February 27, 2008
Submitted by Matt Castors, Chief of Data Integration, Pentaho
Mappings got a feature bump in version 3.0. So it's good to take a look at what has changed since the mappings in version 2.
Not much has changed here, except for these 2 things:
- It is now possible to have 0 or more than 1 "Mapping input" step in one and the same mapping.
- It is now possible to let unspecified fields flow through the "mapping input" step with a special option
The way we handle mappings has changed a bit since previous versions. It is no longer needed to specify anything in the "mapping output" step.
The information that is in there, possibly left over from previous versions is informational and no longer considered by the 3.x algorithm.
Options like "Include unspecified fields, ordered by name", also available in the "Select Values" step, are much more cleaner to deal with situations where you need to remove "temporary" fields from the output streams.
As with the "Mapping input" step, the "Mapping output" step has become a placeholder that now can also be used 0 or more times in the same mapping.
Consider the following simple mapping :
We can use this simple mapping in for example this transformation:
A new field X is calculated based on 2 input fields A and B. How do we map "A" to "leftValue" and "B" to "rightValue" on the input side and "res" to "X" on the output?
When we open the Mapping dialog, we notice the following tabs:
- One tab for "Parameters"
- One "Input" tab per "Mapping input" step in the mapping
- One "Output" tab per "Mapping output" step in the mapping