Feature request: Allow use of a custom PYTHONPATH

[realh]

Issue Prelude

Please complete these steps and check these boxes (by putting an x inside the brackets) before filing your issue:

• [x] I have read and understood YCM's [CONTRIBUTING][cont] document.
• [x] I have read and understood YCM's [CODE_OF_CONDUCT][code] document.
• [x] I have read and understood YCM's [README][readme], especially the [Frequently Asked Questions][faq] section.
• [x] I have searched YCM's issue tracker to find issues similar to the one I'm about to report and couldn't find an answer to my problem. ([Example Google search.][search])
• [ ] If filing a bug report, I have included the output of vim --version.
• [ ] If filing a bug report, I have included the output of :YcmDebugInfo.
• [ ] If filing a bug report, I have attached the contents of the logfiles using the :YcmToggleLogs command.
• [ ] If filing a bug report, I have included which OS (including specific OS version) I am using.
• [ ] If filing a bug report, I have included a minimal test case that reproduces my issue.
• [x] I understand this is an open-source project staffed by volunteers and that any help I receive is a selfless, heartfelt gift of their free time. I know I am not entitled to anything and will be polite and courteous.
• [x] I understand my issue may be closed if it becomes obvious I didn't actually perform all of these steps.

Issue Details

I've been struggling to get ycm to support GObject Introspection via fakegir. fakegir's build-jedi-cache.sh didn't seem to do enough to get it working so I created a simple wrapper script to add ~/.cache/fakegir to PYTHONPATH and set its path in g:ycm_python_binary_path.

It would be easier if YCM had another variable to override PYTHONPATH instead of having to create a wrapper script.

[vheon]

Wouldn't be a better approach to use a virtualenv? Looks like is mentioned in the README of fakegit as well on how to use it. Once you have the virtualenv you can either set g:ycm_python_binary_path to the Python binary inside the virtualenv or you can do it at runtime with YcmCompleter Restart <path_to_binary_of_virtyalenv>. I think this would be nicer instead of acting directly on the PYTHONPATH. But I'm not a Python dev so maybe @micbou knows better?

[realh]

fakegir's README also says that the virtualenv method doesn't seem to work with python3, which is what I want to use.

[micbou]

fakegir's README also says that the virtualenv method doesn't seem to work with python3, which is what I want to use.

Did you try? I was able to get some completions of the gi module by copying the ~/.cache/fakegir folder to the site-packages of a Python 3 virtualenv created with pyenv-virtualenv.

[puremourning]

The author of the fakegir script says:

The other option which I use is to create a virtualenv and copy the fakegir gi package into it:

mkvirtualenv fakegir cdvirtualenv cd lib/python*/site-packages cp -a ~/.cache/fakegir/gi .

This appears to be what @micbou has had success with. So, you could try that? Adding yet more options for a very specific use case is not really something we want to do. Closing this for now, as it seems there is a workaround. We can re-open if there really is a bug or a feature request that will benefit large numbers of users.