David E. Wheeler comments

Results 436 comments of


                                            David E. Wheeler

create table_constraint_checksum function for pg_tapgen

Looks okay to me, though I agree that it could get annoying if the output of `pg_get_constraintdef` changes between releases.

create table_constraint_checksum function for pg_tapgen

Agreed, but maybe check with [pgsql-hackers](https://www.postgresql.org/list/pgsql-hackers/).

`has_pk(:schema, :table)` Not defined

Oh well perhaps we're missing `has_pk(:schema, :table)`. If we had it you could make sure it's called instead of `has_pk(:schema, :desc)` by specifying the type: ``` sql SELECT has_pk('utility', 'eqpt_model_info'::name);...

`has_pk(:schema, :table)` Not defined

> Would that fill in the missing :desc like SELECT has_pk('eqpt_model_info');? Yes, exactly.

Improve handling of types

No, I think that would be fine, as long as we updated `compat/install-8.4.patch` to switch it back to `NAME[]`.

Improve handling of types

pgTAP has patch files for support back to 8.1. I dropped 8.0 back in Nov 2011, so I'd say we could start dropping the other 8.x versions. Hasn't really been...

Change pgtap_version() from numeric?

PGXN supports only semantic versions, so 1.001 is not allowed.

Change pgtap_version() from numeric?

Sorry, was reading too quickly. Please don't change its return value, that would break compatibility. The value is set by the `$NUMVERSION` variable in the Makefile. I think the idea...

Any way to check VIEW privileges similar to table_privs_are?

Sure, makes sense. Care to submit a pull request?

trigger_is(): allow testing function arguments

Jesus, trigger parsing code? In PL/pgSQL? o_O