Re: Adding support for batches that return generated keys - Mailing list pgsql-jdbc

From Giuseppe Sacco
Subject Re: Adding support for batches that return generated keys
Date
Msg-id 543D21A2.1050204@eppesuigoccas.homedns.org
Whole thread Raw
In response to Adding support for batches that return generated keys  (Craig Ringer <craig@2ndquadrant.com>)
Responses Re: Adding support for batches that return generated keys  (Dave Cramer <pg@fastcrypt.com>)
List pgsql-jdbc
Hello Craig,

On 10/10/2014 11:02, Craig Ringer wrote:
> Hi all
>
> I've implemented support for batches that return generated keys.
>
> https://github.com/pgjdbc/pgjdbc/pull/204
>
> and I'm looking for feedback before I merge this into master.

I spent some time in looking your patch. Since I am just beginning
studying how pgjdbc works, I cannot speak about correctness of the code,
but I still have two questions:

First one is about the java version you are using. It seems to me that
the driver should be compileable in java 1.4, i.e., on a java version
that doesn't  use int and Integer in a mixed way. At the end of
getMaxResultRowSize() in org/postgresql/core/v3/SimpleQuery.java you
assign an int value to an Integer variable. Is this correct?
On the very same method, on the first lines, you extract the intValue()
from the Integer variable.

Second, in method flushIfDeadlockRisk(...) in file
org/postgresql/core/v3/QueryExecutorImpl.java, about line 1169, you use
the word "uresults". Is this a typo or a data structure used by pgjdbc?

Bye,
Giuseppe



pgsql-jdbc by date:

Previous
From: Enrico Olivelli - Diennea
Date:
Subject: R: Disk buffering of resultsets
Next
From: "Lussier, Denis"
Date:
Subject: Re: Disk buffering of resultsets