try to prove this is the case.
* Make sure company-mode is enabled.
* Undo cursor shifts.
* Company mode still not catching.
* Travis unhappy with undo and anonymous completion.
Tests work locally, but fail on travis. Let's play around a bit.
* Revert undo position jiggers.
I think the issue is character or line ending encoding in WSL (i.e. Windows) and
a "normal" Linux setup.
* ein-ac: Fail silently for unexpected ein:completion-backend values.
Fixes#608, though we really need a test cases for completion in polymode.
* Only define ein ac-sources when the user explicitly sets the backend.
* Test case for issue #608.
* Only test jedi completion when testing polymode.
* Add epc dependency for emacs-jedi.
* Flanging up with travis is always a challenge.
* Forgot to update the scenario.
* Give jedi time to install.
* Add virtualenv to travis OS X build.
Also add some addition TOX environments for future testing cases.
* OS X travis is failing - is it a timeout issue?
Update help message in travis as well. Get rid of case-fold check for the
moment. It works fine locally, but travis can't seem to wrap its head around the
test.
Hard to reproduce exactly what is happening, but try to work around the
possibility of a company completion popup appearing in the buffer or minibuffer.
Unfortunately, no good way to test font-lock as font-lock-mode won't
activate under noninteractive ecukes (can try --no-win or --win, but
won't fly under travis)
In practice, polymode font-lock bugs often manifest themselves much as
issue #456 did. The rear-nonsticky property on the input prompt gets
washed away by font-lock and brokenness ensues.
Some users run jupyter via `jupyter-notebook` instead of `jupyter`,
and break because we unilaterally impose the "notebook" subcommand.
Those users can now:
```
M-x customize-group ein
Jupyter Server Use Subcommand: Omit
```
Currently, an org block that has no output (e.g., `foo = 5`) never
clears the `[....]` upon block completion.
Also, no `:session` should default to localhost.
Before: Exit emacs with modified notebooks. Be asked "You have
unsaved notebooks. Discard changes?" Respond no, and manually figure
out which notebooks haven't been saved.
After: Exit emacs with modified notebooks. Be asked "Save A?".
Respond yes. Quit without further ado.
Also, bugfix killing of buffers in polymode. Must kill host buffer,
not indirect buffers.
Initial SIGTERM prompts server to dutifully shutdown kernel, but
another SIGTERM necessary to kill server. Perhaps ipython changed
since the last time I tested this.
Ipython doesn't send a display_data message upon initial plot
request. This is probably our bug, or is it? Document this bug in an
ecukes test.
Also enforce emacs version >= 25
M-x customize-group RET ein
Toggle Ein:Polymode
Avoid trying to emulate jump-to-definition, eldoc, and autocompletion
functionalities that Elpy will always do better.
Fixes#497#482#418
* Testing for shared eval and connecting buffers.
* Forgot the feature.
* Need a small delay to make sure code is evaluated.
* Testing for shared eval and connecting buffers.
* Forgot the feature.
* Need a small delay to make sure code is evaluated.
* Add scenario for company completion in a connected buffer.
* Works better if we test completion first.
* Update ipython version tested in travis.
* Stab in the dark to fix travis erroring under ipython 6.x
* Revert "Stab in the dark to fix travis erroring under ipython 6.x"
This reverts commit 7255d31fdb.
We can't be certain when the shared output cell can be safely turned
over to the next org block. I have witnessed `display_data` updates being
received *after* both the `execute_result` and `execute_reply`
messages. As a HACK, we `run-at-time 2` seconds the turning over of
the cell to accommodate any eleventh hour straggler display_data's.
In light of #425 and #456, provide more guidance on curl errors.
Also, clean up eldoc stub. Leveraging elpy's eldoc perfection seems
the only sensible choice, however.
