fix: paramaters of type character and bit not ignoring length

[laurenceisla]

Closes #1586.

[laurenceisla]

According to PostgreSQL docs on character types:

If specified, the length must be greater than zero and cannot exceed 10485760. character without length specifier is equivalent to character(1)

But, the docs for CREATE FUNCTION say this:

The full SQL type syntax is allowed for declaring a function's arguments and return value. However, parenthesized type modifiers (e.g., the precision field for type numeric) are discarded by CREATE FUNCTION. Thus for example CREATE FUNCTION foo (varchar(10)) ... is exactly the same as CREATE FUNCTION foo (varchar) ...

It made me think that CREATE FUNCTION foo(character(10)) is the same as CREATE FUNCTION foo(character), where character is of length 1 by default, as the docs on chars say. But that's not the case, because it allows to use parameters with a maximum length specified for a character which is the same of a varchar and text (10485760).

I couldn't find in the source code how they handle this change, so I casted both character and bit to their character varying and bit varying respectively. I'm not sure if it's a bit opinionated on my part.

[steve-chavez]

so I casted both character and bit to their character varying and bit varying respectively

If they fail when exceeding the length I guess it's fine. (maybe the error message changes but it wouldnt be a dealbreaker)

[steve-chavez]

Is ppType now unused? Would it make sense to have a single ppType with the added length for these cases?

[laurenceisla]

No, ppType is still used on every other place (e.g. errors, OpenAPI, etc.) Only when building the Query it transforms e.g. character -> character varying.

The thing is, the length is ignored by PostgreSQL, it's not specified in the function definition, so no length can be added, it's the character -> character varying change that's done here.

To clarify a little more, this change could also be done using only ppType. It wouldn't be at Schema Cache level, but per request instead, e.g.

case ppType of
  "character" -> "character varying"
  "bit" -> "bit varying"
  ...

[laurenceisla]

Interestingly, only the PG14 io-test keeps failing. Could not find a reason why for now. Maybe a change needs to be done in the schema query.

[steve-chavez]

@laurenceisla I don't see the pg 14 failure from CI. Can you link to the error?

[laurenceisla]

@steve-chavez Ah, true. Strangely it failed twice in a row (1st and 2nd attempt) https://github.com/PostgREST/postgrest/actions/runs/5661060491/attempts/2?pr=2867, then the last ones run without problems.