Detailed Behavior

Each instance of rate must specify the "co-ordinates" of required cell in from a rate table. The co-ordinates which uniquely identify a cell are:

• the code for the name of the rate table (i.e. the code for the entry from the RateTableType code table);
• the code for the name of the row in the rate table (i.e. the code for the entry from the RateRowType code table); and
• the code for the name of the column in the rate table (i.e. the code for the entry from the RateColumnType code table).

Recall that each rate table can have multiple "version" with values effective from a specified date; whereas the rule object created by the Rate Rule Object Propagator has a timeline of numbers representing a varying rate value. Thus the value returned by the rate is the value of the rate cell as it varies over different versions of its rate table.

When an instance of rate is evaluated, a rate rule object matching the required rate cells co-ordinates is sought. Ordinarily the required rate rule object will be found, and its varying rate cell value returned as a timeline of numbers.

However, if the required rate rule object has not been found (e.g. if rate propagation has not been run, for example in a unit test that creates a rate but does not publish changes), then the rate expression will create an internal rule object to hold the varying value from the rate table (subject to the same constraints that apply to the Rate Rule Object Converter - i.e. cells in sub-rows and sub-columns are not supported). If such internal rule object are created, then if a rate table change is subsequently published, then the rate expression will safely recalculate to pick up the external rule object created by the propagator. This feature means that it is possible to write unit tests that interact with rates without having to worry about rate propagation.