JPS60142734A - Character string comparator - Google Patents

Abstract

PURPOSE:To attain dissidence of a character string at high speed by detecting the dissidence of the character string through the comparison of hash values. CONSTITUTION:A character string data I functioning as a reference is inputted to input data. The data I is applied with hash processing with the 1st hash processing section 3 and the hash value obtained as the result is set to a section 5-2 of a character string data register 5. On the other hand, a character string data IIcompared with the identity is inputted to the input register 2, the data II is applied with has processing by the 2nd hash processing section 4 and the hash value obtained is set to a section 6-2 of a character string data register 6. The set hash value to sections 5-2, 6-2 is compared by the comparator 7 and when the hash value is different, the comparator 7 outputs 0 and an inverter 8 outputs logical 1. Thus, when the hash value is effective and dissident with each other, a dissident signal is outputted from an AND gate 9.

Description

[Detailed Description of the Invention] [Technical Field of the Invention] The present invention relates to a character string comparison device, and in particular to a preprocessing method for detecting coincidence by quickly finding out whether character strings to be compared are inconsistent. The present invention relates to a character string comparison device that can perform high-speed comparison.

[Technology background]

A data processing device not only processes numbers, but also processes characters. For example, processing is performed to detect whether two character strings match or to extract specific characters from a character string. and rI AM
Processing such as changing rAMJ of A BOY, j to j"WASJ is also performed. In order to extract rAMJ in such a case, it is necessary to compare each character string.

By the way, when processing characters, each character is represented by a 1-byte code, so to compare whether strings match or not, you must compare the character codes that make up each string. . However, when there are many character strings to be compared, high-speed processing is not possible because the entire process requires a very long time when matching is detected by comparing each character code.

[Conventional technology and problems]

Therefore, the number of candidates to be compared is reduced by first detecting those that do not match among the comparison targets and reducing the number of comparison targets for character codes.

Conventionally, when comparing string data like this, in order to speed up the detection of inconsistencies, the length information of the strings is used to calculate the length before comparing the strings themselves. A comparison was made and a comparison candidate was selected. However, with this method, the lengths of the strings are the same and the strings themselves are often different, so it is insufficient to narrow down the comparison targets, and this method still takes a considerable amount of time to process. There were drawbacks.

[Purpose of the invention]

SUMMARY OF THE INVENTION An object of the present invention is to provide a character string comparison device that can improve such drawbacks and more accurately detect inconsistencies in character strings.

[Structure of the invention]

In order to achieve this purpose, in the character string comparison device of the present invention, in a data processing device that processes character string data,
a hash processing unit that outputs a hash value of a character string; a validity indication data holding unit that indicates whether or not the hash value is valid;
Comparing means for detecting whether the hash values of a plurality of character strings match is provided, and it is detected that the hash values of each character string are valid and that the hash values of each character string are different.

[Embodiments of the invention]

An embodiment of the present invention will be described based on the accompanying drawings. In the figure, ■ is the first manual register, 2 is the second manual register, 3 is the first hash processing section, 4 is the second hash processing section, 5 is the first character string data register, and 6 is the second character string data. - Register, 7 is a comparator, 8 is an inverter, 9 is an AND gate.

The first human input register 1 is a register into which internal representation data of a character string to be used as a reference at the time of comparison, such as character string data such as a character code, is input. This is a register into which a comparison character string is input to detect whether or not the character string has a hash value. The first hash processing unit 3 performs processing based on a hash function, such as adding up the sum of each character code, on the character string data set in the first manual register 1, and calculates the hash value of the calculation result. It outputs
The second hash processing section 4 performs the same processing as the first batch processing section 3. The first character string data register 5 is a register in which various data of a standard character string is written, and the length of the character string is written in a section 5-1.
, a section 5-2 in which the hash value calculated by the first hash processing unit 3 is entered, a section 5-3 in which a validity flag Vl indicating whether this hash value is valid or invalid is entered, and the characters of the character string itself. It has sections 5-4, etc. in which column data is entered.

Similarly to the first character string data register 5, the second character string data register 6 also has a division 6-1.6-2.6-3.6.
-4, etc.

Note that the valid flags Vl and V2 have the following effects.

In other words, when processing is performed where string data is frequently rewritten, a valid flag is provided to deal with the overhead of hash value calculations occurring many times. This indicates that the hash value is invalid when the data is rewritten. This allows for flexible handling depending on the processing target.

Comparator 7 compares the hash values entered in section 5-2 and section 6-2, and when the hash values are the same, "
1”, outputs rOJ when different.

Next, the operation of the present invention will be explained.

First, character string data of a reference character string is input into the first manual register 1 . This character string data is sent to section 5-4 of the first character string data register 5 and hashed by the first hash processing section 3, and the hash value obtained as a result of the operation is stored in the first character string data register. 5, section 5-2. The length of the character string data is determined by a length detecting means (not shown), and this length is set in the section 5-1.

The character string data of the character string to be compared for identity with the reference character string is inputted to the second human input register 2, and the character string data is hash-processed in the second hash processing section 4 to obtain a hash. The value is set in section 6-2 of second string data register 6, and the string data itself is also set in section 6-4. Data regarding the length of the character strings to be compared is also similarly set in section 6-1. Then, the hash values set in section 5-2.6-2 are compared in comparator 7, and when these hash values are different, comparator 7 outputs "0" and inverter 8 outputs "1". do. Therefore, when these hash values are valid and do not match, the AND gate 9 outputs a mismatch signal indicating the mismatch of the hash values, so that the character string set in the second character string data register 6 is It is easy to see that this character string is different from the character string set in the 1 character string data register 5.

If these hash values match, the comparator 7 will be "1"
Since the inverter 8 outputs rOJ, the AND gate 9 does not output a mismatch signal. Therefore, at this time, each valid flag v1, ■2 is "1" indicating the valid state.
”, the string data set in the second string data register 6 is transferred to the first string data register 5.
This means that there is a possibility that it will match the string data set in .

In this way, it is possible to speed up the detection of inconsistency between strings with high accuracy, so when there are many strings to be compared, the number of strings that need to be accurately compared for identity can be greatly reduced by character codes. , can be reduced quickly, and as a result, consistency can be determined quickly.

In the above description, an example has been described in which the sum of each character code is calculated as a hash function, but the hash function is of course not limited to this.

Although an example has been described in which two input registers and two hash processing sections are provided, it is of course possible to provide one each.

〔Effect of the invention〕

According to the present invention, inconsistency in character strings can not only be detected quickly by comparing their hash values, but also effective in cases where character string data is frequently rewritten. By using the flag, recalculation of the hash value can be suppressed, so it can be handled flexibly depending on the processing target.

[Brief explanation of the drawing]

The figure is a configuration diagram of an embodiment of the present invention. In the figure, 1 is the first manual register, 2 is the second manual register,
3 is a first hash processing section, 4 is a second hash processing section, 5
is a first string data register, 6 is a second string data register, 7 is a comparator, 8 is an inverter, and 9 is an AND gate. Patent Applicant: Fujitsu Ltd. Representative Patent Attorney Akira Yamatani

Claims (1)

[Claims] In a data processing device that processes character string data, a hash processing unit that outputs hash values of character strings, a validity indication data holding means that indicates whether or not the hash values are valid, and hash values of a plurality of character strings match. A character string comparison device includes a comparison means for detecting whether the hash values of each character string are valid and that the hash values of each character string are different.