Error detection
During map execution, the input data is compared to the data definition in the type tree. If the data does not match the definition, it is invalid or "in error".
To validate a data object as belonging to a certain type, the data must be matched to its type definition. For data to be valid, the following must be true:
- The data must have the properties that were defined for the type.
- If the type is an item that has restrictions, the data object must match one of the restrictions.
- If the type is a group, the components of the data object must match those defined in the group view, and each component rule must evaluate to "true" at map execution time.
Ultimately, all data objects, no matter how complex, consist of items because items represent the smallest unit of data. When all of the items that collectively comprise a component are found, the component has been "found."
When executing a map, invalid data is recorded in the trace file. You can decide what you want the system to do when errors are encountered: you can map the errors to an output and also have the system ignore invalid data. For methods on ignoring invalid data, see "Restart attribute" .