JPH02136969A - Retrieving system for character string - Google Patents

Abstract

PURPOSE:To retrieve the character strings with reduce back tracks at collation by reducing gradually the table collating range when the collation is carried out between the character string to be retrieved and a retrieving character string table. CONSTITUTION:A retrieving character string table is prepared to sort and hold plural retrieving character strings. Then a partial character string starting at an optional position of a character string to the retrieved is collated with the character strings stored in the retrieving character string table with no occurrence of back tracks by reducing gradually the collating range of the table with use of the wild characters. Thus the back tracks are decreased at retrieval of character strings and the collating process efficiency is improved.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a string search method, and particularly to a string search method that can be applied to string processing tools such as text editors and file conversion programs.

(Prior Art) Conventional search methods for searching for multiple search strings from a search string include a method in which the search string table is a hash table, and a method in which the search string is expressed as a finite O-man. etc.

(Problem to be Solved by the Invention) Among the conventional methods described above, the method using a hash table can divide a searched character string into appropriate words, and search for some of those words. It is valid in some cases. However, if a plurality of arbitrary character strings are given as the search string, in the worst case, the hash function must be applied to all substrings of the search string. In other words, the method using a hash table is inappropriate when searching for a plurality of arbitrary character strings. Furthermore, it is difficult to apply this method when the search string contains wild characters.

Furthermore, the method using a finite automaton has the disadvantage that when there are many or long search strings, the finite automaton becomes large and backtracking during the search increases, resulting in a decrease in processing efficiency.

(Means for Solving the Problem) The character string search method according to the present invention includes a sorting step of sorting a plurality of search strings including wild characters in the order of character codes and storing them in a search string table; backtracking between a substring starting from an arbitrary position in the column and the string stored in the search string table by successively narrowing the matching range of the search string table using the wild character; and an output step of outputting the character string successfully matched in the matching step or the index of the character string in the searched character string.

(Example) Hereinafter, the present invention will be explained with reference to the drawings.

FIG. 1 is a flowchart showing the configuration and processing procedure of the present invention. Reference numeral 11 holds a plurality of character strings containing wild characters.The plurality of character strings containing wild characters held in the search string table 11 are sorted in the order of character codes in a sorting step 12. On the other hand, the searched character string 13 is compared with the contents of the searched string table 11 in a matching step 14 .

The character string successfully matched in the matching process 14 or the index of that character string in the searched character string is outputted as output data 16 through the process 15.

That is, as described above, the processing of this embodiment includes the sorting step 12, the collation step 14. This is performed in the order of output step 15. Sorting of search strings in the sorting step 12 can be performed using a well-known algorithm such as quick sort.

The output data 16 can be output to a suitable medium such as a computer memory, secondary storage such as a magnetic disk, a printer, or a CRT display.

Next, it is a flowchart showing in detail the processing procedure of the verification step 14 in this embodiment. Second
This will be explained in detail with reference to the drawings.

First, in step 21, the search string pointer is positioned at the beginning of the search string, and in step 22,
Checks whether scanning of the searched string is finished,
If not completed, perform pattern matching processing (
In step 23), in step 24, it is determined whether there is a matched character string, and if there is, in step 25, that character string or the index of that character string in the searched character string is written to the output buffer. , Then, in step 26, the searched character string pointer is counted up. If there is a matched character string at this time, the length of that character string is counted up, and if there is no match, only one is counted up, and step 22
Return to processing.

FIG. 3 shows a data configuration diagram when performing the pattern matching process 23. In the figure, 31 is a part of the character string to be searched, and 32 is a part of the search string table, in which individual search strings are sorted and stored in each row of the table. Search strings starting with wild characters are stored sorted towards the back of the table. Here, “?” is a wild character for alphabetic characters, “&
” indicates a wild character for numbers.

The algorithm for pattern matching processing will be explained below.

First, assume that the searched character string pointer is positioned at "A" of character string 31. ? lFirst, set the collation pointer to 1, and register the positions in the table of the character string 33 starting with "A" in the search string table and the character string 34 starting with the wild character "?" in the collation range table. , is the initial value of the matching range.

When viewed from the matching range table, the search string table appears in the form shown at 35 in the figure. That is, the character string 33 starting with A" and the character string 34 starting with the wild character "7" in the original search string table 32.
It will look like it has been compressed. Note that if the range within the table for each first character is separately stored in an index table in the sorting step, there is no need to search the search string table when setting the initial value of the collation range. “A” or ゛? If there is no search string that starts with "pa", there will be no matching string.9If the matching range has been set, move the matching pointer one step and match the next character of the searched string. In this case, the search strings registered in the matching range table are compared, but if there is a search string that ends there (for example, the search string in the first line of 35), a match is made. Assume that there is a string. For search strings that are successfully matched, the indexes in the search string table are compressed into a matching range table. As a result, ranges 36 and 37 of 35 in the figure are compressed, and the search string table appears as shown in 38 in the figure when viewed from the matching range table.

Further, the verification pointer is moved and the same process is repeated while narrowing the verification range one after another, and the repetition ends when the verification range reaches O (stage 1).

(Effects of the Invention) As explained above, the present invention includes a search string table that sorts and holds a plurality of search strings, and when matching a searched string with the search string table, the table matching range is set. Since the matching process is performed to narrow down the strings one after another, it is possible to perform a string search with less backtracking during matching.

[Brief explanation of the drawing]

FIG. 1 is a flowchart showing the configuration processing means of an embodiment of the character string search method according to the present invention, FIG. 2 is a flowchart showing in detail the processing procedure of the collation step in FIG. 1,
FIG. 3 is a data structure diagram when performing the pattern matching process in FIG. 2. 11... Search string table, 12... Sorting process, 13... Searched string, 14... Collation process, 15
...Output process, 16...Output data.

Claims (1)

[Claims] a sorting step of sorting a plurality of search strings including wild characters in the order of their character codes and storing them in a search string table; and a partial string starting from an arbitrary position of the searched string and storing it in the search string table. A matching process in which a character string is matched without backtracking by successively narrowing the matching range of the search string table using the wild character, and a character string successfully matched in the matching process. or an output step of outputting an index of the character string in a searched character string.