Bit string merge sort device, method, and program

Abstract

It is possible to realize a high-speed merge sort method using a coupled node tree. A minimum or a maximum key is extracted from a plurality of sorted key storage regions containing a key formed by a sorted bit string. A coupled node tree for merging is generated by adding a processing source identifier to identify the sorted key storage region as an extraction source. A write-out of a key into the merged key storage region by the minimum value or the maximum value search of the coupled node tree, a deletion of the key, and an extraction of the key from the sorted key storage region are repeatedly executed for an insert process.

Description

technical field [0001] The present invention relates to a device and method for merging and sorting bit sequences, and a program for enabling a computer to execute the method. Background technique [0002] In recent years, the informatization of society has continued to develop, and large-scale databases have been used everywhere. In order to search for records from such a large-scale database, the items in the records corresponding to the addresses where the records are stored are usually searched as index keys to search for desired records. Moreover, character strings in full-text search can also be regarded as index keywords of documents. [0003] Moreover, since these index keywords are expressed by bit sequences, the retrieval of the database can be attributed to the retrieval of bit sequences. [0004] On the other hand, as the processing related to the database, as described in the following patent document 1 and the following patent document 2, record merge sorting...

AI Technical Summary

Problems solved by technology

[0021] As described above, in the retrieval process using the Patricia tree, there are advantages such as that the retrieval can be performed only by checking the necessary bits, and that all keywords need only be compared once. However, there are disadvantages such as the following. Two links from the node increase the storage capacity, complex judgment processing due to the existence of backlinks, delay in retrieval processing due to comparison with index keys after returning from the backlink, addition and deletion, etc. Data maintenance is difficult

[0023] However, in the technique disclosed in the following Patent Document 3, since one node necessarily occupies the index key area and the pointer area, one pointer is used to store the lower left and right nodes in a continuous area, so for example Figure 1B The left pointer 1740c and right pointer 1741h shown as the bottom part of the Patricia tree also need to allocate a storage area with the same capacity as the node, and the reduction effect of the storage capacity is not very good

In addition, it does not improve the delay of retrieval processing caused by backlinks, and it is difficult to perform processing such as addition and deletion.

[0024] Therefore, when sorting and merging a large amount of data is to be performed, a large amount of computer resources will be occupied for a long time, and the cost is very high

Images