I like your approach to divide the work into moderate-sized steps. Clear descriptions of those steps help those of us who are overwhelmed with email (all of us!) track the situation. The first step seems to be ’Support TIMESTAMP and TIME WITH TIME ZONE in the parser’. (I’ve reversed your current phrasing, ‘Parser does not…’ which makes a missing feature sound like a bug.)
After this email thread, let’s move the conversation to https://issues.apache.org/jira/browse/CALCITE-208. Jira is a better medium for long-running conversations. As the type names are so verbose, I recommend that we all adopt Snowflake’s abbreviations: TIMESTAMP_NTZ, TIMESTAMP_LTZ, TIMESTAMP_TZ, TIME_NTZ, etc. Be very careful to avoid vendor-specific terms that do not match the SQL standard (e.g. BigQuery’s TIMESTAMP type is a TIMESTAMP_LTZ, i.e. an instant; call it “BigQuery TIMESTAMP” or “TIMESTAMP_LTZ” but never “TIMESTAMP”). I have a feeling that in order to support TIMESTAMP_TZ and TIME_TZ we will need a notion of the ‘current session time zone’. Can you check whether we have adequate support for that? If not, it could be a separate task. Julian > On Dec 15, 2023, at 11:04 AM, Mihai Budiu <mbu...@gmail.com> wrote: > > Hello all, > > There is a feature open for Calcite to support time zones > https://issues.apache.org/jira/browse/CALCITE-208 > It has been open for 9 years. > > I have submitted last week a PR which starts to address this issue: > https://github.com/apache/calcite/pull/3569 > I would appreciate reviews on this PR. > > This will be a lot of work, and I don't think we want it solved in a monster > PR. > I plan to do at least some of the follow up work as well once this is merged. > > I notice that it's not easy to get reviews on PRs that touch many parts of > the code. I understand, myself I am reluctant to review code that touches > parts I am not familiar with. I will appreciate any advice on making progress > with similar contributions. > > Mihai