Platform specific extras for dependency

[darsnack]

• Poetry version: 1.6.1
• Python version: 3.10.12
• OS version and name: macOS 13.5
• pyproject.toml: https://gist.github.com/darsnack/981ce7e407c549579e092ec81de6e21e

• I am on the latest stable Poetry version, installed using a recommended method.
• I have searched the issues of this repo and believe that this is not a duplicate.
• I have consulted the FAQ and blog for any relevant entries or release notes.
• If an exception occurs when executing a command, I executed it again in debug mode (-vvv option) and have included the output below.

Issue

I am primarily having an issue with the dependency:

deeplabcut = [
    {version="~2.3", extras=["gui", "tf"], markers="sys_platform!='darwin'"},
    {version="~2.3", extras=["gui", "apple_mchips"], markers="sys_platform=='darwin'"}
]

My expectation is the the apple_mchips extra is only installed on darwin (and the opposite for tf). Instead, I get that all three extras: gui, tf, and apple_mchips are installed.

#5506 is listed as a duplicate in related issues, but I don't understand exactly how that is the case. Can the use of markers be clarified here? Why are these specifications being merged?

[dimbleby]

I do not reproduce. with that dependency and on an ubuntu VM:

$ poetry install --dry-run | grep tensorflow
  • Installing tensorflow-estimator (2.10.0)
  • Installing tensorflow-io-gcs-filesystem (0.33.0)
  • Installing tensorflow (2.10.0)

note no tensorflow-macos or tensorflow-metal, which are the things provided by the apple_mchips extra.

[darsnack]

You are right, I was confused about poetry show --tree deeplabcut which displays both tensorflow and tensorflow-macos. I guess poetry show is based on the lock file which is platform agnostic, but the documentation for poetry show --all seems to suggest the poetry show is platform-specific.

My downstream errors are unrelated to poetry.

[github-actions]

This issue has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.