Hi Vladimir,
Thank you for the timely response.
I am new to this community. Can you provide specifics as to how I would go about bringing this to the attention of the “postgresql developers” – that is, I’m not sure if you’re directing me to send this same message out to the pqsql-hackers mailing list, or to submit a bug report via the form located at www.postgresql.org/account/submitbug/, or maybe to do something else, altogether.
And yes, in the mean time we can work with the lowercase style just fine (the hard part was figuring out the cause of the bad behavior).
--Al
From: Vladimir Sitnikov [mailto:sitnikov.vladimir@gmail.com]
Sent: Tuesday, January 24, 2017 5:34 PM
To: Al Presseller <apresseller@commoninf.com>
Cc: pgsql-jdbc@postgresql.org
Subject: Re: [JDBC] "Cached plan must not change result type" error when switching search_path via explicit execution of SET SEARCH_PATH = ...
That's unfortunate.
Would you please report that ustream? (I mean to the postgresql developers)
The thing is it would be so much easier if pgjdbc would not have to parse SQL in order to identify if the statement was "set search_path".
In the mean time: is lowercase style fine for you?
Vladimir
The fix for issue #496 ("Cached plan must not change result type" when using PreparedStatement cache across schemas) doesn’t work when the default schema is being assigned via explicit execution of a SET statement that renders the search_path keyword in UPPER CASE.