[unixODBC-support] Implementing an ODBC *server*?

Peter Valdemar Mørch swp5jhu02 at sneakemail.com
Mon Oct 17 14:13:10 BST 2005

Hi there,

We have an application running on Linux where we would like to export 
data via ODBC - primarily so that MS Office users can create "external 
data" links to our application and have the refresh automatically.

The easy way is to just e.g. put all the data in a MySQL database and 
export it via ODBC, but that is not practical given the large amounts of 
data involved and the fact that they are stored elsewhere in other 
formats already.

How does one implement an ODBC server? Can/will unixODBC help with this?

Ideally what we're hoping for is that a library will do all the 
encoding/decoding, SQL parsing etc. and allow us to register callback 
routines that get called when someone issues an SQL query against "us".

Do we have other options? I guess one could start reading the ODBC 
specs, and start implementing ODBC from the ground up, but that tends to 
be the hard way! :-D

Our application is written in perl but we're not frightened if we have 
to write some C, C++, java, python or whatever is required. We'd prefer 
to stay in perl, however.

Any good ideas?

Thank you for reading this far.

Kind regards,


Peter Valdemar Mørch

More information about the unixODBC-support mailing list