Horton

Re: Non-text mode for pg_dumpall

Noah Misch <noah@leadboat.com>

View thread
From: Noah Misch <noah@leadboat.com>
To: Mahendra Singh Thalor <mahi6run@gmail.com>
Cc: Andrew Dunstan <andrew@dunslane.net>, Álvaro Herrera <alvherre@alvh.no-ip.org>, jian he <jian.universality@gmail.com>, Srinath Reddy <srinath2133@gmail.com>, pgsql-hackers@lists.postgresql.org
Date: 2025-07-24T17:50:36Z
Lists: pgsql-hackers

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

  1. Add non-text output formats to pg_dumpall

  2. Improve pg_dump/pg_dumpall help synopses and terminology

  3. Non text modes for pg_dumpall, correspondingly change pg_restore

  4. Doc: manually break lines in wide UUID examples.

On Thu, Jul 17, 2025 at 03:46:41PM +0530, Mahendra Singh Thalor wrote:
> On Wed, 16 Jul 2025 at 05:50, Noah Misch <noah@leadboat.com> wrote:
> > On Thu, Jul 10, 2025 at 12:21:03AM +0530, Mahendra Singh Thalor wrote:
> > > On Wed, 9 Jul 2025 at 02:58, Noah Misch <noah@leadboat.com> wrote:
> > > > On Fri, Apr 04, 2025 at 04:11:05PM -0400, Andrew Dunstan wrote:
> > > > > +/*
> > > > > + * get_dbnames_list_to_restore
> > > > > + *
> > > > > + * This will mark for skipping any entries from dbname_oid_list that pattern match an
> > > > > + * entry in the db_exclude_patterns list.
> > > > > + *
> > > > > + * Returns the number of database to be restored.
> > > > > + *
> > > > > + */
> > > > > +static int
> > > > > +get_dbnames_list_to_restore(PGconn *conn,
> > > > > +                                                     SimpleOidStringList *dbname_oid_list,
> > > > > +                                                     SimpleStringList db_exclude_patterns)
> > > > > +{
> > > > > +     int                     count_db = 0;
> > > > > +     PQExpBuffer query;
> > > > > +     PGresult   *res;
> > > > > +
> > > > > +     query = createPQExpBuffer();
> > > > > +
> > > > > +     if (!conn)
> > > > > +             pg_log_info("considering PATTERN as NAME for --exclude-database option as no db connection while doing pg_restore.");
> > > >
> > > > When do we not have a connection here?  We'd need to document this behavior
> > > > variation if it stays, but I'd prefer if we can just rely on having a
> > > > connection.
> > >
> > > Yes, we can document this behavior.
> >
> > My review asked a question there.  I don't see an answer to that question.
> > Would you answer that question?
> 
> Example: if there is no active database, even postgres/template1, then
> we will consider PATTEREN as NAME. This is the rare case.
> In attached patch, I added one doc line also for this case.

If I change s/pg_log_info/pg_fatal/, check-world still passes.  So no test is
reaching the !conn case.  If one wanted to write a test that reaches the !conn
test, how would they do that?