I didn't find such discussion in the SQL standard (maybe I have missed something).
My current thought is not to convert "rowtime" to upper-case is the best: 1. for those not column name case-sensitive database, it works. 2. for those case-sensitive database, assume users are aware of their sources that are case-sensitive, I think their intention to use descriptor will consider that factor. Converting column names to upper-case causes confusion. -Rui On Wed, Nov 20, 2019 at 6:55 PM Julian Hyde <jh...@apache.org> wrote: > If I write > > DESCRIPTOR(rowtime) > > and I am running Calcite’s parser in a mode that converts unquoted > identifiers to upper-case, do you think that it should convert “rowtime” to > upper-case? > > I am undecided. It’s possible that “rowtime” references a column in a > case-sensitive database. > > Julian > > > > On Nov 20, 2019, at 11:14 AM, Rui Wang <amaliu...@apache.org> wrote: > > > > Forgot to mention, the DESCRIPTOR support means the following query will > be > > able to run: > > > > SELECT * > > FROM TABLE(Tumble( > > TABLE ORDERS , > > DESCRIPTOR(ROWTIME) , > > INTERVAL '1' MINUTES)) > > > > > > Note that the second parameter to indicate the watermarked column is > > changed from a string(varchar) to DESCRIPTOR(col_name). > > > > > > -Rui > > > > > > On Wed, Nov 20, 2019 at 11:04 AM Rui Wang <ruw...@google.com> wrote: > > > >> I have an update: > >> > >> I tried to add DESCRIPTOR support (CALCITE-3339). Amazingly it was not > >> complicated at all to have a working version. So I created #1599[1], > which > >> is built on top of #1587[2], to demonstrate DESCRIPTOR. The PRs are > >> separated because DESCRIPTOR support currently is less mature. > >> > >> > >> [1]: https://github.com/apache/calcite/pull/1599 > >> [2]: https://github.com/apache/calcite/pull/1587 > >> > >> > >> -Rui > >> > >> On Thu, Nov 14, 2019 at 9:35 PM Rui Wang <amaliu...@apache.org> wrote: > >> > >>> Hi community, > >>> > >>> I have created a new PR ( https://github.com/apache/calcite/pull/1587) > >>> to demonstrate the progress of TUMBLE table value function > (CALCITE-3272). > >>> Julian suggested me to have a working version that adds a stream.iq > and > >>> have an enumerable implementation. Those are in the PR. > >>> > >>> High level speaking, the PR is adding a support of the following: > >>> > >>> SELECT * > >>> FROM TABLE(Tumble( > >>> TABLE ORDERS , > >>> 'ROWTIME' , > >>> INTERVAL '1' MINUTES)) > >>> > >>> > >>> One missing feature so far is adding support of DESCRIPTOR, which is > >>> intentionally cut off from the PR because that will make the PR more > >>> complicated. Thus DESCRIPTOR is left as future work. > >>> > >>> The PR solves not only CALCITE-3272, but also it's blockers: > >>> https://jira.apache.org/jira/browse/CALCITE-3340 > >>> https://jira.apache.org/jira/browse/CALCITE-3501 > >>> https://jira.apache.org/jira/browse/CALCITE-3499 > >>> https://jira.apache.org/jira/browse/CALCITE-3418 > >>> > >>> > >>> I will probably need some guidance on how to proceed to get the PR > >>> merged. Please let me know if you have any thoughts. > >>> > >>> -Rui > >>> > >> > >
