The index can be local or remote. Each string section within the index can be any length of Unicode text including punctuation, spaces, symbols, letters, etc.

Character Overlap
For example, a node in the index may simply store the single character ‘N’ or it may store the entire string of ‘Northern Inc.’. It is at this point that the kannuu index analyzes character overlap between items in the list and designs the hierarchy tree to take advantage of this overlap. As a result, the index can provide faster lookup of phrases that share common character sequences such as ‘No’, ‘North’, and ‘Northern’. The source strings may also combine multiple identifiers allowing for simultaneous selection based on multiple levels. For example, a single lookup session can provide selection of both a music track and artist.

Prioritization
At the time of indexing, items in the list are assigned a priority metric in the form of an integer. The priority is determined by the relation between items in that list. This metric is a major factor in determining the order in which that item is presented to the user for selection. The kannuu index allows for customization of this integer to ensure that higher priority items show up sooner in the selection process. The process of indexing source data is extremely fast. To ensure this, kannuu employs a linear indexing methodology so that as items are added to source list, the additional time required for indexing the list is as efficient as mathematically possible.