Test multiple python versions & revert python 3.10 not being supported #31
Conversation
Signed-off-by: David Black <dblack@atlassian.com>
…n 3.10 & a specific version of numpy being pulled in by pyarrow. Signed-off-by: David Black <dblack@atlassian.com>
dbaxa
requested review from
susodapop,
arikfr,
moderakh and
yunbodeng-db
as code owners
This reverts commit c8ab431.
| @@ -11,8 +11,8 @@ include = ["CHANGELOG.md"] | |||
| [tool.poetry.dependencies] | |||
| python = "^3.7.1" | |||
| thrift = "^0.13.0" | |||
| pyarrow = "^5.0.0" | |||
| pandas = "^1.3.0" | |||
There was a problem hiding this comment.
IMHO this library should drop support for python 3.7 & the update pandas minimum version to 1.4.3.
There was a problem hiding this comment.
Understandable. Because many Databricks users connect from older Pythons we prefer to maintain as much backwards compatibility as possible. For your use-case, is there any reason you cannot install a newer pandas? Or does databricks-sql-connector use pandas methods that were deprecated in newer versions?
There was a problem hiding this comment.
No -^ it isn't an issue for us. It is rather that many libraries have dropped support for python 3.7 & you may wish to update pandas to a newer version at some point which will mean dropping support for python 3.7.
dbaxa
changed the title
|
You can see test execution at https://github.com/dbaxa/databricks-sql-python/runs/7837469701. |
|
cc @susodapop |
|
I'm running the e2e test suite against this change now. |
|
All e2e tests pass in the following Python versions:
|
* Test with multiple python versions. * Update pyarrow to version 9.0.0 to address issue in relation to python 3.10 & a specific version of numpy being pulled in by pyarrow. Closes databricks#26 Signed-off-by: David Black <dblack@atlassian.com>
Hi, we are making use of this library and have for some time been using it in combination with newer versions of
pandas(1.4.3) &numpy(1.23.1), and pyarrow (8.0.0, however version9.0.0seems fine). While version 2.0.2 of this library works for us ... version 2.0.3 does not due to the changes to the dependency specification which break its usage in python 3.10.This fixes #26 .