2009/11/18 Peter Eisentraut <peter_e@gmx.net>:
> On m=C3=A5n, 2009-11-16 at 12:28 +0100, Pavel Stehule wrote:
>> 2009/11/16 Peter Eisentraut <peter_e@gmx.net>:
>> > On m=C3=A5n, 2009-11-16 at 10:19 +0100, Pavel Stehule wrote:
>> >> wrong:
>> >>
>> >> pavel@nemesis ~]$ psql postgres -v x=3D10 -c "select :x"
>> >> ERROR: =C2=A0syntax error at or near ":"
>> >> LINE 1: select :x
>> >> =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0^
>> >
>> > This is documented in the psql man page.
>>
>> I don't see it, Peter?
>
> =C2=A0 =C2=A0 =C2=A0 -c command
>
> =C2=A0 =C2=A0 =C2=A0 --command command
> =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0Specifies that psql is to=
execute one command string,
> command, and then exit. This is useful in shell scripts.
>
> =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0command =C2=A0must =C2=A0=
be =C2=A0either a command string that is
> completely parsable by the server (i.e., it contains no psql specific
> features), or a single backslash command. Thus you cannot
> =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0mix SQL and psql meta-com=
mands with this option.
>
>> Is it some reason for it? I don't understand, why this order is correct:
>>
>> execute statement
>> process external variables
>> finish
>
> Well, -c works a bit different so that it is possible at all to send a
> command to the server without any psql processing in the way. =C2=A0It's a
> poor excuse, from a user's point of view, but that's historically why
> it's been kept that way.
ok - thank you. I could to live with it.
Regards
Pavel Stehule
>
>