Fix datetime conversion with mixed timezones when ignore_tz is False

[mreiche]

When querying multi symbols at the same time, the Pandas Timestamp to datetime64 conversion will fail, when utc=False.
It looks like this is the most stable implementation, because Pandas suggests to use this feature in the future.
Downside: The original timezones are lost. But every datetime has now tz=timezone.utc
Upside: ignore_tz may be obsolete.

I added a test for testing Pandas datetime conversion as well.

[ValueRaider]

Downside: The original timezones are lost. But every datetime has now tz=timezone.utc

That's fine as long as 9:30am EST is aligned with 2:30pm EU (or whatever the Atlantic offset is). ignore_tz means 9:30am EST aligned with 9:30am EU.

[mreiche]

Downside: The original timezones are lost. But every datetime has now tz=timezone.utc

That's fine as long as 9:30am EST is aligned with 2:30pm EU (or whatever the Atlantic offset is). ignore_tz means 9:30am EST aligned with 9:30am EU.

That's covered by the test TestPandas.test_mixed_timezones_to_datetime() where the datetime64 utc timestamps are compared to the tz-aware Timestamps.

[ValueRaider]

Approved. Just squash the commits #1084

[mreiche]

Approved. Just squash the commits #1084

Can't you just enable squash commits in your repo? https://docs.github.com/de/repositories/configuring-branches-and-merges-in-your-repository/configuring-pull-request-merges/configuring-commit-squashing-for-pull-requests

[ValueRaider]

Hmm. Could, but then the squashed commit message lists each individual commit e.g. Add assertion for timezone. Why would you not want to rewrite the message?