feat: Support multiple models in S3 batch export

[tomasfarias]

Problem

Users wish to export person information to allow identifying unique persons in their event batch exports.

Changes

This commit introduces a new batch_export_model input for all batch export destinations. This input is of type BatchExportModel which can be used to indicate which model a batch export is supposed to target. This opens the door to creating "persons" model exports, and eventually extending this further with more changes.

The change was done in a backwards compatible way, so batch_export_schema was not removed, and existing export configurations should continue to work (as evidenced by unit tests passing without changes).

However, moving forward all batch exports will be created with batch_export_schema set to None and batch_export_model defined in its place.

After updating existing exports, batch_export_schema and any other code associated with backwards compatibility can be deleted.

👉 Stay up-to-date with PostHog coding conventions for a smoother review.

Does this work well for both Cloud and self-hosted?

How did you test this code?

Updated all S3 unit tests to guarantee person batch export support. Follow-up PRs will work on other destinations.

This required generating person data, and checking it was successfully exported. To do this, new utilities were added under temporal/tests/utils/persons.py, they work similar to event utilities in temporal/tests/utils/event.py

Also, a bug was fixed in the test_s3_export_workflow_with_minio_bucket_and_custom_key_prefix test as it wasn't properly using custom keys.

[tomasfarias]

Err, ignore the last few test failures, they seem to be on github's side:

 livestream Error Get "https://ghcr.io/v2/posthog/livestream/manifests/sha256:748d94295ed87bad468c059edd756d8bbed16059156812f4d17c626aaddbf76a": net/http: TLS handshake timeout
Error response from daemon: Get "https://ghcr.io/v2/posthog/livestream/manifests/sha256:748d94295ed87bad468c059edd756d8bbed16059156812f4d17c626aaddbf76a": net/http: TLS handshake timeout
Error: Process completed with exit code 18.

I'll retry the tests.

[tomasfarias]

Hmm not sure why the hogql_query tests are failing, don't see any changes of mine around that part of the codebase. I'll investigate tomorrow.

[EDsCODE]

lgtm, nice pattern for extending beyond persons

[tomasfarias]

Removed this as locally was failing by builds. Shouldn't have been pushed though.

[tomasfarias]

Hey @fuziontech! You got tagged for review as I rebased this PR and commits from the previous PR you've reviewed popped up in this PR's git history.

Feel free to ignore, and apologies for the spam. Unless you are curious, then obviously comments are welcome!

[sentry-io]

Suspect Issues

This pull request was deployed and Sentry observed the following issues:

• ‼️ ClickHouseError: Code: 46. DB::Exception: Unknown table function events_batch_export: or incorrect parameterized ... posthog.temporal.common.clickhouse in check_res... View Issue
• ‼️ ClickHouseError: Code: 46. DB::Exception: Unknown table function events_batch_export_backfill: or incorrect param... posthog.temporal.common.clickhouse in check_res... View Issue
• ‼️ ClickHouseError: Code: 46. DB::Exception: Unknown table function events_batch_export: or incorrect parameterized ... posthog.temporal.common.clickhouse in check_res... View Issue

_{Did you find this useful? React with a 👍 or 👎}