extras_require and length of name > 5 fails.

[dfee]

I have a setup.py that looks like this:

from setuptools import setup, find_packages

setup(
    name='graphene_utils',
    version='1.0',
    packages=find_packages(),
    include_package_data=True,
    install_requires=[
        'graphene>=2.0.dev',
    ],
    extras_require={
        'sqlalchemy': [
            'sqlalchemy',
        ],
        'test': [
            'pylint',
        ],
    },
    entry_points='''
    [console_scripts]
    run=server:run
    ''',
)

When attempting pipenv install "-e .[sqlalchemy]" it fails with:

~/code/graphene_utils (virtualenvs/graphene_utils-Dw3J0BOS)
$ pipenv install "-e .[sqlalchemy]"
Courtesy Notice: Pipenv found itself running within a virtual environment, so it will automatically use that environment, instead of creating its own for any project.
Discovering imports from local codebase…
Installing -e .[sqlal…
⠸
Error:  An error occurred while installing -e .[sqlal!
.[sqlal should either be a path to a local project or a VCS url beginning with svn+, git+, hg+, or bzr+

However, renaming the extra_requires to sqla works fine.

[nateprewitt]

Yeah, this is definitely a bug. I haven't had a chance to look into it yet, but I'll try to set aside some time this weekend. If anyone else would like to pick this up before then, feel free.

[spott]

Are there any workarounds for this?

[nateprewitt]

We still haven't root caused the issue, so the only suggestion at this time is you'll have to use pip or a non-wheel (I wouldn't recommend with something like tensorflow).

[spott]

What are the consequences of using pip inside pipenv shell?

Does this just screw with the lock file?

[nateprewitt]

@spott yes, it will install but nothing will be added to the lockfile or Pipfile. This means you'd have to run this step on every install.

[spott]

@nateprewitt: I'll have to run which step? pipenv lock?

[nateprewitt]

Tensorflow won't be added to the Pipfile when you install it via pip. This means it will also not be included in the lockfile when locking.

You'll need to run pipenv shell and install Tensorflow via pip anytime you run pipenv install.

[spott]

@nateprewitt: Ah. Thanks. hopefully this bug will be fixed when I next need to do that.

[techalchemy]

@nateprewitt I've got the bug and part of the solution (it's the same as the local file installation issues, relating to passing relative paths to the Requirements library which can't parse anything with slashes) -- just need to fix the piece for markers as well. Edited because I read the actual issue-- will add that to todos