Data comes from a monitor device, with test results of different elements. Those elements have 3 values, their ID, Their Value and any error messages.
The customer wants to see a flattened list of just Monitor Device and timestamp with a result of each element on seperate lines.
- Using the above we can achieve that, we import the data ignoring the header line. (hence labeling them with numbers)
- First we label the monitor “Parent” fields by using a field renamer
- We build a Empty map for us to correctly parse the records using expression evaluator
- we use a field mapping processor, to map those groups of 3 fields ( checking we only remap columns that are numeric)
- now we have groups, we split the groups into records using a field pivot processor
- This leaves a single group per record but as a group, Lets tidy that up with a field flattener so all the records are at the same level
- Finally we use a field renamer to label the 3 fields we have produced
- We ship that off to our secure storage facility…. or as this is a demo, into trash.
A Bit more detail on the field mapper
Before
After
