Re: JDBC String to Bool spec

Seems to me we can introduce a breaking change in version 42. If this
becomes an issue then we can put it back behind an option as Kevin
suggested.

Dave Cramer

davec(at)postgresintl(dot)com
www.postgresintl.com

On 16 January 2017 at 16:20, Kevin Wooten <kdubb(at)me(dot)com> wrote:

> As the OP I’d like to make a clarification….
>
> I didn’t want to use a string of “1.0” and convert it to a Boolean. The
> driver, in its test suite, tests for this specific case (and a few others).
> These seemed really out of the norm since the database doesn’t support
> this. Thus my investigation began.
>
> On Jan 16, 2017, at 2:16 PM, rob stone <floriparob(at)gmail(dot)com> wrote:
>
> Hello,
>
>
> Since we now know the JDBC spec doesn’t support this (thanks to all
> of Mark Rotteveel’s sleuthing) & we know Postgres doesn’t support
> this. It seems the “feature" should be removed or placed behind an
> option. Postgres results in an error here and considering this is
> a driver tied to that database I look to that to clarify ambiguity.
> These examples should all result in an error.
>
>
>
>
> Yes, I agree, a "feature" that allows bad code should be considered
> a "bug".
>
>
>
> The OP in the original post used a string of "1.0" and wanted to
> convert it into a boolean. If you use setBoolean in a prepared
> statement it expects the argument to be a "numeric" type, so it is not
> surprising that the string was converted to a double. You'd expect
> something along the lines of ps.setBoolean(5,"xyz") to throw an error.
>
> I don't know how many string values are to be considered "true" in the
> application. Maybe use some regular expressions to suss out "true"
> values?
>
> My two cents.
>
> Rob
>
>
> --
> Sent via pgsql-jdbc mailing list (pgsql-jdbc(at)postgresql(dot)org)
> To make changes to your subscription:
> http://www.postgresql.org/mailpref/pgsql-jdbc
>
>
>