FreeTDS Development Group

[Bob Kline <bkline AT rksystems.com>]

Craig:

Well, I've accomplished my own goal, which was to hack the freetds JDBC
driver until I could talk to log onto a SQL Server database using
version 7.0 of TDS and talk to NTEXT columns.

What I have is very limited.  In particular:
 1. I haven't added the NCHAR or NVARCHAR types yet.
 2. The method for specifying 7.0 for TDS is somewhat crude.
 3. Lots of the logic has been taken directly from the freetds C code,
    and I don't have any better idea of what all the magic values mean
    than the original authors; and most important:
 4. I haven't done anything approaching a code review of the entire body
    of source code to find out where everything is and how it works.

On point #1, I did modify Tds.getCharValue() to handle wide characters,
so once the token values are plugged into all the right places for
SYBNCHAR and SYBNVARCHAR it has a good chance of working properly for
retrieval of those types.

On point #2, I couldn't find any more elegant and unobtrusive way of
switching on the 7.0 processing flavor than System.getProperty().  If
the "TDSVER" property is set to "70" then the Tds and TdsComm classes
know to use that version of TDS.

On point #4, I'm fairly sure that 4.2 code that was working properly
before will continue to do so, and it looks as if the new functionality
I've added is working with my limited testing, but I've haven't done
anything to verify either of these assumptions thoroughly.

If, with all of these caveats, you still want the modifications, I'll
post the patches here or mail them to you directly (as you prefer).
Caveat receptor! :->}

Brian:

I noticed as I was shamelessly copying your work that the C code to
extract column information for TDS 7.0 is missing the logic to pull out
the precision and scale bytes for DECIMAL and NUMERIC columns.

Bob