Conditional processing means you apply attribute values to certain elements of your text. These values are then entered into a separate DITAVAL (DITA values) file in which you write rules that specify whether elements with those values need to be included, excluded (filtering) or flagged (highlighting). While generating output, the rules in this DITAVAL file are applied to the end-result. Below, the three main steps in conditional processing are described.

Before worrying about applying conditions, you should first establish a scheme and define exactly what you need to create conditions for. This way you can plan out which attributes you need and which values to assign to each of them. To do this, define your:

1. Audience: who will use your product and what specific information do they need?
2. Platform: which operating systems, platforms and other environments are used with your product?
3. Product: which models/versions are you planning and how do they differ?

Once you have specified which values you need and in which circumstances you will need them, you can start applying them as you write. Applying them while writing is far easier than having to hunt through your documentation afterwards to find the elements that are product/audience specific.

You can apply conditional attribute values in:

Note: It is recommended to apply attributes to the topic references in the DITA map, and not to the <topic> element within the topic in question.