Question regarding independent timestamp variables support

[cliuj]

Hi there,

I have a Cassandra schema where the time column is broken down into various columns identical to:

CREATE TABLE IF NOT EXISTS temperature (
    year int,
    month int,
    day int,
    minute,
    sensor_id int,
    temperature int,
    PRIMARY KEY (sensor_id)
);

Is it possible for the current Query Editor to support this format? Or would it be a good idea to add a feature to support this?

[HadesArchitect]

Hey @unflag , that should be possible with the #167 done?

[unflag]

No, datasource makes no assumptions about columns meaning, thus it cannot create timestamp by itself using set of int columns. C* has its own tools to do such things. It should be done by user - for example, converting that values to seconds and passing to dateOf(maxTimeuuid()) functions. Or by using some user-defined function.

[HadesArchitect]

Yep, that's exactly what I mean. A user could do some computations or UDFs to get the first column combined of other columns, could he?

[unflag]

Certainly. If I were that user I would use UDF:) It is a quite powerful feature - basically, embedding Java/JavaScript code in cql queries.