nk4um

Many thanks cj.

Bit clunky, but i guess jtds do it for optimisation reasons.

I just changed it to use the MS driver and that seems ok (for now)

This is happening because the NetKernel RDBMS module generates an HDS representation made up of the POJOs that are returned by the JDBC API. The default serialization of HDS, I believe, outputs the toString of the representation class (which in this case is a net.sourceforge.jtds.jdbc.ClobImpl). It looks like ClobImpl doesn't have a toString implementation, so uses the default Java one (which returns the Object's class and hash). If traverse the HDS in Java, you will get access to the actual Clob Object.

I can see two possible/sane solutions:

• Processing the HDS after the query is performed, converting the Clob into a String.class.
• Creating a Clob to String transreptor.

-cc

Im doing a db request using freemarker and jdts driver.

<#if component = "article-full">
select top 1 headline, articleBody
from PCA_Features f
</#if>

For some reason the articleBody which is nvarchar(max) comes back as a reference and not the field value.

[

    {
        "headline": "Boeing Designs In-Flight Surfing",
        "articleBody": "net.sourceforge.jtds.jdbc.ClobImpl@42ecf8c4"
    }

]

any ideas?

My config file for the RDBMS looks like