Re: [pg_dump] 'create index' statement is failing due to search_path is empty

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: tushar <tushar(dot)ahuja(at)enterprisedb(dot)com>
Cc: PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: [pg_dump] 'create index' statement is failing due to search_path is empty
Date: 2020-04-28 14:31:03
Message-ID: 28215.1588084263@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

tushar <tushar(dot)ahuja(at)enterprisedb(dot)com> writes:
> While testing something else ,i found 1 scenario  where pg_dump  is failing

> CREATE FUNCTION do_analyze() RETURNS VOID VOLATILE LANGUAGE SQL
>         AS 'ANALYZE pg_am';
> CREATE FUNCTION wrap_do_analyze(c INT) RETURNS INT IMMUTABLE LANGUAGE SQL
>         AS 'SELECT $1 FROM do_analyze()';
> CREATE INDEX ON vaccluster(wrap_do_analyze(i));
> INSERT INTO vaccluster VALUES (1), (2);

You failed to schema-qualify the function reference. That's not
a pg_dump bug.

While we're on the subject: this is an intentionally unsafe index.
The system doesn't try very hard to prevent you from lying about the
volatility status of a function ... but when, not if, it breaks
we're not going to regard the consequences as a Postgres bug.
Basically, there isn't anything about this example that I'm not
going to disclaim as "that's not supported".

regards, tom lane

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Stephen Frost 2020-04-28 14:31:54 Re: More efficient RI checks - take 2
Previous Message Jesse Zhang 2020-04-28 14:15:49 Re: Fix compilation failure against LLVM 11