We have a partner sending us csv data to a web service of ours that gets sent to a messaging system that is SQL Server based.
I am writing data to disk and then importing to another database by using this file as a flat file source in an SSIS Data Flow task.
Although this works I'd like to avoid using the file system and treat the database field as a flat file source.
Is this possible?
I'm very new to SSIS packages and searching for this scenario is very difficult because of the number of hits just relating to importing a csv file.
Best Answer
My first inclination would be to use a script component (e.g. C#) as source in a Data Flow Task. The script would just read the column contents, parse the individual records and fields, and send records down the pipeline. The individual fields (perhaps strongly-typed) can be defined in the output record definition of the designer.
https://msdn.microsoft.com/en-us/library/ms136060.aspx contains an example of parsing a CSV file and processing a SQL query. You can combine the two examples for your needs. Below is an example of this. I don't have time to test right now but it should get you started.