JPS6289134A - Character string difference extracting method and its device - Google Patents

Abstract

PURPOSE:To efficiently execute a document and computer program processing maintaining work by searching for the common part string from the first and second character strings, deleting successively the common part string matching with the standard and extracting the remaining dissident part as the difference. CONSTITUTION:When a part string length is the reference length or above specified beforehand, it is unnecessary to scan the the comparing pair area of respective elements of the common part string, the scanning starting position is renewed, and when the length is the reference length or below, the next scanning position is obtained by a scanning position occurring part 301. By using the common part string obtained by searching the common part string, a searching result deciding part 303 divides two gives character strings into three, decides the coincidence or the difference of the part string and accommodating it to a deciding result memory part 202. When the procedure is repeated and the divided part string is all decided, it is judged to be the completion. Finally, by a difference extracting part 305 from a difference part string which accommodates the deciding result memory part 202, the difference is outputted to a CRT printer driving part 500 by three expressions of the adding, deleting and replacing and the whole processing is completed.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a method and apparatus for character string difference extraction used in the field of information processing using electronic computers. The method and apparatus according to the present invention are capable of comparing similar character strings and character strings, which are part of the work of processing and maintaining character strings such as documents written in everyday language and computer programs stored in information storage media. This is related to difference extraction.

[Prior art and problems to be solved by the invention]

The conventional method for character string comparison is to shift the combinations of the addition/deletion correction elements such as characters, lines, and terms that make up the string as appropriate, then compare and match them, and then use the matching elements as common elements, and repeat this process. , the final discrepancy is the difference.

For example, when comparing the first character string "Aiuni" and the second character string "Aie" using characters as elements, "A" and "A",
Since "i" and "i" and "tech" and "tech" are equal, [the second character string is the first character string with "tsu" removed. ’”.

This is a method of extracting differences.

However, the conventional method has the following problems.

In other words, appropriate results are given for character strings designed so that the same character does not appear again, but for normally written documents, computer programs, etc., even if these character strings are In most cases, it consists of a large number of identical elements, even on a line-by-line basis, and the differences tend to be inappropriately understood.

For example, when comparing the first character string "Inugairu" and the second character string "Nekogairu" character by character, the first character "i" of the first character string and the fourth character "i" of the second character string are the same. As a result of making it an element, ``The second string is r cat g j before the first string
``r nugai'' is added and ``nugai'' is removed. ”.

The simple difference expression in photometry was corrected from ``r dog'' to ``cat.'' ”, and the above result is not appropriate.

In this way, conventional methods cannot obtain appropriate results when comparing character strings that have many identical elements, such as ordinary documents, computer programs, etc. Therefore, it is necessary to mechanize the work of processing and maintaining these character strings. It is considered impossible,
The problem is that it places a heavy burden on humans.

In view of the above-mentioned problems, one object of the present invention is to efficiently mechanically obtain a simple difference expression when comparing a character string composed of many identical elements and a similar character string. The goal is to greatly improve the efficiency of document and computer program processing and maintenance work, which could only be done manually and with low efficiency in the past.

Another object of the present invention is to efficiently perform a comparison operation that requires a great deal of manual labor with a huge number of character strings, and to quickly and accurately extract character differences.

[Means and effects for solving the problem] In the present invention, as a basic form, a first character string consisting of individual elements and an end of a second character string that is a modified first character string are provided. A common subsequence is searched by sequentially scanning from the part to the center and from elements close to each other to distant elements, and within the obtained common subsequence, a length is determined by a predetermined standard. Provided is a character string difference extraction method characterized by sequentially deleting common substrings that match the two character strings from the two character strings, and extracting unmatched portions remaining as a result of the common substring deletion as differences.

In addition, in another embodiment of the present invention, a first character string consisting of individual elements and a second character string that is modified from the first character string are provided.
a document storage means for storing a character string; a document storage device that sequentially scans the first and second character strings read from the document storage device from the ends to the center, from elements close to each other to elements further away; a common subsequence search/determination means for searching for a common subsequence and determining the search results; a means for sequentially deleting common subsequences whose lengths match a predetermined criterion;
A character string difference extraction device is provided, comprising a difference extraction means for extracting a mismatched portion remaining as a result of the sequential deletion of the common portion as a difference between the first and second character strings.

In the present invention, the following means are used to simply and quickly extract the changed part, that is, the difference, when the original character string is changed by operations such as addition, deletion, correction, etc. It will be done.

In the present invention, a standard length (hereinafter referred to as a common substring) of the longest or sufficiently long partial sublength commonly present in two given character strings is searched for, and each string is divided at the found common substring. As a result, the given character string is divided into three sets, the matched set of common substrings is judged to be a "match", and the two before and after the remaining common substring are
One set will be temporarily suspended. This procedure is applied recursively to the retained sets, dividing them based on the longest common subsequence, and repeating until at least - or more elements are judged to be a match or a mismatch, that is, a difference.

Through the above process, at least - elements are judged as "-defeat" or "difference" in - times, so this procedure will eventually stop and all elements will finish the judgment and -defeat. By deleting subsequences, subsequences with only differences are extracted.

Thereby, by first associating the longest or sufficiently long substrings, it is possible to extract appropriate differences in character strings containing many identical elements.

〔Example〕

Prior to specific description of embodiments of the present invention, the principle of the present invention will be explained with reference to FIGS. 2 and 3.

FIG. 2 is an explanatory diagram for explaining the overall procedure, and FIG. 3 is an explanatory diagram showing the procedure for searching for a common subsequence.

In Figure 2, this is an example of comparing and matching the first character string "Aiueueoioi", which is made up of the same elements, and the second character string similar to it, "Aielanioaooiei", where letters are used as elements. .

The overall procedure is to search for common substrings that exist in common in both character strings, and search for the longest common substring [Elanio JP3
is adopted as the first common subsequence.

As a result, both character strings are divided, and the next character strings to be compared become two sets: the front character string "aiu" vs. "ai" and the rear character string "ii" vs. "aoiei".

Next, the longest common substring "i-JPI" is associated with the preceding character string, but since there is no common substring, the processing of the preceding character string ends.

The same procedure is applied to the backward character string, and the longest common substring "Oi JP6" is associated and this string is divided, but since there is no longer a common substring in each of the forward and backward character strings, the entire process is stops.

As a result, in the second character string, "tsu" P2 is removed from the first character string, [A JP4 is replaced with "A JP5, [A J
Obtain that P7 has been added.

Next, details of the common subsequence search in the overall procedure will be explained using FIG. In FIG. 3, the vertical axis represents the arrangement of first character strings, the horizontal axis represents the arrangement of second character strings, and the first intersection of the two represents a set of elements.

In FIG. 3, solid line arrows indicate a scanning order that provides only the starting position for searching for common subsequences, from the edge to the center and from close to remote sets; is a list of comparison sets of elements in which both the vertical and horizontal axes are advanced one by one, starting from the position given by scanning, and indicates the search order for common subsequences.

In addition, the ○ mark indicates the case where the elements are equal through the search, and the
The marks also indicate cases where the elements are not equal. In this example, the reference length for a sufficiently long common subsequence length is set in advance to be 3 characters or more.

First, scanning is started from the element set 10 at the upper left corner, which is the end. Next, a search for a common subsequence is started using this point as a starting point, and since element set 10 are equal, both elements are advanced one by one to element set 11, and it is found that these are also equal. After that, no subsequent common element can be found, so it is recognized that the length of the common subsequence at this scanning position is two characters.

In the case of the example shown in Figure 3, the standard length of the predetermined common subsequence is set to 3 characters or more, so the common subsequence discovered here, ``IJPI'', is temporarily put on hold because it does not meet the standard length. The scanning position is changed back again from the edges to the center and from the nearest set to the farthest set for the next element comparison [12
Move to.

Next, a common element 13 is found, but since the common subsequence that follows it also does not meet the standard length, the matching is suspended and scanning moves on to the next element comparison set 14.

The length of the common subsequence following the next discovered common element 15 is 4 characters, which is enough to meet the standard length.
3 is recognized as a common subsequence.

As a result, comparisons before and after the common subsequence are no longer necessary, so it is no longer necessary to scan the area of the hatching part 16, and the next scanning start position is determined by the common subsequence "Elanio JP".
Proceed to position 17, which is plus 1 from the end of 3.

By using the reference length in this way, it is possible to significantly reduce the scanning range, and it becomes a means of obtaining results quickly.

Thereafter, the remaining lower right area is scanned in a similar manner, but when the common element 1B is discovered and the subsequent (common element 19) is discovered, the first character string is exhausted and the search for the common substring ends.

As a result, the first common element that exists in common to the two character strings, the first character string and the second character string, is 10°13, 1
Four common subsequences, 5 and 18, can be quickly found.

In the example shown in Figure 3, while taking advantage of the principle of recognizing and deleting the longest common subsequence as a common subsequence, in order to carry out this process more efficiently, a certain length standard is used. Measures are taken to proceed by determining that the common subsequence extending over the length (length) is the longest, and in this case, the predetermined specific length, that is, the reference length, is determined as follows.

The length of the reference length may be at least a length that can be determined to be a match, but if it is too long, the efficiency in obtaining results will be impaired. Therefore, normally, the value may be appropriately determined to be about 10 times the elements constituting the target character string or about 1/2 of the comparison range of the character string. According to the inventor's experiments, good results were obtained when the predetermined reference length for comparing computer programs was set to about 20 to 100 lines.

The following points can be drawn from the examples in FIGS. 2 and 3.

(1) By performing matching, division, and deletion using the longest common substring existing in two given character strings, appropriate results can be obtained even when each of the character string components appears frequently.

(2) By sequentially deleting from the longest common subsequence, the range becomes solved, so the problem is reduced to the next two remaining comparison problems, and the unsolved range thereafter is reduced.

(3) Early detection of common elements is achieved by scanning for common elements from the edges to the center and from close to distant sets. This is because the occurrence of similar character strings is mainly due to the accumulation of partial additions, subtractions, and corrections to the original, and there is a high probability that a common substring exists near the end or in a close group.

(4) In addition to the means described in (2) and (3) above, it is also determined that the common subsequence is long enough by determining that it has reached a predetermined standard length, so that the results can be obtained. The scanning range, which is the comparison area, is significantly reduced. If it is determined that it is sufficiently long only by being the longest, then in order to guarantee that it is the longest, elements will be compared for combinations that are the product of the number of elements of both character strings, and the above (2 ), (3) are not taken advantage of.

An apparatus for performing a character string difference extraction method as an embodiment of the present invention is shown in FIG. In the apparatus shown in FIG. 1, a floppy disk 100a on which documents created using a word processor are recorded is used as an input medium, and differences in results are recorded on a CRT 50.
0 or output to the printer 600. In the apparatus shown in FIG. 1, a floppy manual section 100, a document storage section 200, a common subsequence storage section 201, a judgment result storage section 202, a scanning position generation section 301, a common subsequence search section 302, a search result judgment section 303, a common Subsequence deletion unit 304, difference extraction unit 30
5. CRT/printer drive unit 400, CRT 50
0 and a printer 600 are provided.

The operating procedure of the apparatus shown in FIG. 1 will be explained with reference to the flowchart shown in FIG.

Step 1 (SO to S2): First, read the first and second documents for which differences should be extracted using the floppy human resource department (SL
) is stored in the document storage unit (S2).

Step 2 (33 to S9): Next, the characters of the two read documents are used as comparison elements, and common subsequences that exist in common are used as elements obtained from the scanning position generator until it is determined that all scanning is complete (S3). The common subsequence search unit searches for the set position as a starting point (S4). Determine that the subsequence length is greater than zero (S5
), it is stored in the common subsequence search result storage unit (S
6) Do. Next, if it is determined that the subsequence length is equal to or greater than the predetermined reference length (S7), it is no longer necessary to scan the total comparison area of each element of this common subsequence.
The scanning start position is updated to a position that is plus one from the end of the common subsequence found here (S8).

Otherwise, the next scan position is obtained by the scan position generator (S
9).

Step 3 (S13 to 515): Using the common subsequences obtained by the common subsequence search, the search result determination unit divides the given two character strings into three, and determines whether these subsequences are the same or different. is determined (5IO) and stored in the determination result storage section (511). This procedure is repeated, and when all the divided subsequences have been evaluated, it is determined that the process is complete. (S12).

Step 4 (S13 to 515): Finally, the difference extracting unit extracts the difference from the difference subsequence stored in the judgment result storage unit using three expressions (S13): adding, deleting, and replacing the difference on the CRT.
Output to the printer drive unit (S14) L completes the entire process (515).

A specific example of the arithmetic processing shown in FIG. 3 will be shown using the above procedure.

First, according to step 1, a first document is stored as a first character string and a second document is stored as a second character string in the document storage unit. As a result, the first character string is stored as shown in Table 1, and the second character string is stored as shown in Table 2.

Next, in step 2, a substring that exists in common between the two character strings is searched for and stored in the common substring storage section. The results are shown in Table 3.

Step 3 is to divide the two given character strings using the common substring obtained in step 2, repeat the judgment of match or difference for the divided substrings, and store the divided parts in the judgment result storage section. Stores columns and their judgment results. This division process is sequentially shown in Tables 4 to 9.

Initially, the beginning to end of both vehicle character strings are stored as partial string A of the determination result of the pending state as shown in Table 4.

Table 4 Next, among the common subsequences searched in step 2, item No. 3 of Table 3, which is the longest, is used to divide subsequence A. As a result,
Subsequence A is divided into three subsequences B, C, and D. C is determined to be a match, and the preceding and succeeding pairs B and D are determined to be reserved. The division results are shown in Table 5.

Next, we focus on the reserved subsequence B and search Table 3 for common subsequences within the range of this subsequence.The common subsequence that satisfies the condition is item number 1, so this is adopted. Then, subsequence B is divided into three sets E, F, and G shown in Table 6. Here, since subsequence C was determined to be a match, it has been deleted from Table 6.

Table 6 Here, substring E is deleted because both character strings are empty sets (0).

In addition, substring G can no longer be divided because the second string is an empty set (0), and the second string contains the first string element position i = 3, which is deleted because JP2 does not exist. As a result, the partial strings G and D shown in Table 7 remain in the determination result storage section.

Table 7 Again, the subsequences for which the judgment results are pending are divided using the common subsequences, but in this case, only one subsequence name is held in the judgment result storage section. Similarly, common subsequences included within the range of the subsequences are searched from the common subsequence storage unit and Table 3. In this case, item number 4 is adopted and D is divided into three parts, H, 1, and J shown in Table 8.

Table 8 Here, since the first string side of substring J is an empty set (0), it can no longer be divided, it does not exist in the first string, and the second string element position 1 = 10 ~ 11, “Ei JP7
Since there is, it is determined that the difference is an added difference. Next, subsequence ■ is deleted from the judgment result storage unit because it matches, and as a result of the search, there is no common subsequence within the range of subsequence H, so subsequence H cannot be judged for difference. receive. The results are shown in Table 9.

Table 9 Here, since there are no subsequences for which the determination result is in a pending state, the difference between subsequences G, H, and J is extracted, and the difference extraction process is completed.

Finally, in step 4, the difference extraction section outputs each difference from the partial strings G, H, and J of differences existing in the determination result storage section to the CRT screen or printer via the CRT/printer drive section.

In the substring G, since the second character string is the empty set (0) and the character [JP2 exists at the first character string element position i=3,
"" is displayed as deleted.

In the substring H, since the characters exist in both character strings, the character r・i, P4'' at the first character string element position i=8 is replaced with the character ``AJP5'' at the second character string element position j=7. is displayed.

Substring J displays that the character "Ei" has been added because the first character string is an empty set (0) and the character "Ei JP7" exists in the second character string element position j = 10 to 11. be able to.

〔Effect of the invention〕

According to the present invention, it is possible to mechanically understand the addition/subtraction correction portion of a document or computer program written in everyday language by comparing the addition/subtraction correction portion rather than its progress. According to the present invention, it is possible to greatly improve the efficiency of document and computer program processing and maintenance work, which in the past could only be done manually and with low efficiency. Further, according to the present invention, it is possible to perform comparison work that requires a great deal of manual labor with a large number of character strings with high efficiency, and character string differences can be extracted quickly and accurately.

[Brief Description of the Drawings] Fig. 1 is a block diagram of an apparatus for performing a character string difference extraction method which is an embodiment of the present invention; Fig. 2 is an explanatory diagram for explaining the present invention in detail; 3
The figure is an explanatory diagram of the common subsequence search procedure, and FIG. 4 is a flowchart for explaining the operation of the apparatus of FIG. 1. DESCRIPTION OF SYMBOLS 100... Floppy human resource department, 100a... Floppy disk, 200... Document storage section, 201... Common subsequence search result storage section, 202...
Judgment result storage section, 301... Scanning position generation section, 302... Common subsequence search section, 303... Search result determination section, 304... Common subsequence deletion section, 305... Difference extraction section , 400...CRT/printer drive unit, 500...
C.R.T. 600...Printer.

Claims (1)

[Claims] 1. A first character string consisting of individual elements, and a second character string that has been modified from the first character string, from the end to the center and located close to each other. A common subsequence is searched by sequentially scanning from element to remote element, and among the obtained common subsequences, a common subsequence whose length matches a predetermined criterion is extracted from the two character strings. A character string difference extraction method characterized by sequentially deleting common substrings and extracting unmatched portions remaining as a result of the common substring deletion as differences. 2. The method according to claim 1, wherein the predetermined criterion is a common substring that has the largest number of characters among the common substrings or a common substring that has reached a predetermined number of characters. 3. Document storage means for storing a first character string consisting of individual elements and a second character string obtained by modifying the first character string; Common subsequence search/determination means for searching for a common subsequence and determining the search results by sequentially scanning from the end of the character string of No. 2 to the center, from elements close to each other to elements further away; means for sequentially deleting common substrings whose length matches a predetermined criterion; and difference extraction for extracting a mismatched portion remaining as a result of the sequential deletion of the common portions as a difference between the first and second character strings. A character string difference extraction device comprising: means.