If the functions you implement work within the supported range of Calcites DATE type, that is sufficient.
Changing the range of the DATE type would be an orthogonal change. I’m not sure if anyone is asking for that. > On Aug 28, 2024, at 8:09 AM, Cancai Cai <caic68...@gmail.com> wrote: > > I tried many ways to write it. As Julian said, Maybe BigQuery's DATE data > type has a different range than Calcite's. I am not sure if I can fix this > problem. > > If I adapt some other date functions of Spark in the future, do I need to fix > this problem first? For example, > https://issues.apache.org/jira/browse/CALCITE-6527 > > Best wishes, > Cancai Cai > >> 2024年8月27日 05:11，Norman Jordan <norman.jor...@improving.com.INVALID> 写道： >> >> Re https://issues.apache.org/jira/browse/CALCITE-6544. Is it possible that >> BigQuery’s DATE_ADD can return BC dates because BigQuery’s DATE data type >> has a different range of allowable values than Calcite’s? In other words, >> it’s not a difference with the DATE_ADD function. >
