zeroshade commented on issue #2843: URL: https://github.com/apache/arrow-adbc/issues/2843#issuecomment-2997597232
Okay, I did a small test: I created a polars dataframe, wrote a parquet file and then followed the same strategy we use to insert the file into a snowflake table. After the insertion, the data was the correct time instants, but still displayed in my local session timezone just as what happened in the original issue here. It was still treated appropriately as a `TIMESTAMP_TZ` though, if I changed my session timezone it still outputs the results in my original timezone. I did a bit of digging and found this:  You'll notice that when `USE_LOGICAL_TYPE = TRUE` a `TimestampType(isAdjustedToUTC=true, ....)` gets converted to `TIMESTAMP_LTZ`. This happens before it is ultimately converted to `TIMESTAMP_TZ`, leading to what you're seeing. The raw data that is being held is still the correct value under the hood (because `TIMESTAMP_TZ` and `TIMESTAMP_LTZ` both store the value in UTC internally), the difference appears to solely be in how it is displayed as far as I can tell. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
