Range-conditional character string retrieving method and system

Abstract

A range-conditional character string retrieving method and system capable of performing retrieval of a numerical value from a character string at an increased speed by shortening the time taken for generation of finite automaton, range condition retrieval for a character string containing admixedly numeric characters and non-numeric characters such as alphabetic letters and highly intelligent retrieval of a numerical value with designation of preceding and succeeding characters. Given range condition is partitioned in accordance with difference in the digit number between upper and lower limit values, whereon retrieval is performed in each of partitioned ranges in parallel. When a finite automaton transits from a predetermined state to at least two state in dependence on the result of collation of a character string subjected to retrieval, conditions for the state transitions are designated in terms of corresponding codes. A numerical value detecting unit for detecting a numerical value of interest from the character string subjected to retrieval is provided in association with a range decision unit for deciding whether the numerical value detected by the numerical value detecting unit falls within a specified range. A character string collating unit for retrieving a specific character string from the string subjected to retrieval is provided in association with a range condition collating unit for detecting a numerical value falling within a specific range from the specific character string.