On 02/01/2010 08:06 AM, Tom Lane wrote:
> Heikki Linnakangas <heikki(dot)linnakangas(at)enterprisedb(dot)com> writes:
>> Fujii Masao wrote:
>>> In HEAD, psql using conninfo fails in connecting to the server as follows.
>>>
>>> $ bin/psql "host=localhost"
>>> psql: FATAL:  database "host=localhost" does not exist
>>>
>>> This is because the recently-introduced PQconnectStartParams()
>>> doesn't handle correctly the dbname parameter containing '='.
> 
>> Hmm, I don't think that was ever really supposed to work, it was
>> accidental that it did.
> 
> No, it was intentional.

Here's a patch.

If "=" is found in the dbname psql argument, the argument is assumed to
be a conninfo string. In that case, append application_name to the
conninfo and use PQsetdbLogin() as before. Otherwise use the new
PQconnectdbParams().

Also only uses static assignments for array constructors.

Objections?

Thanks,

Joe

In response to

• Re: BUG #5304: psql using conninfo fails in connecting to the server at 2010-02-01 16:06:11 from Tom Lane

Responses

• Re: BUG #5304: psql using conninfo fails in connecting to the server at 2010-02-03 01:46:14 from Fujii Masao
• Re: BUG #5304: psql using conninfo fails in connecting to the server at 2010-02-03 01:59:20 from Tom Lane

pgsql-hackers by date

Next:	From: Tom Lane	Date: 2010-02-03 01:14:17
	Subject: pgsql: Assorted cleanups in preparation for using a map file to support
Previous:	From: Jeff Davis	Date: 2010-02-03 01:05:08
	Subject: Re: Listen / Notify - what to do when the queue is full

pgsql-bugs by date

Next:	From: Fujii Masao	Date: 2010-02-03 01:46:14
	Subject: Re: BUG #5304: psql using conninfo fails in connecting to the server
Previous:	From: Tom Lane	Date: 2010-02-02 22:18:06
	Subject: Re: whole-row functional index?