Algorithm for dividing a sequence of values into chunks using breakpoints

摘要

A method of dividing a sequence of data into chunks using a sliding window uses an algorithm to compare fingerprint values for each position within the sequence against sets of criteria to create breakpoints. When a fingerprint value does not satisfy a first set of criteria a second set are applied and if satisfied a potential breakpoint is identified. Subsequently, if a fingerprint value that satisfies the first set of criteria is not found before the maximum chunk size is reached the potential breakpoint is designated as a breakpoint. Additional constraints on minimum and maximum sizes of chunks can be used to further refine the method. Further sets of criteria may be used if a fingerprint value does not meet wither of the two initial criteria sets. Preferably the fingerprint value can be identified using Rabin's Fingerprint algorithm.