Presuming that by datasources you mean an external data read by a custom control or module, the easiest way to index such content is using the built-in PageIndex.
The website crawler will scan through all the pages included and index the data visualized at client pages as HTML.
A disadvantage of such approach is the dynamically loaded paged data using server controls like GridView or custom paged result set because the crawler will index only the first page.
Another much more complex approach is building Custom Index Provider that implements methods and logic for reading your data directly from the external datasource when indexed.
More info on your scenario would help us to point you to the right solution.
Let us know if there is anything else with could assist you with.
All the best,
the Telerik team