Re: [JDBC] Re: JDBC2 Array Patch (New feature) - Mailing list pgsql-patches

From Rene Pijlman
Subject Re: [JDBC] Re: JDBC2 Array Patch (New feature)
Date
Msg-id fu83nto5a0u6gcc4jd0nkt0j3ih80n8iin@4ax.com
Whole thread Raw
In response to Re: JDBC2 Array Patch (New feature)  (Bruce Momjian <pgman@candle.pha.pa.us>)
List pgsql-patches
On Wed, 8 Aug 2001 11:19:32 -0400 (EDT), Greg Zoller wrote:
>> Specifically you can now use the ResultSet.getArray() method
>> to return an java.sql.Array object. Its a really convenient
>> way to load up a Java array with one columns' worth of results.

I don't think this patch is a correct implementation of the JDBC
spec. I've mailed this to Greg:

R>Isn't ResultSet.getArray() supposed to fetch *one* value
R>(the value of one column in one row) and interpret that as
R>a structured value of an SQL array field
R>(http://www.postgresql.org/users-lounge/docs/7.1/postgres/arrays.html)?
R>It looks like your code is looping through the ResultSet
R>and gathering data from multiple rows.

And received this reply:

G>You know what... I may have misinterpreted the JDBC spec!
G>My original understanding was that getArray was supposed to
G>iterate through a ResultSet and load up a Java array with
G>the contents of a given field (i.e. column).  But now as
G>I re-read the specification it does actually mean the
G>interpretation you gave (load a single field of type array
G>into a Java array object).
G>
G>So it seems that my implementation of Array is based on a
G>misunderstanding. :-(
G>
G>Massive bummer.  Please disregard my patch.

Greg, I do hope you continue your efforts! Not bad for a first
patch anyway, mine was only 2 lines. We still need Array support
badly... (hint, hint) :-)

Regards,
René Pijlman

pgsql-patches by date:

Previous
From: Dave Blasby
Date:
Subject: Re: [HACKERS] PostGIS spatial extensions
Next
From: Dave Blasby
Date:
Subject: Re: [HACKERS] PostGIS spatial extensions