I wrote:
> Rather what we've got is that constraints are uniquely named among
> those associated with a table, or with a domain. So the correct
> unique key for a table constraint is table schema + table name +
> constraint name, whereas for a domain constraint it's domain schema +
> domain name + constraint name. The current patch provides sufficient
> information to uniquely identify a table constraint, but not so much
> domain constraints. Should we fix that? I think it'd be legitimate
> to re-use SCHEMA_NAME for domain schema, but we'd need a new nonstandard
> field DOMAIN_NAME (or maybe better DATATYPE_NAME) if we want to fix it.
Here's an updated patch (code only, sans documentation) that fixes that
and adds some other refactoring that I thought made for improvements.
I think this is ready to commit except for the documentation.
I eventually concluded that the two ALTER DOMAIN call sites in
typecmds.c should report the table/column name (with errtablecol),
even though in principle the auxiliary info ought to be about the
domain. The reason is that the name of the domain is nearly useless
here --- you probably know it already, if you're issuing an ALTER for
it. In fact, we don't even bother to provide the name of the domain
at all in either human-readable error message. On the other hand,
the location of the offending value could be useful info. So I think
usefulness should trump principle there.
regards, tom lane