From: | "Johann 'Myrkraverk' Oskarsson" <johann(at)2ndquadrant(dot)com> |
---|---|
To: | PostgreSQL JDBC <pgsql-jdbc(at)postgresql(dot)org> |
Subject: | Re: Implementing setBinaryStream(int, InputStream, long) |
Date: | 2012-07-10 09:09:20 |
Message-ID: | byody.x5byody.ehok.ugn3.gnus@asuka.myrkraverk.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-jdbc |
Dave Cramer <pg(at)fastcrypt(dot)com> writes:
> Also it would be nice if a test case came with this patch.
I've written a testsuite for this patch. It tests the three cases of
length < 0, length > Integer.MAX_VALUE and that it actually works.
The testcase that tests if the feature is working properly uses the
prepared statement SELECT ?::TEXT and compares the returned hex value.
This is not a good idea because it depends on the bytea -> text
conversion which has changed in recent Postgres versions, right?
What is a good way to isolate the test code from version specific
behaviour? In this case, whether bytea to text is _escape_ or _hex_.
Now that we're using git, is there any point in CVS identification
strings in new files? Also, I only state Copyright 2012 for new files,
and not 20xx-2012, right?
When tests are catching SQL exceptions and comparing the error code, do
we want to hard code them, or refer to PSQLState constants? Personally,
I do not believe in using the same constants in test and production code.
--
Johann Oskarsson http://www.2ndquadrant.com/ |[]
PostgreSQL Development, 24x7 Support, Training and Services --+--
|
Blog: http://my.opera.com/myrkraverk/blog/
From | Date | Subject | |
---|---|---|---|
Next Message | Johann 'Myrkraverk' Oskarsson | 2012-07-10 09:36:55 | SQL Error Code Anomaly |
Previous Message | Craig Ringer | 2012-07-05 23:29:29 | JDBC 4.2 spec: expert group proposed refcursor support |