| US 7,526,497 B2 | ||
| Database retrieval apparatus, retrieval method, storage medium, and program | ||
| Hideo Kuboyama, Kanagawa (Japan); and Makoto Hirota, Tokyo (Japan) | ||
| Assigned to Canon Kabushiki Kaisha, Tokyo (Japan) | ||
| Filed on Mar. 13, 2007, as Appl. No. 11/717,044. | ||
| Application 11/717044 is a division of application No. 10/391578, filed on Mar. 20, 2003, granted, now 7,222,129. | ||
| Claims priority of application No. 2002-096179 (JP), filed on Mar. 29, 2002; and application No. 2002-127329 (JP), filed on Apr. 26, 2002. | ||
| Prior Publication US 2007/0174261 A1, Jul. 26, 2007 | ||
| Int. Cl. G06F 17/00 (2006.01) | ||
| U.S. Cl. 707—101 [707/102; 707/103 Y; 707/104.1] | 6 Claims |
| 1. A database retrieval method of performing retrieval processing for data stored in a database, characterized by comprising:
a storage step of dividing the database into a plurality of data blocks that include character data corresponding to search
keys, and storing the data blocks;
a determination step of determining a combination of the data blocks on the basis of retrieval priority levels and data sizes
of the respective data blocks such that a total sum of the data sizes of the data blocks becomes smaller than a predetermined
data size set by the user;
a loading step of loading the determined data blocks;
a memory step of temporarily storing the loaded data blocks to perform retrieval processing for a database of the loaded data
blocks; and
an output step of outputting data having undergone retrieval processing
wherein each data block has corresponding index information, and the plural index information are sorted based on the retrieval
priority levels of the corresponding data blocks, and
wherein in the determination step, a combination of the data blocks is determined on the basis of the plural index information,
each index information having a cumulative data size obtained by adding the data size of the corresponding data block to a
total sum of data sizes of data blocks higher in priority level than the corresponding data block.
|