JPH02109166A - Retrieving device for character string - Google Patents

Abstract

PURPOSE:To perform an independent program searching action of a field at a high speed and to realize a process in terms of a pipeline by using the head address of the field stored simultaneously with the stream data at a process part. CONSTITUTION:An input part 20 detects the division of the record of the stream data and writes one record into a record memory part 30. At the same time, a field control part 21 obtains the head address of each field out of the input stream data and writes it into an address memory part 40. A process part 22 processes the data on a record memory part 31 where one record is already written. In this case, the field number is used a the address of an address memory part 41 to read out the head address of a designated field of the part 31. An output part 23 outputs the contents of a record memory part 32 where the processed record is written to a line 11. Then the parts (30, 40), (31, 41) and (32, 42) are connected again to the parts 22, 23 and 20 respectively via a memory part selector 50 for execution of each process.

Description

DETAILED DESCRIPTION OF THE INVENTION FIELD OF INDUSTRIAL APPLICATION The present invention relates to a character string search device used in database systems, etc., which performs processing centered on searching large amounts of structured data.

2. Description of the Related Art Recently, in fields such as database systems, high-speed processing such as searching and replacing large amounts of data is desired. Conventional processing is performed in software using a general-purpose processor.

Problems to be Solved by the Invention However, there is a problem in that software processing using such a general-purpose processor takes too much time.

The present invention solves the above-mentioned problems of the prior art, and aims to provide a character string search device that enables processing such as search and replacement at high speed.Means for Solving the Problems The present invention provides an input section for inputting a stream delta having a data structure in which one record is a set of one or more fields, a field control section for obtaining the start address of each field from the input stream data, and a field control section for calculating the start address of each field from the input stream data. A processing unit that searches and processes the data in record units based on preset search data, an output unit that streams the processing results and outputs them in record units, and n units (n units) that store stream data in record units. n≧3) record storage units, and n (
n≧3: However, the same number of address storage sections as record storage sections are provided, and the record storage section and address storage section are paired for the input section, the processing section, and the output section, respectively, and are sequentially switched and used, The above object is achieved by the input section, processing section, and output section operating independently.

Effect of the present invention In the above configuration, the input section and the processing section output section operate independently, and the processing section uses the start address of the field stored at the same time as the stream data. Field cueing is performed independently at high speed by -0, and processing can be performed in a so-called lifeline manner.

EXAMPLE Below, an example of the present invention will be described in detail with reference to the drawings. FIG. 1 shows the relationship between the stream data stored in the record storage section and the start address of each field stored in the address storage section in the character string search device of the present invention. In Figure 2, A is S) I
It is a record storage unit that stores J-mu data, and
This is an address storage section that stores the start address of each field. The contents stored in the record storage section A are one record written in solid form as is, and no distinction is made between fields or the like.

The contents stored in the address storage section B are the addresses of each field of the contents stored in the record storage section A, and are offset addresses from the beginning of the record storage section. That is, the first address of the first field of the record stored in the record storage unit is stored in the first location of address storage unit B.
In the second location of the address storage section B, the start address of the second field of the record stored in the record section 8 is stored. Therefore, when reading, for example, the 5th field from the record storage section A, the content stored in address storage section 1305 is read out, and that value is used as the read address of the record storage section A. You can easily find the cue.

FIG. 1 is a block diagram of a character string search device according to an embodiment of the present invention.

In the figure, 10 is an input line for transmitting data to be processed, and 20 is an input unit for inputting stream data from the input line 10. 2] is a field control unit that calculates the start address of each field from manually-processed stream data; 22 is a processing unit that performs processing such as searching and adding data to the manually-processed data; and 23 is a processed data 30.31.32 is a record storage section that stores stream data in record units; 40, old; 7I2 is an output section that outputs J: to the field control section;
The record storage unit 30 is an address storage unit that stores the start address of each field in the record obtained by
．． 31 and 32, respectively, operate as bare. 50 is the bear (3o, 40), (3
This is a storage unit selector that selects the storage units (30, 40), (31, 41,), (32, 42).
2, 42) is connected to any one of the input section 20, the processing section 21, and the output section 22 via the storage section selector 50. Reference numeral 11 is an output line, which is a transmission line for sending processing results to other devices.

The operation of the above configuration will be explained below.

First, the processing unit 22 is given an instruction specifying a search character and processing content, and an insertion character string in advance. For example, the command may be such that when data in a certain field matches a search string, it is replaced with an inserted string.

The operation of the character string search device in this embodiment when it is performing steady VC operation will be described below.

At any time, store (:(0,40), (:(■
41) and (32, 42> are connected in pairs to either the manpower section 20, the processing section 22, or the cutting edge section 23.

1 storage unit (30, 40), (31,, 41), (32
, 42) are the input section 20, the processing section 22, and the output section 2, respectively.
The explanation will be based on the assumption that 3& is connected.

The input unit 20 detects the break between records of the manually inputted stream data and transfers one record to the record storage unit 30.
Enter VC4. At the same time, the field control unit 21 determines the start address of each field from the input stream data and writes it into the address storage unit 40. The processing unit 22 processes the data in the storage unit 31 whose I record has already been written by the input unit 20 . This process is
This is performed according to a command set in advance in the processing unit 22, and the field specified by the command uses the field number as the address of the address storage unit 41, and the content, that is, the record storage unit, is transferred from the address storage unit 41. This is done by reading out the start address of the designated field of No. 31. The output unit 24 has already received the processing unit 22.
Storage unit 3 in which records processed by are written
Output the contents of 2 on line l 1. , &ko output.

As described above, when all of the human power section 20, processing section 22, and output section 23 finish processing one record stored in the storage sections connected to each,
The storage unit selector 50 selects the storage units (30, 40), (3
1, 41) and (32, 42) are reconnected to the processing section 22, output section 23, and human power section 20, respectively. In this state, when an operation similar to the above-mentioned operation is performed, the processing unit 22 processes the record that was written to the record storage unit 30 by the human power unit 20Vc before reconnecting. The output unit 23 outputs the records processed by the unit 22 and written to the record storage unit 31 to the output line 11, and records that have already been output to the output line 11 and do not need to be stored are stored. The input unit 20 writes a new record into the record storage unit 32 that is currently being stored.

As described above, each time the input section 20, the processing section 22, and the output section 23 finish the operation for one record, the storage section selector 50 selects the storage section ( 30, 40) to the processing unit 22, and then the storage unit (31, 41) that was connected to the processing unit 22.
After reconnecting to the output unit 24 and reconnecting the storage unit (32, 42) that was connected to the output unit 23 to the human power unit 20, the input unit 20, processing unit 22, and output unit 23 each store one record. Repeat this action. In this way, data can be processed one after another.

However, before moving to this regular operation and after the regular operation ends, some of the human power unit 20, the processing unit 22, and the output unit 23 operate, and some do not.

A situation arises where there is no such thing.

As described above, in this embodiment, the input section 20, the processing section 22, and the output section 23 operate independently, and the processing section uses the start address of the field stored simultaneously with the stream data. Since the beginning of fields is performed independently and at high speed, so-called vibe-line processing is possible, stream data can be processed efficiently and high-speed processing can be achieved.

As described in detail, the present invention provides a method for determining the start address of each field when inputting system data.
It records stream data at high speed, inputs, processes, and outputs them independently and synchronously, and outputs them as IJ-system data, so processing is performed in a pipeline. We can achieve high-speed character string searches.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing the overall configuration of a character string search device in an embodiment of the present invention, and FIG. 2 shows stream data stored in a record storage unit and address storage unit in an embodiment of the invention. FIG. 3 is a diagram showing the relationship between the starting addresses of each field. ]0・Input line, 11...Output line, 20
- Input unit, 21... Field control unit, 22 - Processing unit, 23... Output unit, 30.31. , 32...Record storage unit, 40, 41.42...Address storage unit, 5
0...Storage section selector.

Claims (1)

[Claims] An input section for inputting stream data having a data structure in which one record is a collection of one or more fields, a field control section for calculating the start address of each field from the input stream data, and presetting of input data. Search record by record based on search data,
a processing unit that performs processing, an output unit that outputs processing results in record units as stream data, a storage unit for n records (n≧3) that stores stream data in record units, and a field control unit that n address storage sections are provided for storing the start address of each field in the record, and the record storage section and the address storage section are paired for the input section, the processing section, and the output section, respectively, and are used by switching sequentially. . A character string search device, wherein the input section, the processing section, and the output section operate independently.