Recently, a number of my projects have required that Scribe Insight connects to a home-grown web service in order to write or retrieve data. While there is some information out there on how to use Scribe’s web service adapter, there were a number of surprises along the way that I’d like to share.
Scribe’s web service adapter only supports SOAP-based web services with a valid WSDL (does not work with RESTful web services). That having been said, the web services does not work with all SOAP-based web services. Be sure to build extra time into your project to complete a thorough proof of concept, including a test of all relevant web methods both in the Workbench and through automated runs in the Console.
When Scribe initializes a DTS and authenticates with a source connection, it actually runs the source query to validate the schema. This is also true when using a web service method as your source query. Make sure the web method is written in such a way that is can be called with the same input more than once and return the same data set. For example, the web method should not mark source data as “processed” after execution, otherwise you will lose that data upon DTS initialization.
When selecting a method to use as your source you’ll notice there is no room to enter parameters. Simply navigate to the “Filters” tab and list your input parameters as a series of filter clauses. The only comparison operator available is “=” and the only condition operator available is “and”.
![Scribe_Filter]()
You can only use each method once in the Step Configuration per DTS. Either wait until the end of the DTS (if possible) to make your call to the web method or be sure to capture all necessary return values in Target Variables for use later.
