Capturing Field Name Metadata from a CSV File in Altova MapForce

Question

I've been asked to prototype a replacement "file transformation process" (that currently is a mess of SQL) using Altova's MapForce. My input is a CSV file with headers. My problem is that I need to capture both the data AND the column name to use in downstream processing.

I need to have MapForce feed a C# method (imported as that takes two parameters: fieldName and value. I can access the value trivially, but after hours pouring over the manual (1000 pages!) I haven't found any examples of how to access the field name as an output.

The reason each output needs the field name and the value has to do with how all our mappings/transformations are currently managed - on a database. The .NET code jumps in at this point and does any necessary database lookups.

For example, if I had the following file:

"Symbol", "Account", "Price", ...
"FOO", "10101", "1.23", ...
"BAR", "10201, "13.56", ...

And a static method string TransformField( string fieldName, string value ),

I'd like to map the CSV file's Symbol data output to the method's value parameter and the Field Name "Symbol" to the method's fieldName parameter.

Some limitations:

1. I need to keep the "wiring" visible in the MapForce GUI. I'll have non-programmers maintaining the mappings in the future. So doing all this in code is not an option.
2. MapForce is the tool of choice by the company. Part of the reason our original process is such a mess is because the original programmer rolled his own mapping/transformation tool (out of TSQL no less - ouch).
3. We can treat all inputs/outputs to the method call as strings. Conversions will happen later.
4. I would like to avoid using scalar literals as inputs. I already have the column names from the file - I do not want to re-type each one and feed it to my method.

I'm not sure how many users out there have experience with this tool, but after 3 days of tinkering with it, I see much potential. If only I can get past this current sticking point, I think the company will have a solid alternative to their current mess.

Thanks for any/all suggestions.

Answer 1

I solved my issue and, for future reference, want to post a solution. I handled my problem by using MapForce's FlexText. This allowed me to extract the header from the CSV file and "invert" the column names as data inputs to the transformation process. Once I knew the approach to take, I was able to find more information directly from Altova.

I found a couple helpful tutorials while digging through their website:

• Altova Online Videos
• Web Tutorial

Hope this can help someone else in the future!