The positional indexing introduced in Section 6.1 has the benefit of allowing for a simple, yet efficient, implementation. Section 7.4.2 discuss the situations when even much simpler indexing schemes are possible. However, its major advantage is its flexibility. Since each daughter is allocated a separate entry in the chart, further information collected about the peculiarities of each mother-daughter pair (with respect to the unification of the respective mother and daughter) can be integrated into the indexing scheme. This is one of the fundamental differences between indexing and filtering, and another reason for preferring indexing over filtering.
As indicated in the indexing timeline introduced in Section 4.2.2, three points during parsing can be identified for indexing. In the first, a static analysis of grammar rules can determine what information can be extracted from categories that can be used as indexing keys. At this point, the Static Cut will be determined for each mother and daughter that are unifiable, and index keys will be identified through the Static Cut. During parsing, the second point occurs at the completion of a rule. At this moment, its mother is introduced as an edge into the chart. The Dynamic Cut is used here to further refine the index keys. The third point (occurring when a matching edge for a daughter is searched for in the chart) involves the extraction of the daughter's index key.