[Bug]: symbols are not exported from apache_beam, causing type errors from Pylance

[sebastian-correa]

What happened?

When doing

from apache_beam import Pipeline

Pylance complains about:

"Pipeline" is not exported from module "apache_beam". Import from "apache_beam.pipeline" insteadPylance[reportPrivateImportUsage (https://github.com/microsoft/pyright/blob/main/docs/configuration.md#reportPrivateImportUsage)

I think that that (and other symbols) imported in the packages __init__.py are meant to be part of the public interface. However, Pyalnce requires the following semantics to detect them as such:

Imported symbols are considered private by default. If they use the “import A as A” (a redundant module alias), “from X import A as A” (a redundant symbol alias), or “from . import A” forms, symbol “A” is not private unless the name begins with an underscore (...).
A module can expose an __all__ symbol at the module level that provides a list of names that are considered part of the interface. The __all__ symbol indicates which symbols are included in a wildcard import. All symbols included in the __all__ list are considered public even if the other rules above would otherwise indicate that they were private.

[source]

I propose that either redundant aliases or an __all__ be added to apache_beam's __init__.py to prevent this.

Probably the transforms module, which is * imported will need this done to it as well (and, well, recursively as transforms also imports other modules).

Issue Priority

Priority: 3 (minor)

Issue Components

• Component: Python SDK
• Component: Java SDK
• Component: Go SDK
• Component: Typescript SDK
• Component: IO connector
• Component: Beam YAML
• Component: Beam examples
• Component: Beam playground
• Component: Beam katas
• Component: Website
• Component: Infrastructure
• Component: Spark Runner
• Component: Flink Runner
• Component: Samza Runner
• Component: Twister2 Runner
• Component: Hazelcast Jet Runner
• Component: Google Cloud Dataflow Runner

[sebastian-correa]

.take-issue

[sebastian-correa]

I'll tackle the issue, but will wait for some from the Beam team to confirm this is something they want and to voice an opinion on redundant aliases (my preference) vs __all__. This way, I don't open an unwanted PR.