JPH04145579A - Content retrieving device - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a data processing device for manipulating data groups in a database, and in particular, to a data processing device for manipulating a data group in a database, and in particular for speeding up update processing with search conditions for columns that are not indexed in a database. The present invention relates to a content search device.

[Prior Art] A database update process with search conditions is a process of updating data in a specific column of a record that matches a specified search condition. The conventional update process with search conditions will be explained with reference to the drawings.

FIG. 3 is a diagram for explaining conventional update processing with search conditions. For the sake of explanation, it is assumed that the database has a format in which m records are stored divided into N pages. Each page (a fixed-length physical record that is read or written by one input/output command) 31 consists of a plurality of records 32, and each record 32 stores data in a first column 33, a second column 34, and a record. It consists of a third column 35 for storing numbers.

In such a database, a specific example of update processing with a search condition will be described below, in which the value "4n" is added to the second column 34 of a record whose first column 33 has a value of 5". However, it is assumed that the first column 33 does not have an index. Therefore, in order to find the record with the value 5'' in the first column 33, all pages are sequentially searched from the file storage device (not shown) in the main unit (not shown). The search condition for each record in the read page is determined (the value of the first column 33 = "5"), and the second column 34 of the record that matches the condition is updated.

That is, in the example of FIG. 3, since there is no record matching the search conditions on the first page, that page is discarded from the work area at the time the search ends. Since the second page has a record that meets the search conditions, the record is updated in the work area, and the updated second page is saved in the work area. Similarly, since there are no records that match the search conditions on the N-1st page, that page is discarded from the work area at the end of the search, and since there are records that match the search conditions on the N-th page, the page is discarded from the work area. The record is updated and the updated Nth page is saved in the work area. When the above processing for all pages is completed in this way, all updated pages in the work area are stored in the file storage device.

[Problem M that the invention seeks to solve]

In the above-mentioned search method, all pages are sequentially read from the file storage device into the memory and search processing is performed as described above, and therefore, high-speed processing cannot be performed.

Therefore, as a technology to speed up the search process, multiple pages of the file storage device are successively read based on commands from the main unit, and records that match the specified search conditions are selected from the records within the page. There is a known technique for speeding up the determination of search conditions for columns that are not indexed using a content search device that transfers the content to the main unit.

Figure 4 is a system configuration diagram equipped with a content search device.
1 is a content search device, 42 is a main unit, 43 is a database management system on the main unit, 44 is a memory work area, 4
5 is a file storage device. The content search device 41 searches the file storage device 45 based on instructions from the main device 42.
Continuously reads multiple pages of the page and selects records that match the specified search conditions from the records within the page. Transfer to 42. Such a search condition matching record group 50 is shown as a fifth factor, where record 51 is a record in page number 2, and record 52 is a record in page number N. That is, the conventional content search device 41 blocks a group of records matching the search conditions, configures the blocks into one page, and transfers the block to the main device 42 . The database management system 43 uses the unique key (in this example, the record number in the third column) of the transferred search condition matching record as an index to store the file storage device 4.
5 is searched, the page storing the corresponding record is read into the memory work area 44, and the second column of the record matching the search condition is updated. When the page that stores all the records matching the search conditions has been updated, the work area 44
All updated pages in the file are written to the file storage device 45.

By using the content search device, the search condition determination process for records across all pages is sped up, so the entire update process with search conditions for columns that are not indexed is sped up.

However, in the conventional processing method described above, in addition to searching by columns with specified search conditions, it is necessary to search by unique keys, which are not normally necessary, and the page to be updated must be searched twice. , the data must be read from the file storage device, which poses a problem in that speeding up is hindered.

An object of the present invention is to provide a content search device that further speeds up update processing with search conditions for columns that are not indexed.

[Means to solve the problem]

In order to achieve the above object, the present invention sequentially reads a plurality of pages stored in a file storage device in response to a command from a main unit, and extracts records that match specified search conditions from a group of records on each page. In a content search device that selects a group and transfers it to the main device, information indicating a page including a record matching the search condition, information indicating a record matching the search condition, and information indicating a page number of the page. The present invention is characterized by comprising means for transmitting the information to the main body device.

[For production]

According to the above-described means, it is possible to immediately update a record that matches a search condition on the work area of the main device, and it is possible to speed up update processing with a search condition for columns that are not indexed.

[Example] Hereinafter, an example of the present invention will be specifically described using the drawings.

FIG. 1 is a block diagram of an embodiment of the present invention, which includes a content search device 1, a main device 2, and a file storage device 3 of the present invention.
The mutual connection relationship with the conventional one is the same as that of the conventional one.

The content search device 1 of the present invention includes a file control unit 11 that controls an interface with the file storage device 3, and 2 that stores pages read from the file storage device 3.
an input buffer 12 configured with a screen, a search device 13 that determines search conditions, a search control unit 14 that controls the search device 13 and extracts record IDs of records that match the search conditions, and A two-page second report information buffer 15 stores second report information
, first report information consisting of a group of pages including records matching the search conditions in the input buffer 12 , and a channel control unit 16 that controls the transfer of the second report information to the main unit 2 .

FIG. 2 shows first report information 21 and second report information 22.
It is a figure showing a specific example. The first report information 21 is made up of pages in which records matching the search conditions exist, and according to the example of FIG. 3, it is made up of a group of pages with page numbers 12 and N. The second report information 22 is a record D of a record matching the search conditions, and this record ID is the information for recognizing the records matching the search conditions from among the pages transferred to the main device 2. The page contains information for writing the page to the file storage device 3.

That is, specifically, the record ID is composed of a set of a page number, an intra-page record number, and a relative page number. Here, the in-page record number is a number used to recognize the record that matches the search condition in the page. For example, in-page record number 2 of page number N means that the second record in page N is the number that matches the search condition. Indicates that it is a matching record. The relative page number is a number used to recognize each page transferred from the content search device 1 to the main device 2 on the work area.For example, relative page number 1 in page number 2 means that the page with page number 2 is This is the first #th page, indicating that the page with page number N is the second page. Then, as will be described later, the database management system performs record update processing in the order of the smallest relative page number and the smallest intra-page record number within the relative page number.

Further, the page number is used when writing the page for which update processing has been completed in the work area to the file storage device 3.

To explain in detail the present invention configured as above, the file control unit 11 continuously reads pages from the file storage device 3 and stores them alternately on two sides of the input buffer 12 (input buffer 12 The capacity of one side is N pages). The search 111#IWT 14 and the search device 13 determine search conditions for each record in the input buffer 12, generate record IDs of records that match the search conditions as shown in FIG. 2, and store them in the second report information buffer 15. do. When the search for all the records on the first side of the input buffer 12 is completed, the channel control unit 16 searches the page (first report information 21) in which the record matching the search condition exists in the input buffer 12 and the page in the second report information buffer 15. The second report information 22 is transferred to the main device 2.

The database management system 4 on the main device stores records that match the search conditions in the page, which is the first report information 21.
Based on the record ID which is the second report information 22, it is detected sequentially, that is, in order of the smallest relative page number and the record number within the page within the relative page number, and the second column of the detected record is stored in the work area. Update processing is performed on 5. When all update target pages have been updated on the work area 5, the updated pages are written to the file storage device 3 via the channel control n16 and the file control unit 11 based on the page number of the second report information 22. issued.

Subsequently, all records for the remaining side of the input buffer 12 are searched in the same manner as described above, and at the same time, the first side of the input buffer 12 searched earlier is continuously searched from the file storage device 3. N pages are stored.

In this way, by using the first and second report information of the present invention, it is possible to immediately update a record that matches the search condition in the work area of one main body device.

Note that the processing method of the database management system on the main body device and the configuration of the database are not limited to the above embodiments.

[Effects of the Invention] As described above, according to the present invention, when a record matching the search conditions is transferred to the main device, the page on which the record matching the search conditions exists and the process for recognizing the record matching the search conditions are provided. Since the information and the information indicating the page number of the page are transferred to the main unit, the record to be updated can be immediately updated without referring to the index or re-reading the page containing the record to be updated from the file storage device as in the past. Update processing can be performed. As a result, it is possible to further speed up update processing with search conditions for columns that are not indexed.

[Brief explanation of drawings]

Fig. 1 is a block diagram of an embodiment of the present invention, Fig. 2 is a diagram showing first report information and second report information of the present invention, and Fig. 3 explains conventional update processing with search conditions. Figure 4 for
Figure 5 is a system configuration diagram equipped with a conventional content search device.
The figure is a diagram showing a conventional search condition matching record group. 1...Content search device, 2-...Main device, 3...
File writing device, 11... File control unit, 12... Input buffer, 13... Search device, 14... Search control unit, 15
...Second report information buffer, 16...Channel control unit. Figure 4 Figure 5

Claims (1)

[Claims] (1) Continuously read multiple pages stored in the file storage device according to a command from the main unit, select a group of records that match the specified search conditions from the record group of each page, and send the selected records to the main unit. In the content search device to be transferred, means for transferring a page including a record matching the search condition, information indicating a record matching the search condition, and information indicating a page number of the page to the main device. A content search device comprising: