feat: add a basic async python client starting point #1014
Conversation
westonpace
force-pushed
the
feat/async-python
branch
2 times, most recently
from
b76ffae
to
dbd316e
Compare
.github/workflows/pypi-publish.yml
Outdated
| name: PyPI Publish | ||
| name: Build and upload python wheels |
There was a problem hiding this comment.
For continuity of the jobs, I think it would be nice to keep the name the same. We already have enough dead GA names.
There was a problem hiding this comment.
I've reverted this change so the workflow will still be the same. Do dead GA names matter for workflows? Or jobs? The job name has to change because its splitting one job into three.
There was a problem hiding this comment.
I think workflows. I'm mostly thinking of this left sidebar on https://github.com/lancedb/lancedb/actions
There some some dead names there (e.g. "Node Release"), which I find annoying.
| region: Optional[str], | ||
| host_override: Optional[str], | ||
| read_consistency_interval: Optional[float], | ||
| ) -> Connection: ... |
There was a problem hiding this comment.
Doesn't this return a Future?
There was a problem hiding this comment.
I changed it to async def which should imply the future. I'm on the fence about these having these pyi files at all and curious for your opinion. These methods aren't exposed to the user so this is only for our own internal understanding.
There was a problem hiding this comment.
For internal only things, I don't think we have to write them. Sometimes we might publicly expose them though, if the Python wrapper is redundant.
If I add actual mypy checks to the repo, my opinion on this might change.
westonpace
force-pushed
the
feat/async-python
branch
from
0bdab9b
to
a2d818b
Compare
This changes `lancedb` from a "pure python" setuptools project to a maturin project and adds a rust lancedb dependency. The async python client is extremely minimal (only `connect` and `Connection.table_names` are supported). The purpose of this PR is to get the infrastructure in place for building out the rest of the async client. Although this is not technically a breaking change (no APIs are changing) it is still a considerable change in the way the wheels are built because they now include the native shared library.
This changes `lancedb` from a "pure python" setuptools project to a maturin project and adds a rust lancedb dependency. The async python client is extremely minimal (only `connect` and `Connection.table_names` are supported). The purpose of this PR is to get the infrastructure in place for building out the rest of the async client. Although this is not technically a breaking change (no APIs are changing) it is still a considerable change in the way the wheels are built because they now include the native shared library.
This changes `lancedb` from a "pure python" setuptools project to a maturin project and adds a rust lancedb dependency. The async python client is extremely minimal (only `connect` and `Connection.table_names` are supported). The purpose of this PR is to get the infrastructure in place for building out the rest of the async client. Although this is not technically a breaking change (no APIs are changing) it is still a considerable change in the way the wheels are built because they now include the native shared library.
This changes
lancedbfrom a "pure python" setuptools project to a maturin project and adds a rust lancedb dependency.The async python client is extremely minimal (only
connectandConnection.table_namesare supported). The purpose of this PR is to get the infrastructure in place for building out the rest of the async client.Although this is not technically a breaking change (no APIs are changing) it is still a considerable change in the way the wheels are built because they now include the native shared library.