JP2590698B2 - Character string data retrieval device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for retrieving character string data, and more particularly to a method for retrieving a character string at high speed from a data structure to be searched.

[0002]

2. Description of the Related Art In a conventional character string data search method, when registering character string data in a data structure to be searched, a character string serving as a data classification key is fixed to one, so that the data is stored later. Even in the case of addition, the character string serving as the data classification key is fixed as it is, and the search target data to be added is added.

[0003] As a reference of such a conventional character string data search method, there is "Japanese Patent Laid-Open No. Sho 62-241026 Character String Search Method".

[0004]

In the above-described conventional character string data search method, the number of data classification keys used when creating a data structure to be searched is fixed to one. There is a disadvantage that when the search target data is registered, there is a possibility that a large deviation may occur in the numbers registered in the data structure.

FIG. 6 is a diagram showing an example of a hash table representing a data structure in which search target data is registered by a conventional character string data search method. For example, an open hash method is used as a data structure to be searched, and the number of hash elements is set to 5
When the data classification key is fixed to the first character from the beginning of the character string, {A0, B0, C
When 0, D0, D1, D2, D3, D4, D5, D6} are input, a large bias as shown in FIG. 6 occurs for each number registered in the created data structure. From such a search target data structure, even if the hash method is used, there is a position with a large amount of data.
This means that character strings cannot be searched at high speed.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a character string data search method capable of searching for a character string in a short time while reducing a bias in a data structure of a search target.

[0007]

According to the present invention, a character string data detection method according to the present invention is provided.
The search device determines the rule for determining the data classification key and the number of data to be searched, and determines the number of data classification keys to be reviewed. The initial setting unit performs initial registration, and inputs the character string data to be searched. A search target data input unit that sends out the search target data for the number determined by the initial setting unit, and classifying the search target data received from the search target data input unit using each of the data classification keys. A data registration unit for registering the search target data in the data structure, and a difference between the maximum number and the minimum number of search target data included in the data structure registered by the data registration unit for each data classification key. And a data analysis unit for calculating the value and storing the value as a data structure analysis condition;
The data classification key used in the data registration unit is reviewed for each number of search target data determined in the initial setting unit, the data structure analysis condition of the data analysis unit is changed to a smaller data classification key, and the data registration key is changed. A data classification key changing unit for re-registering the search target data input in the search target data input unit with the changed data classification key by returning to the unit, a search character string input unit for inputting a search character string to be searched, Searching the search character string input in the search character string input unit using the data structure analysis condition stored in the data analysis unit using a classification with a small data classification key from the data structure of the search target registered by the data registration unit. includes a character string search unit that, and an output unit for outputting a search result of a search with the character string search unit.

Further, the character string data search method according to the second invention is characterized in that: (A) a rule for determining a data classification key and an initial setting by defining how many data to be searched should be reviewed every time data to be searched is registered; (B) a search target data input unit that inputs character string data to be searched and sends out the search target data for the number determined by the initial setting unit; and (C) the search target data. A data registration unit for classifying the search target data received from the input unit using each of the data classification keys to register the search target data in a data structure; and (D) for each of the data classification keys, A data analysis unit that calculates a difference between the maximum number and the minimum number of search target data included in the data structure registered by the data registration unit, and stores the value as a data structure analysis condition; E) The data classification key used in the data registration unit is reviewed for each number of search target data determined in the initial setting unit, and the data structure analysis condition of the data analysis unit is changed to a smaller data classification key. A data classification key changing unit for returning the search target data input at the search target data input unit by the changed data classification key by returning to the data registration unit; and (F) a search character for inputting a search character string to be searched A column input section,
(G) a search character input from the search character string input unit using a classification based on a data classification key having a small data structure analysis condition stored in the data analysis unit from the data structure to be searched registered by the data registration unit; A character string search unit for searching a column; and (H) an output unit for outputting a search result searched by the character string search unit.

[0009]

Next, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing one embodiment of a character string data search method according to the present invention. As shown in FIG. 1, the character string data search method according to the present embodiment includes a search target data creation unit 109 and a search result creation unit 110.

In the search target data creation unit 109, the initial setting unit 101 determines the data classification key determination rule and the number of data search keys to be changed each time the search target data is registered. . Further, the search target data input unit 102 inputs the search target data and sends the data registration unit 103 with the search target data of the number determined by the initial setting unit 101.

The data registration unit 103 uses a part of the character string of the search target data for the search target data sent from the search target data input unit 102 as a data classification key, and performs the search according to the data classification key. The data is registered in the data structure, and the data analysis unit 104 is operated.

The data analysis unit 104 calculates the difference between the maximum number and the minimum number of search target data included in the data structure as a data structure analysis condition, saves the result, and then classifies the data. The key changing unit 105 is operated.

Therefore, the data classification key changing unit 105
The data classification key used in the data registration unit 103 is
The data is reviewed for each number of search target data determined by the initial setting unit 101, and the data structure analysis condition of the data analysis unit 104 is changed to a smaller data classification key, and returned to the data registration unit 103 through the data analysis unit 104. The search target data input unit 102 is
Is re-registered in the data structure.

Then, for all data classification keys,
When the registration and analysis of the data structure are completed, the data registration unit 103 reconstructs the data structure of the search target using the data classification key having the smallest data structure analysis condition, and executes the next search target data. Proceed to registration.

Thus, the search target data input unit 1
By repeating the transmission of the number of search target data determined by the initial setting unit 101 from 02 to the data registration unit 103, these processes are repeated until there is no more search target data to be registered.

Further, in the search result creating unit 110, the search character string input unit 106 inputs the search character string data, and operates the character string search unit 107. The character string search unit 107 searches the search target data created by the search target data creation unit 109 for the search character string input by the search character string input unit 106, and operates the output unit 108. Note that this search uses the data classification key finally used when the data registration unit 103 registers the search target data. The output unit 108 outputs a search result searched by the character string search unit 107.

FIG. 2 is a flowchart showing an example of the operation of the data registration unit 103, the data analysis unit 104, and the data classification key change unit 105 according to this embodiment. Also, FIG.
FIG. 5 is a diagram illustrating an example of a hash table representing a data structure in which search target data is registered in the present embodiment.

Next, the operation of this embodiment shown in FIG. 1 will be described with reference to FIG. Here, the open hash method is used for the data structure to be searched, and the number of hash elements is set to 5, and {A0, B0, C0, D0, D1, D2,
An example will be described in which data D3, D4, D5, D6} are registered as search target data.

First, the initial setting unit 101 extracts one character from the head of character string data and sets the remainder obtained by dividing the character code by 5 as a data classification key. The key is reviewed, and when the data to be registered is completed, the data classification key is reviewed again. Also, the search target data input unit 102 stores the registered data {A0, B0, C0, D0, D
1, D2, D3, D4, D5, D6} are first read into the data registration unit 103 {A0, B0, C0, D0, D1}.
Has been sent.

Therefore, as shown in FIG. 2, the data registration unit 103 determines in step 201 the remainder obtained by dividing the character code of the first character from the beginning of the character string by 5 as a data classification key, and in step 202 The registered data is classified into hash elements using the data classification key, and registered as shown in FIG. 3, and the data analysis unit 104 is operated.

Then, in step 203, the data analysis unit 104 stores a value obtained by subtracting the minimum number of hash elements from the maximum number of hash elements as a data structure analysis condition, and the data classification key change unit 105 It is working. In this case, the data structure analysis conditions range from a maximum of 2 to a minimum of 0.
To 2

In step 204, since the data classification key changing unit 105 has not yet checked the second character from the beginning, the process proceeds to step 205, and
At step 05, the data classification key is changed to a value obtained by dividing the second character from the head of the character string by 5, and the process returns to step 202.

In step 202, the data registration unit 103 classifies the currently registered search target data using the data classification key obtained in step 205 and re-registers the data as shown in FIG. The analysis unit 104 is operating. Then, in step 203, the data analysis unit 1
In step 04, when the data structure analysis condition is obtained, the minimum 0 is subtracted from the maximum 4 to become 4. The result is stored and the data classification key changing unit 105 is operated.

In step 204, the data classification key changing unit 105 determines in this example that only the remainder of the first character divided by 5 and the second character divided by 5 is used as the data classification key. Since all the data classification keys have been checked, the process proceeds to step 206.

In step 206, the data analysis unit 104 compares the data structure analysis conditions obtained in step 203, and in this example, uses the remainder obtained by dividing the first character by 5 as the data classification key. Since the data structure analysis condition is smaller, the data registration unit 103 is operated using the remainder obtained by dividing the first character by 5 as a data classification key. In step 207, the data registration unit 103 sets up the data structure. From the beginning, five A0,
Data registration for B0, C0, D0, D1} has been completed.

Next, the search target data input unit 102
{D2, D3, D4, D5, D6} are stored in the data registration unit 10
3, and {A0, B0, C0, D0, D1} are additionally registered in the same manner as when registered, and the registration of the search target data is completed. As a result of this additional registration, the bias is reduced when the second character is used as the data classification key. Therefore, the data classification key is changed to a value obtained by dividing the character code of the second character by 5 as shown in FIG. Are classified.

On the other hand, to perform a character string search from the registered search target data, the search character string input unit 106 inputs the search character string data and sends it to the character string search unit 107. Is the search target data creation unit 10
The search character string is searched using the search target data created in step 9 and the search character string input by the search character string input unit 106, and the output unit 108 outputs the result.

For example, when searching for the character string D6, the value obtained by dividing the character code of the second character by 5 is 4
Therefore, if the position of 4 in the hash table of FIG.
D6 comes out second. In this way, the character string can be searched at high speed by reviewing the data classification key of data registration every time five data to be searched are registered.

The processing time formula when registering the search target data by the open hash method is shown below.

(N * s + c) * k * y + (x * s + c) * k + b In this equation, the number of data classification key candidates is k,
The number of search target data is z, the time for registering one search target data is s, the time for the search target data to obtain a data structure analysis condition in the data structure of the search target is c, and the best one of the conditions is selected. The data classification key is reviewed every time n search target data is added with the time b, and the quotient obtained by dividing d by n is represented by y, and the remainder obtained by dividing d by n is represented by x.

[0031]

As described above, according to the character string data search method of the present invention, every time a fixed number of character string data are registered, the data classification key is reviewed and the data to be searched is re-registered. By doing so, even if you do not know what character string the search target data contains, you can perform data registration with the optimal data classification key, reducing the bias in the search target data structure This has the effect that a character string can be searched in a short time.

[Brief description of the drawings]

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

FIG. 2 is a flowchart illustrating an example of an operation of a data registration unit 103, a data analysis unit 104, and a data classification key change unit 105 according to the embodiment.

FIG. 3 shows the first character as a data classification key of 5 in the present embodiment.
FIG. 6 is a diagram illustrating an example of a hash table representing a data structure in which pieces of search target data are registered.

FIG. 4 shows the second character as a data classification key of 5 in the present embodiment.
FIG. 6 is a diagram illustrating an example of a hash table representing a data structure in which pieces of search target data are registered.

FIG. 5 is a diagram illustrating an example of a hash table representing a data structure in which a total of ten search target data are registered using a second character as a data classification key in the embodiment.

FIG. 6 is a diagram illustrating an example of a hash table representing a data structure in which ten search target data are registered using the first character as a data classification key in a conventional character string data search method.

[Explanation of symbols]

 Reference Signs List 101 initial setting unit 102 search target data input unit 103 data registration unit 104 data analysis unit 105 data classification key change unit 106 search character string input unit 107 character string search unit 108 output unit 109 search target data creation unit 110 search result creation unit

Claims (1)

(57) [Claims] 1. An initial setting unit for performing initial registration by determining a rule for determining a data classification key and how many data to be searched are to be reviewed each time data is registered, and inputting character string data to be searched. A search target data input unit for transmitting the search target data of the number determined by the initial setting unit, and classifying the search target data received from the search target data input unit using each of the data classification keys. A data registration unit for registering the data to be searched in the data structure, and a search for the maximum number and the minimum number of data to be searched included in the data structure registered by the data registration unit for each data classification key. A data analysis unit that calculates a difference and stores the value as a data structure analysis condition, and a data classification key used by the data registration unit are defined by the initial setting unit. The search target data input in the search target data input unit by reviewing for each number of search target data, changing the data structure analysis condition of the data analysis unit to a smaller data classification key, and returning to the data registration unit. A data classification key changing unit for re-registering with a changed data classification key, a search character string input unit for inputting a search character string to be searched, and a data analysis unit based on a data structure of a search target registered by the data registration unit. A character string search unit that searches for a search character string input in the search character string input unit using a classification based on a data classification key having a small data structure analysis condition saved in the search character string search unit; A character string data search device comprising: an output unit that outputs the character string data.