1.3.4.10.4 Output Selector: First Non-empty Value

The First Non-Empty Value output selector searches through the records being merged and selects the first non-empty value it finds for the output attribute. Where input records are being merged from multiple data sources, you can specify the order in which the records are inspected, in order to specify a preferred source for the merged value.

Use the First Non-Empty Value output selector when you want to prefer the values from one data source over another. For example, when enhancing records, you might prefer to use the value from the reference data table, provided it is not empty. The value from the working data would only be chosen if it was input to the output selector, and if the reference data table has an empty value for the attribute.

You might also use the First Non-Empty Value selector if you want to ensure you have some kind of default output value for each output attribute even if there is no way of selecting data using a specific business rule, and you need to review and check the output manually to ensure the values are correct. This might decrease the time it takes to resolve your required output.

Note that where a single data source is used but there are multiple records in each group, the First Non-Empty Value selector will select the lowest value for Number attributes, the earliest date for Date attributes, and the first alphabetically sorted value for String attributes. This is purely so that output selection is deterministic if the same match process runs several times.

The following table describes the configuration options:

Configuration	Description
Inputs	Any attributes of any type from any input data sets.
Options	None. The order of the input attributes is important to output selection. Input the attributes in order so that the first attribute is preferred in selection, and the second attribute only used if the first has no value.

Example

In this example, the First Non-Empty Value output selector is used to select a value for the PostCode attribute, when enhancing customer address data from a trusted reference data stream. You want to choose the post code from the reference data stream if it is available. If it is not available, you would like the original post code value in your working data to be preserved. You therefore input both PostCode attributes, but ensure that the output selector will use the reference data stream first.

Example output

The following table shows example output using the First Non-Empty Value selector, configured as above:

Table 1-103 Example Output Using First Non-Empty Value Selector

Input value from reference data stream	Input value from working data stream	Output value
CB4 1UW	CB4 1YW	CB4 1UW
CB4 3DD	CB4 3DD	CB4 3DD
CB4 0WS	null	CB4 0WS
null	SW11 5QB	SW11 5QB
null	null	null (see Note)

Note:

When a null value is selected in output, this will only be considered as a selection error if the Allow NULLs setting has been de-selected for the output attribute. Otherwise, the selection of the null value will be considered as correct.