From: Tom Lane <tgl@sss.pgh.pa.us>
To: Andres Freund <andres@anarazel.de>
cc: pgsql-committers@lists.postgresql.org,
        Andrew Dunstan <andrew@dunslane.net>
Subject: Re: pgsql: ecpg: Output dir, source dir,
 stamp file argument for preproc/*.
In-reply-to: <20220718211408.o2uyu5kvud56ihlj@awork3.anarazel.de>
References: <E1oDWqn-000ADD-Oh@gemulon.postgresql.org>
 <20220718211408.o2uyu5kvud56ihlj@awork3.anarazel.de>
Comments: In-reply-to Andres Freund <andres@anarazel.de>
	message dated "Mon, 18 Jul 2022 14:14:08 -0700"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <822624.1658179641.1@sss.pgh.pa.us>
Content-Transfer-Encoding: quoted-printable
Date: Mon, 18 Jul 2022 17:27:21 -0400
Message-ID: <822625.1658179641@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/822625.1658179641%40sss.pgh.pa.us>
Precedence: bulk

Andres Freund <andres@anarazel.de> writes:
> Jul 18 16:48:25 ./src/interfaces/ecpg/preproc/parse.pl: Bareword file ha=
ndle opened at line 30, column 1.  See pages 202,204 of PBP.  ([InputOutpu=
t::ProhibitBarewordFileHandles] Severity: 5)
> Jul 18 16:48:25 ./src/interfaces/ecpg/preproc/parse.pl: Bareword file ha=
ndle opened at line 31, column 1.  See pages 202,204 of PBP.  ([InputOutpu=
t::ProhibitBarewordFileHandles] Severity: 5)

> afaict that's bogus. It's unnecessary that the code uses "our" instead o=
f
> "my", but there's no bareword there and replacing our with my fixes that
> complaint.

Ah, I think I've got it.  Per "man perlfunc":

           "our" has the same scoping rules as "my" or "state", meaning th=
at
           it is only valid within a lexical scope.  Unlike "my" and "stat=
e",
           which both declare new (lexical) variables, "our" only creates =
an
           alias to an existing variable: a package variable of the same n=
ame.

Since there's not actually any such package variable, the net effect is
that the first argument of open() is an undeclared name.  I can more or
less see why perl might treat that the same as a bareword, though I
definitely agree that this error message is more confusing than helpful.

			regards, tom lane