nvim-ipy
nvim-ipy copied to clipboard
bfredl
Integrating with supertab/default completion
One of the 'problems' I have is that I can add a completion to a shortcut (i.e., change from C-F), but I cannot use vim's default completion with nvim-ipy. In turn, this means I cannot use it with Supertab. Is there a way around this?
If you enable both, what happens? does supertab overwrite the suggestion list that nvim-ipy provides?
It's not an issue with enabling both. I'd like <TAB> to pull the suggestion list from nvim-ipy instead of supertab completions, which are useless to me. Context: I don't ever use jedi-vim or its derivatives because they break on complicated modules like pandas; whereas I can get completions via evaluating code with nvim-ipy, but I'd like to get them via a tab binding.
It should be doable. The problem is that <Plug>(IPy-Complete) is async, so it might not work to just tell SupterTab to use it. "Worst case" would be to just implement a synchronous variant that could be used as &omnifunc. I will look into it.
Thank you for your hard work. This has become an indispensable vim plugin in my workflow.
I took a stab at implementing an omnifunc. So with set omnifunc=IPyOmniFunc it should work if SuperTab is set to use omnifunc. But for an async framework like Deoplete using it is not recommended, I will try implement a Deoplete source as well.
Thanks for taking a look at this. I can't get it to work, although I'm not sure why that is. Doing:
:verbose set omnifunc
gives me:
omnifunc=IPyOmniFunc
Last set from ~/.config/nvim/init.vim
Notes: Attempt is with python3 (renaming directory from python to python3 in rplugin/) and updating plugin. Attempting to pull up omnifunc without supertab also doesn't appear to work. It could be on my end, I'll experiment a bit more soon.
I took a stab at implementing an omnifunc. So with
set omnifunc=IPyOmniFuncit should work if SuperTab is set to use omnifunc. But for an async framework like Deoplete using it is not recommended, I will try implement a Deoplete source as well.
Is there any update on the Deoplete source functionality? For me the set omnifunc=IPyOmniFunc works quite well, but I'm not sure how to specify it as an input for deoplete.
