On Mon, 20 Feb 2012 15:55:54 -0500, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Guillaume Lelarge <guillaume@lelarge.info> writes:
>> On Mon, 20 Feb 2012 14:15:34 -0300, Alvaro Herrera
>> <alvherre@commandprompt.com> wrote:
>>> Excerpts from Susanne Ebrecht's message of s=C3=83=C2=A1b feb 18 18:11:=
54 -0300
>>>> I think that the constraint keyword "DEFAULT" should not be
translated
>>>> here.
>=20
>> I don't see why the modifiers shouldn't be translatable. It's not as if
>> they were intended to be copied and pasted in a query. They are here to
>> help the user understand the different modifiers in each column.
>=20
>> -1 to get rid of the translations.
>=20
> It seems to boil down to whether you think the "Modifiers" column
> contains SQL, or a textual description of the column's properties.
> I could go either way on that, but I notice that the existing code
> is in the habit of truncating the default expression at 128 characters.
> If we want to decide that we're printing real, copy-and-pastable SQL,
> that seems like a bad idea. On the other hand, if we're printing
> informational text, it's not an unreasonable thing.
>=20
My first move was that truncating to 128 characters was bad idea. But in a
terminal, you have a limited number of columns, so it would make sense.
And, to me, it's a textual description. Not SQL. BTW, if it was SQL, when
I do \d tablename, I should get the complete SQL query to create the table,
not a description of the table definition. So, yeah, sure, the "Modifiers"
column doesn't contain SQL, it's a description.
--=20
Guillaume
http://www.postgresql.fr
http://dalibo.com