x

When you use includes, there is a list of included Plans below the work area. The order of includes in the list is based on the order in which the includes were inserted.

Important: You cannot change the order of includes at a later time. The only way to change the order is to remove the includes and add them again. In other words, you must repeat the process of adding the includes to change their order in the list.

The name of the include symbol, visible in a Plan, is derived from its physical file name.

The following image shows the name of an include symbol that is visible in a Plan.

x

You can structure a Plan in one of two ways:

• By include. Use an include for the purpose of structure. The Plan is easier to read. You create an include by selecting a set of steps in a Plan (typically, according to the attribute sets). The steps are structured hierarchically for better orientation. The data flow is the same before the include and inside the include. An include is not reusable.

  One include can be part of another include, allowing for a hierarchy.

• By component. A component is a user-defined, reusable package of a Plan, which provides a single specific function. Available inside a component are only the columns given by the defined interface. Except within the interface, data flow is bypassed. Data flows "around" the component.

  One component can be part of another component, allowing for a hierarchy. A component can be part of an include, but an include cannot be part of a component.

  A component can use parameters. Inside a component, you can map each individual property to a parameter, whose value can be set from the component step.

  If, inside a component, you can use only columns that are defined in input steps, what happens to columns in the data flow before the component step, that is, what happens to columns that are not mapped? It depends on the internal structure of the component. If the component does not discard the data flow (for example, in step Join, Union, and Unification) and creates a new one, the other columns bypass the component and are available in the data flow outside the component. In the latter case, the data flow outside the component is defined by the component.

  When using a component, look at the data flow going to and coming from the component step. You can do that by selecting the connection and clicking Connection in the Properties window.

x

Keep the following tips in mind:

• The caret/tilde (^~) pair of characters is preferred as a field separator. You are unlikely to find this pair of characters in your actual data. However, some systems cannot properly process a double-character field separator like ^~. Therefore, use only one of the characters as a field separator.
• Do not use a semicolon (;) or a pipe (|) as a field separator unless you use string qualifiers. In particular, the semicolon is frequently used in text and may cause a parsing error.

x

To score a change in a Column Assigner (CA) step, be aware that each change between input and output values will be scored. For example, when you initially store a value in an empty field, the score and explanation will be written.

To score only modified values, use the following assignments:

• src_name --> pur_name (first assignment without scoring)
• pur_name --> capitalize(pur_name) (second assignment using score NAME_CAPITALIZED)

The scorer will be applied only if the value was modified. When pur_name is equal to capitalize(pur_name), no scoring is done.

For more complex assignments that occur in CA according to a condition or rule, use the Scoring Simple step or the Scoring step instead of the internal CA scorer. Inside of these steps, you can define conditions that specify when and how to score.