Operations on original columns ​​in SQL-SELECT-JOIN query from different databases using the "superset://" connection

[mastavichus-vp]

Is there any way to transform the original column values ​​in queries from a several different databases by from type super-connection "Superset meta database" with SQLAlchemy URI "superset://"?
First of all, necessary conversion to a string type is necessary for performing comparisons in a query with JOIN.

In superset_config.py:

FEATURE_FLAGS = {
  ...
  ENABLE_SUPERSET_META_DB=True,
  ...
}

Tried syntaxes unsuccessfully,

postgres

select to_char(a.dt, 'YYYY-MM-DD') as dt2, a.* from "pg1.synth.t1" a;
select a.dt::text as dt2, a.* from "pg1.synth.t1" a;

jinja
select {{ a.dt }} as dt2, a.* from "pg1.synth.t1" a;

python
select str(a.dt) as dt2, a.* from "pg1.synth.t1" a;

javascript
select (a.dt).toString() as dt2, a.* from "pg1.synth.t1" a;

[mastavichus-vp]

The SQLAlchemy function works fine.

select SUBSTR(CAST(a.dt AS CHAR),1,10) as dt2, a.* from "pg1.synth.t1" a;

Sorry