Re: Patch for pg_dump: Multiple -t options and new -T option
| От | Tom Lane |
|---|---|
| Тема | Re: Patch for pg_dump: Multiple -t options and new -T option |
| Дата | |
| Msg-id | 7462.1090325928@sss.pgh.pa.us обсуждение исходный текст |
| Ответ на | Re: Patch for pg_dump: Multiple -t options and new -T option (Christopher Kings-Lynne <chriskl@familyhealth.com.au>) |
| Ответы |
Re: Patch for pg_dump: Multiple -t options and new -T
|
| Список | pgsql-patches |
Christopher Kings-Lynne <chriskl@familyhealth.com.au> writes:
> One problem with this patch is that there's no way to dump multiple
> tables in different schemas. Does this matter? It's a bit
> non-orthogonal...
Yeah. With the combination of -n and -t you can pull a specific table,
but as soon as you allow either switch to be multiple you've got an
inexact tool.
I had thought of allowing -t to be schema.table but I'm worried about
backwards-compatibility issues. In particular, since we don't support
SQL-style quoting in -t arguments, how could one then select a table
name that actually contains a dot? Or should we just write off that
case as "stupidity is its own reward"? It would also be good to not
foreclose the possibility of wild-card matching patterns in these
switches in future.
(BTW, does the patch handle multiple -n switches?)
regards, tom lane
В списке pgsql-patches по дате отправления: