Wow, I'd never seen that crosstab() function before! TIL.

https://www.postgresql.org/docs/current/tablefunc.html#id-1....

This raises a problem with the YAML DSL though. Consider the following (truncated) example:

    main:
      select:
      - field: ct.*
      from:
        alias: ct
        fields: ...
        crosstab:
        - select:
          - extra_infos.problem_id
          - extra_infos.info_type
          - extra_infos.info_value
          from: extra_infos
        - select:
          - extra_infos.info_type 
          distinct: true
      from: extra_infos

Inside the "from:" area, those alias and fields and crosstab keys are all at the same level.

But... "alias:" and "fields:" correspond to features of the from clause itself, whereas "crosstab:" is a special custom thing (a macro?) which ends up being compiled into a call to the identically-named crosstab() feature.

So to understand the YAML, I need to understand both how the crosstab() feature in PostgreSQL works AND how that YAML DSL represents "crosstab:" calls using a different syntax.

I looked up crosstab in the PostgreSQL manual but that information won't help me figure out the DSL.

So I think I'd rather stick with regular SQL.