You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I've noticed that passing calling Consumer.commit(msg=None) is a valid case. At least it's being handled in the library's code separately. I'm wondering what's the difference between doing that and calling Consumer.commit with the last read message for the "msg" argument.
It looks like it does pretty much the same thing, but then why the argument is always passed explicitly in all of the tests in examples/integration_test.py?
The argument-less commit() will commit all uncommitted offsets for the current assignment.
The Python integration tests are primarily to verify the APIs with a live broker environment, the actual kafka client tests reside in librdkafka's build tree and are much more detailed.
librdkafka (using this Python client) also runs the full standard Apache Kafka client system test suite in a nightly job.
I've noticed that passing calling Consumer.commit(msg=None) is a valid case. At least it's being handled in the library's code separately. I'm wondering what's the difference between doing that and calling Consumer.commit with the last read message for the "msg" argument.
It looks like it does pretty much the same thing, but then why the argument is always passed explicitly in all of the tests in examples/integration_test.py?
https://stackoverflow.com/questions/47216827/confluent-kafka-python-producer-commit-with-no-parameters
The text was updated successfully, but these errors were encountered: