Error when using relative date query in metabase 0.34.0

[skuroq]

Hi, first, thanks for the driver :)

While testing out the new Metabase 0.34.0 Version i stumbled upon an error when doing relative date queries, for example when filtering 'today' on a date field. I receive this error when using your provided jar from the release section:

I tried rebuilding the athena driver from source. The resulting jar works fine when using metabase 0.33.7.3 but again i get an error in 0.34.0 but a different one this time:

I quickly looked at the metabase source code and saw that in the release a lot a work was done concerning timezones and dates. Also the structure of the files changed, for example metabase/util/date.clj is now called metabase/util/date_2.clj. Maybe this has something to do with this?

[dacort]

Hi @skuroq - thanks for the issue report! There was indeed a lot of time-related work in 0.34 release, so it's likely something changed. I'll try to take a look sometime in the next few days and see if I can figure out what changed.

[dacort]

Yea, looks like that unprepare-date-with-iso-8601-fn got removed. The pre-existing Presto driver has some indication of how it was changed, so modeling that I'm able to get these type of queries working again.

For reference, here's a sample table for testing relative date queries:

CREATE TABLE default.mb_relative
WITH (
  format='PARQUET'
) AS
WITH source_data AS (
  SELECT 0 AS "id", 'damon' AS "name", localtimestamp AS "timestamp"
  UNION ALL
  SELECT 1, 'bob', localtimestamp - INTERVAL '2' DAY 
)

SELECT * FROM source_data

As mentioned in #27 I wanted to see if I can get some basic tests running in order to better catch/validate things like this in the future. I'll try to get a new release of this packaged up early next week.

[skuroq]

Thanks for looking into it @dacort !

[dacort]

As an FYI, I've managed to get integrated Metabase tests working in the feature/integration_tests_v2 branch. There are some failures to resolve, though, so I probably won't let that block rolling out a fix for 0.34. I'll try to get that pushed out tomorrow.