JP3414319B2 - Data retrieval apparatus, method and recording medium - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data retrieval device, method and recording medium for retrieving data by checking whether they match in meaning.

[0002]

2. Description of the Related Art A conventional character string replacement system acquires the contents of a text such as a source program and performs collation processing for checking whether or not there is a character string that matches a specified character string in order from the beginning of the text. When there is a matching character string, the process of replacing the character string with the specified character string is repeated until the end of the text.

However, in the conventional system, the collation process of the designated character string and the character string to be replaced is performed by "word".
Since the comparison is simply performed between the character strings without being aware of the concept of, there is a problem that even a character string that should not be originally replaced may be replaced. For example, if you try to replace "oyster" with "kuri",
"Oyster" in the character string is also replaced.

Therefore, Japanese Patent Laid-Open No. 3-150677 proposes a character string search method in which a program is divided into words (words, tokens, tokens) and the divided words are compared as a unit with a search character string. ing. When this technique is applied, when trying to replace “oyster” with “kuri”, the “oyster” in the character string of “oyster” will not be replaced.

[0005]

However, even in the technique described in the above publication, a character string that is semantically identical to a specified character string but does not match the expression method is not subject to replacement. There is. For example, assume that a hierarchical relationship in which "ta" exists above "u" is represented by "u OFta", "na" exists above "u", and "ta" exists above "na". The existing hierarchical relationships are
It is represented by F ta.

At this time, "U OF" and "U OF"
“OF” is related to whether there is a difference in the syntactic description.
However, they can be recognized as the same object, the meaning is the same, and they may be replaced. However,
In conventional systems (including those described in the above publication),
Since there is no part that matches these expressions, it cannot be handled as the character string to be replaced. in this way,
There is a problem in that even a character string that is originally to be replaced is not replaced.

The present invention has been made in order to solve the above-mentioned problems of the prior art, and is to search for data that is semantically identical to desired data, such as data to be replaced with other data. It is an object of the present invention to provide a data retrieval device, a method, and a recording medium recording a program for executing the method.

[0008]

To achieve the above object, according to an aspect of the data retrieval apparatus according to the first aspect of the present invention, floor
First parsing means for parsing the first character string in which the layer relationship is clearly described according to a predetermined rule, based on the predetermined rule, and a second character string to be searched,
Recognizing that the analysis result by the second analysis unit is the same as the target in the analysis result by the second analysis unit and the first analysis unit that performs syntax analysis based on the predetermined rule.
A first collating means for collating whether or not there is a possible semantically matching portion, and the first character string corresponding to the semantically matching portion in the matching by the first collating means First to extract the character string of as the searched character string
And an extracting means of the above.

In the above data retrieval device, the first collating means determines whether or not there is a match between the first character string and the second character string based on the result of syntax analysis, not as a simple expression of the character string. Is collating. This enables the first extracting unit to extract a character string having the same meaning even if the expression of the character string is different, as the searched character string.

The above-mentioned data retrieving apparatus uses the first instructing means for instructing a third character string in which the second character string is to be replaced, and the character string extracted by the first extracting means, It may further include first replacing means for replacing the third character string instructed by the first instructing means.

Here, the first extracting means is the first extracting means.
The position of the character string in the first character string that corresponds to the portion that is semantically matched by the matching unit may be notified to the first replacing unit. In this case, the first replacement unit may replace with the third character string based on the position of the character string notified from the first extraction unit.

Further, the first analysis means is the first
The analysis result of the character string may be associated with information for performing replacement with the third character string instructed by the instructing means. In this case, the first replacement unit may replace the third character string based on the information associated with the analysis result of the first character string.

In the above data retrieval device, the first,
The second analysis means can divide the first and second character strings into words, respectively, and associate the divided words with a predetermined data structure to obtain respective analysis results. In this case, the first collating means compares the analysis results of the first and second analyzing means in terms of words included in each, and collates whether there is a semantically matching portion. You can

Here, the first and second character strings may include a plurality of words described in a hierarchical structure.
In this case, the first and second analysis means sequentially connect a plurality of words described in a hierarchical structure from the words on the upper side or the lower side to obtain respective analysis results, and the first matching means Can check the results of analysis by the first and second analyzing means while tracing each list to see if there is a semantically matching portion.

Further, the first analysis means stores the highest or lowest word of the hierarchical structure in a predetermined storage position according to a hash value obtained by the hash method, and the first analysis means The matching unit obtains a hash value from the highest or lowest word of the words described in the hierarchical structure in the second character string, and stores the first character string at the position of the hash value. It is possible to check whether there is a semantically matching part by checking the list in order from the word inside.

In the above data retrieval device, the first character string may be described in a predetermined programming language. In this case, the first and second parsing means may parse the first and second character strings based on the specifications of the predetermined programming language.

Further, when the first character string is a data definition in the predetermined programming language,
The data retrieval device includes an operand analysis unit that divides a fourth character string, which is a processing description in the predetermined programming language, into operands, and a third analysis unit that parses the operands divided by the operand analysis unit. And second collation means for collating whether or not the operand analysis result by the third analysis means has a part that semantically matches the analysis result by the first analysis means.
And a second extracting unit for extracting a character string in the fourth character string corresponding to a portion that is semantically matched by the matching by the first matching unit as a searched character string. You can

[0018] Here, the third analysis means can target the operands other than the reserved words among the operands divided by the operand analysis means as the target of syntax analysis.

As described above, in the further analysis of the fourth character string, the second instructing means for instructing the fifth character string to replace the fourth character string, and the second instructing means. The character string extracted by the extracting unit may be further replaced with a second replacing unit that replaces the fifth character string instructed by the second instructing unit.

In order to achieve the above object, in the data search method according to the second aspect of the present invention, the first character string in which a hierarchical relationship is clearly described and described according to a predetermined rule is used as the predetermined character string. A first parsing step for parsing based on the rule, and a second parsing step for parsing the second character string to be searched based on the predetermined rule; and the first parsing step. It is possible to recognize that the target is the same as the analysis result in the second analysis step in the analysis result
A matching step for matching whether there are semantically matching part a portion, the character string in the first character string corresponding to semantically matched part matching in the matching step, retrieved And an extraction step of extracting as a character string.

In the data retrieval method, the first character string extracted in the instructing step of instructing a third character string in which the second character string is to be replaced and the first character string extracted in the extracting step is used.
The replacement step of replacing with the third character string designated in the designation step may be further included.

In order to achieve the above object, a computer-readable recording medium according to the third aspect of the present invention comprises:
A first parsing step of parsing a first character string in which a hierarchical relationship is clearly described according to a predetermined rule in the computer based on the predetermined rule, and a second character string to be searched are provided. , It is possible to recognize that the analysis result of the second analysis step is the same as the target in the analysis results of the second analysis step and the first analysis step, which are syntax-analyzed based on the predetermined rule. a matching step for matching whether there is meaning <br/> taste congruent portions a moiety are characters in the first character string corresponding to semantically matched part matching in the matching step And a program for executing an extraction step of extracting a string as a searched character string.

The computer-readable recording medium further comprises an instruction step for instructing a computer of a third character string to be replaced with the second character string, and the character string extracted in the extracting step, A program for further executing the substitution step of substituting the third character string designated in the first designation step may be recorded.

[0024]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the accompanying drawings.

[First Embodiment] FIG. 1 is a functional block diagram showing a data replacement system according to this embodiment. As shown in the figure, the data replacement system is provided with a source program 1 including a character string to be replaced, and a parameter 3 including a character string to be replaced and a character string after the replacement. The system also includes a character string conversion function 2, which replaces the character string in the source program 1 in accordance with the given parameter 3.

The source program 1 is described in accordance with a predetermined programming language (high-level language), and includes a data definition section 11 and a processing description section 12. Parameter 3 specifies the character string to be replaced (character string to be replaced) and the character string after replacement (character string to be replaced).

The character string replacing function 2 is a data definition dictionary 2.
1, a data analysis unit 22, a data collation unit 23, an operand analysis unit 24, a character string replacement unit 25, and a data / process collation unit 26. In the character string replacement function 2, the processing is controlled so as to be transferred in the order of the data analysis unit 22 → data collation unit 23 → operand analysis unit 24 → character string replacement unit 25.

In the data definition dictionary 21, one information item is registered for each defined data item. According to the hierarchical structure of the data, the dictionary is constructed so as to be registered with the hierarchical structure. For example, even if the data named "A" is defined multiple times, it is expressed in the dictionary that the data is different when the upper data is "KA" and "SA". To be done.

The data analysis unit 22 uses the source program 1
The data definition unit 11 therein is parsed, and the definition content of the data is registered in the data definition dictionary 21 as a hierarchical structure of the relationship between words (lexical characters) (this is referred to as construction of the data definition dictionary 21).

The data collating unit 23 uses the data definition dictionary 2
After the construction of 1, the character string to be replaced (replaced character string) given as the parameter 3 is acquired, and this is parsed. The data collating unit 23 compares the syntactic analysis result of the replaced character string with the syntactic analysis result registered in the data definition dictionary 21, and determines whether the syntactic description is the same or not.
It is checked whether or not there is a definition of data that can be recognized as a target and is semantically consistent. As a result of the above collation, the data collating unit 23 sets, in the data definition dictionary 21, information indicating that the data is semantically identical to the character string to be replaced, which is the data to be replaced.

After matching the replaced character strings, the operand analysis unit 24 lexically analyzes the process description unit 12 in the source program 1 and divides the content of the process description into syntactically independent units (operands).

The data / process collation unit 26 receives the operands divided by the operand analysis unit 24, and the operand which is not a reserved word in the programming language of the source program 1 matches the data name existing in the data definition dictionary 21. Check whether to do.

The character string replacing section 25 is a processing description section of the source program 1 with the replaced character string given as the parameter 3 for the operand successfully collated in the collating process performed by the data / process collating section 26. Replace 12. Further, the character string replacing unit 25 assigns, as the parameter 3, data that has been successfully collated with the replaced character string (the character string to be replaced) given as the parameter 3 in the collation performed by the data collating unit 23. Replace with the replaced character string. That is, the data successfully collated in the data definition unit 11 of the source program 1 is replaced with the replaced character string.

Next, the hardware configuration for realizing the above-described system functions will be described with reference to the block diagram of FIG. As shown, this hardware includes a secondary storage device 51, a main memory 52, and an input unit 53.
, An output unit 54, and a central processing unit (hereinafter referred to as CPU) 55, and these components are connected to each other via a bus 56.

The secondary storage device 51 is composed of, for example, a hard disk and the like, and is a program for executing the above-mentioned source program 1 and the above character string replacing function 2,
A dictionary 21 of data definition of file format is stored.

The main memory 52 is composed of, for example, a RAM or the like, and has storage areas 52A and 52B for storing a source program 1 read from the secondary storage device 51 and a program for executing the character string replacement function 2.
A work area 52C for temporarily storing data required in the replacement process, but are assignment have. The data definition dictionary 21 is created in the work area 52C and stored in the secondary storage device 51 as needed. Further, the parameter 3 and the result of the parsing are also temporarily stored in the work area 52C.

The input unit 53 is composed of a keyboard, a mouse, etc., and specifies the parameter 3, for example. The output unit 54 is composed of a display device, and displays, for example, the source program 1, the designated parameter 3, and the source program 1 after the replacement process.

The CPU55 via the bus 5 6 controls the above components, loading a program from the secondary storage device 51 into main memory 52 by executing the program, the string replacement function 2 Achieve each included function.

The operation of each part of the data replacement system according to this embodiment will be described below according to the processing order. Here, the programming language describing the source program 1 is described Tsu preparative an example where a COBOL.

In COBOL, the "DATA DIV
ISION. ”And“ PROCEDURE DIVISION. ”
Since the data is defined between the part described with, the description between this part is defined as the data definition part 11. On the other hand, "PROCED
Since the processing is described after the portion described as “URE DIVISION.”, The description after this is referred to as the processing description section 12.

The data definition and the process description in COBOL will be described by taking "a.DELTA.OF.DELTA.sa." As an example. Here, Δ means blank. This Δ is shown for convenience of explanation, and is not actually used in the description by COBOL. The meaning of this description is as follows.

(1) "Blank" is treated as a separator. (2) "OF" means a word representing a hierarchical relationship. In the above description example, a hierarchical relationship in which "sa" exists above "a" is expressed. (3) The period (.) Means the end of the instruction. In the above description example, the operand will be determined by "a OF sa".

Next, regarding the processing of the data analysis unit 22,
This will be described with reference to FIG. The data analysis unit 22 uses COB
According to the specifications of the OL, the data definition part 11 which is the part where the data name in the source program 1 is defined is targeted for processing. The data analysis unit 22 registers the data defined by writing a character string in the data definition unit 11 in the data definition dictionary 21 on the memory so that it can be easily applied to the subsequent matching process.

Here, the registration format that is easily applied to the subsequent collation processing is (1) so that the hierarchical structure of data can be traced.
Have a dependent data and a pointer that can refer to the dependent data mutually, and (2) Obtain a hash value from the data name using the hash method so that the data name can be searched quickly. Has a pointer that can continuously reference the same data,
It is a format with features such as.

For example, the case where the data analysis unit 22 analyzes the character string (description) described in the data definition unit 11 in FIG. 3 will be described as an example. "01? // 02 a.
The data analysis unit 22 obtains an analysis result that the word “a” exists in a higher order of the word “a” by the syntax analysis of the character string (description) “(// represents a line break)”. .

Then, the data analysis unit 22 adds a pointer indicating a hierarchical relationship to each of the data "a" and "ka", and the data name ("a") is added to the "a" data by the hash method. A pointer corresponding to the obtained hash value is added.

Similarly, when the second and second character strings (descriptions) are also analyzed, the hierarchical structure information as shown in the data definition dictionary 21 in FIG. 3 is obtained.

Next, regarding the processing of the data collating section 23,
This will be described with reference to FIG. The data matching unit 23 uses the replaced character string given to the parameter 3 (the character string to be replaced).
Determines which data definition corresponds in the data definition dictionary 21.

The data collating unit 23 determines the contents of the replaced character string in the language of the source program 1 (COBO in this case).
L) The syntax is parsed according to the specification, the hash value of the first word divided as the analysis result is obtained, and the data definition dictionary 2
1, the definition of data having a name tag (name) represented by the hash value is scanned one after another, and the data of the data name that matches the first word is provisionally determined. After that, the data matching unit 2
If the description content of the replaced character string has an expression of a hierarchical structure, 3 performs collation processing as to whether this hierarchical structure and the hierarchical structure in the data definition dictionary 21 match.

For example, if the character string to be replaced is "a OF sa", refer to the hierarchical structure in the data definition dictionary 21 to determine whether "sa" exists as higher data of "a". The pointer is traced upward from the data name indicating "A", and the comparison is performed by the data name in each layer and the data name indicating "A". And “a”
If a data name indicating “sa” exists above the data name indicating “”, it is determined that the data name is the replacement target.

In FIG. 4, the symbol "x" is the "a" of "a OFsa" from the parameter 3 and the dictionary 2 of the data definition.
The "a" that exists below "ka" in 1 was first matched, but the character string expressed in this hierarchical structure was "a OF.
"A" of "a OF sa" from parameter 3 and "sa" in the dictionary 21 of the data definition. As a result of next matching with the lower "a", it is shown that the character string represented by this hierarchical structure has the same meaning as "a OF sa". Note that the symbols “x” and “◯” described in the figures described later have the same meaning.

The data collating unit 23 sets, in the data definition dictionary 21, information indicating that it is a replacement target for the data (data name) determined to be a replacement target as described above.

In the example shown in FIG. 4, the symbol “⊚” is added to the confirmed data name (“a”), but this symbol “⊚” indicates information indicating that the data is to be replaced. I mean. Note that the symbol “⊚” described in the drawings described later has the same meaning.

The data collating unit 23 further notifies the character string substituting unit 25 of the start position and end position of the character string in the data defining unit 11 which is semantically matched, and the data defining unit 25 Instruct replacement processing of the corresponding data definition.

Next, the processing of the operand analysis unit 24 will be described with reference to FIG. Operand analysis unit 24
Processes the process description part 12 in the source program 1 where the process description is defined according to the COBOL language. The operand analysis unit 24 performs the following process.

(1) A "word" is separated and determined from the process description described in the process description section 12 as a character string based on the separator in the language specification. (2) When the separated "word" is a user word that is not a reserved word in the language specification, the word "OF" indicating the hierarchical relationship in the COBOL language is used as a clue, and the hierarchy of the user word in the process description According to the relationship, the range of the character string is fixed as the "operand". (3) The data / process collation unit 26 is instructed to perform the process of determining whether or not the target is the replacement target in the unit of the determined “operand”.

Here, in the processing description, "MOVE △ 100 △ TO △
Oh △ OF △ If a character string indicated by "(△ is a blank character) is described,"
OF △ The analysis for the description of "is described.

The operand analysis unit 24 first searches for a separator. In the COBOL language, "blank" is treated as a separator, so that "a" is first determined as a "word" (a token). Since there is "OF" after it, it is judged that the word "sa" further expresses the hierarchical relationship. Then, there is a period (.) After "sa", which means the end of the instruction in COBOL, so that the operand is determined within the range of "aΔOFΔsa".

The operand analysis unit 24, which has determined the operand in this manner, uses the operand “AΔOFΔSA”.
Is passed to the data / process collation unit 26, and it is determined whether or not it is a replacement target.

Next, the processing of the data / processing collating section 26 will be described with reference to FIG. Data / process collation unit 2
6 divides the operand transferred from the operand analyzer 2 4 the word, the data comparison unit 23 is similarly parsed and the replaced character string given as the parameter 3. Next, the hash value of the first word of the operand is obtained, the definitions of the data having the name tag represented by the hash value are sequentially scanned on the data definition dictionary 21, and the data name that matches the first word The data of is temporarily confirmed.

For example, when the first word of the operand is "A", pointers that refer to the data of the same hash value as "A" are traced one after another, and the collation process with the data name is performed to obtain "A". The data having the data name of is temporarily determined.

Thereafter, if the description content of the operand has a hierarchical structure expression, the data / process matching unit 26 determines whether this hierarchical structure matches the hierarchical structure in the data definition dictionary 21. Perform matching processing.

For example, when the operand is “A OF SA”, whether or not “SA” exists as higher data of “A” is determined by referring to the pointer for referencing the hierarchical structure in the data definition dictionary 21. The data name indicating “a” is traced to the upper level, and the comparison is performed by comparing the data name in each layer with the data name indicating “a”. Then, when a data name indicating "sa" exists above the data name indicating "a", it is determined that the data name is a replacement target.

Thus, the character string indicating the operand is
By collating the entire character string with the data definition dictionary 21 for each "word" without collating at once, for example, "a △ OF △ SA"
In the case where the number of blank characters between “A” and “OF” is different depending on the description, like the two descriptions of “A” and “A △△ OF △ SA” (△ is a blank character). Also, as a result of the syntactical check, it is possible to determine that the descriptions have the same data name.

In addition, "U OFta" (2 levels) and "U OF
Even if the hierarchical structure is specified differently, such as "OF" (3 layers), it is possible to determine that the description indicates the same data name as a result of the semantic check. .

In the case where information that is a replacement target is set in the matching processing by the data matching unit 23 for the data determined to be a replacement target (in the example of FIG. 6, the symbol “⊚”), , That operand becomes the replacement target (replaced character string).

In this case, the data / process collation unit 26 designates the start position and the end position of the operand in the source program 1 to the character string replacement unit 25, and the corresponding process description in the process description 12 is specified. To replace the.

Next, regarding the processing of the character string replacing section 25,
This will be described with reference to FIG. Both the data collation unit 23 and the data / process collation unit 26 are instructed to the character string replacement unit 25 to perform the character string replacement. The start position, the end position, and the replacement character string are notified in the source program 1.

The character string substituting unit 25 compares the instruction contents of the character string substituting with the data collating unit 23 and the data / process collating unit 26.
After the end of the process, the start positions are arranged in descending order. Then, the character string replacing unit 25 replaces the character string in the source program 1 indicated by the start position and the end position with the character string of the replacement word given as the parameter 3.

The reason for arranging the replaced character strings in descending order of the start positions is as follows. When replacing a character string, if the length of the character string after replacement is different from the length of the character string before replacement, the description after the character string after replacement is relatively displaced (shifted backward). ) It will be. Therefore, by replacing from the back of the source program 1, the replacement target position is sequentially moved to the front, so that it is not necessary to be aware of the relative movement of the description during the subsequent replacement processing. is there. Therefore, the process description section 12 is replaced first, and then the data definition section 11 is replaced.

Note that, in FIG. 7, in the processing description section 12, "A OF
It is shown that “S” is replaced with “A OF S”, and in the data definition section 11, “5 lines and 15 digits ...
It is shown that “a” in “5 rows and 16 digits” is replaced with “a”.

As described above, according to this embodiment, it is determined after confirming semantically that the character string in the source program 1 is suitable as a replacement target. It is possible to accurately perform character string replacement in units of data names.

Further, as described above, the character strings to be replaced can be narrowed down accurately. Therefore, if the character string to be replaced is erroneously replaced after the sentence replacement process is started, the character string to be replaced is replaced. It is no longer necessary to intervene by the manual work of returning to the original character string. Therefore, even when performing character string replacement for a plurality of source programs 1, the operator only needs to specify the parameter 3 at the beginning, and it is possible to perform it all at once with a very small amount of work.

[Second Embodiment] FIG. 8 is a functional block diagram showing a configuration of a character string search device according to a second embodiment. The character string of the replacement word referred to by the character string replacement unit 25 is directly passed as the parameter 3. However, in this embodiment, the information about the character string of the replacement word is temporarily stored in the data definition dictionary 21. , Is different from the first embodiment in that it is given as one of the replacement instructions to the character string replacement unit 25. That is, the processes of the data collating unit 23 and the character string replacing unit 25 are slightly different from the processes in the first embodiment. Therefore, these components will be described.

The part in which the data collating unit 23 determines which data definition in the data definition dictionary 21 the replaced character string given to the parameter 3 corresponds to will be described with reference to FIG. It is basically the same as the processing performed. However, the processing of the data matching unit 23 is different from that of the first embodiment in the following two points.

In the example shown in FIG. 4, the information setting on the dictionary 21 of the data definition for the decided data is only information indicating that it is the replacement target (information indicated by the symbol "⊚" in FIG. 4). However, in this embodiment, in addition to this information, the data collating unit 23 indicates in the data definition dictionary 21, what the character string of the replacement word is, as shown in FIG. The information 21A is set. This information 21A is information indicating "a" when the character string of the replacement word is, for example, "a OFa".

The contents specified by the data collating unit 23 to the character string replacing unit 25 are the character strings to be replaced (for example, "A O
Only the information of the start position and end position of F
In this embodiment, the data collating unit 23 notifies the character string replacing unit 25 of the replacement character string (for example, “A” of “A OF SA”) as the instruction content in addition to these pieces of information.

The character string replacing unit 25 is different from the case described with reference to FIG. 7 in the following one point. In the example shown in FIG. 7, the replacement character string is directly obtained from the parameter 3, but in this embodiment, the character string replacement unit 25 temporarily stores the data from the parameter 3 as shown in FIG. Information about the character string after replacement stored in the definition dictionary 21 (for example, "a" in the case of "a OFsa")
Is given for each replacement instruction.

According to this embodiment, since the information on the character string to be replaced and the information on the character string after replacement are stored as a pair in the data definition dictionary 21, the contents of different parameters 3 are stored. (Character string before replacement and character string after replacement) can be specified simultaneously. In addition, also in this embodiment, it is possible to obtain the same effect as that of the first embodiment.

[Third Embodiment] In the above first and second embodiments, the case where the data defined in the hierarchical structure is replaced has been described as an example. On the other hand, in this embodiment, a case of replacing a mathematical expression will be described as an example. Here, a program including the mathematical expressions (1) to (3) as shown in FIG. 11 is taken as an example.

When a processing device such as a compiler that executes a predetermined program reads the mathematical expressions (1) to (3), the syntax analysis is performed.
A syntax tree as shown in (c) is obtained. In the syntax tree of Expression (1), as shown in FIG. 12A, the uppermost parent node is “=”, the left child node is “a”, and the right child node is a subtree. Has 101. Therefore, when a subtree semantically the same as the subtree 101 appears in the mathematical expression (1) and thereafter, it means that the subtree can be replaced with “a”, and the processing device
To replace the syntax tree that is semantically matched with "a".

In the subtree 101, the parent node is "+", the left child node is "b", and the right child node is "c", but the parent node is "+". Therefore, the left and right child nodes have no order, and have “b” and “c” as child nodes.

The syntax tree of equation (2) is shown in FIG.
As shown in (b), the subtree 102 is included. Since the subtree 102 completely matches the subtree 101, this can be replaced with “a”. Therefore, equation (2)
Is replaced by "d = e + a" by the processor.

On the other hand, the syntax tree of equation (3) is shown in FIG.
As shown in (c), the subtree 103 is included. In the subtree 103, since the parent node is “+”, the left and right child nodes have no order and have “b” and “c” as child nodes. Therefore, the subtree 10
3 becomes a semantic match with the subtree 101, and the expression (3) is replaced with “f = g + a” by the processing device.

As a comparative example, when the same substitution is performed according to the conventional technique, the expression (2) includes “b + c” included in the expression (1) in that order,
It can be replaced with “d = e + a”. However, even if “c + b” included in Expression (3) has the same meaning as “b + c” included in Expression (1), they cannot be replaced because they do not match as the expression of the character string.

Therefore, in the case of executing the program after the replacement shown in FIG. 11 is executed, the number of steps of the addition operation is only three in this embodiment, whereas in the conventional example. Four steps are required. That is, according to this embodiment, it becomes possible to optimize the program in a more efficient form, and it is possible to improve the execution speed of the program.

[Modifications of Embodiments] The present invention is not limited to the above-described first to third embodiments, and various modifications and applications are possible. Hereinafter, modifications of the above-described embodiment applicable to the present invention will be described.

In the first and second embodiments described above, CO
The case where the character string defining the hierarchical structure included in the program described in BOL is replaced with another character string has been described as an example. In the above-described third embodiment, if some of the mathematical expressions included in the program written in a predetermined programming language require a calculation result in advance, then some of the mathematical expressions are The case where the calculation result is replaced with the variable to be substituted has been described as an example. However, the programming language applicable to the present invention is not limited to COBOL and may be another high-level language. Further, the character string in the sentence written in natural language and the specified character string may be subjected to morphological analysis, and the results of the morphological analysis may be compared with each other to determine the character string in the sentence as a replacement target. .

In the above first to third embodiments, the case where the character string determined to be replaced from the specified character string is replaced with another character string has been described as an example. However,
INDUSTRIAL APPLICABILITY The present invention can be applied to all cases in which a character string that semantically matches a specified character string is searched, even if the character string is not replaced.

The program for executing the processing described in the above-mentioned first to third embodiments (the program realizing the character string replacing function 2) is stored in a computer-readable recording medium such as a CD-ROM. It is also possible to distribute it.

[0091]

As described above, according to the present invention,
It is possible to search for character strings having the same meaning even if the expressions of the character strings are different, for example, as a character string to be replaced with another character string.

[Brief description of drawings]

FIG. 1 is a functional block diagram showing a configuration of a data replacement device system according to a first embodiment of the present invention.

FIG. 2 is a block diagram showing a configuration of hardware that realizes the data replacement system shown in FIG.

FIG. 3 is a diagram for explaining a process of a data analysis unit.

FIG. 4 is a diagram for explaining a process of a data matching unit.

FIG. 5 is a diagram for explaining a process of an operand analysis unit.

FIG. 6 is a diagram for explaining a process of a data / process matching unit.

FIG. 7 is a diagram for explaining a process of a character string replacing unit.

FIG. 8 is a functional block diagram showing a configuration of a data replacement system according to a second embodiment of the present invention.

9 is a diagram for explaining the process of the data matching unit shown in FIG.

FIG. 10 is a diagram for explaining the process of the character string replacing unit shown in FIG.

FIG. 11 is an example of mathematical expressions in a program applied in the third embodiment of the present invention.

12A to 12C are diagrams each showing a syntax tree obtained by parsing the mathematical expression of FIG. 11;

[Explanation of symbols]

2 Character replacement function 21 Data definition dictionary 22 Data Analysis Department 23 Data collator 24 Operand analysis unit 25 Character string replacement part 26 Data / Process Collation Unit 51 secondary storage 52 main memory 53 Input section 54 Output section 55 Central Processing Unit (CPU) 56 bus

Claims (15)

(57) [Claims] 1. A first analysis means for parsing a first character string in which a hierarchical relationship is clearly described according to a predetermined rule based on the predetermined rule, and a second character to be searched. It is possible to recognize that the analysis result by the second analysis means is the same as the target in the analysis result by the second analysis means for parsing the string based on the predetermined rule and the analysis result by the first analysis means. Na
First comparing means and said first character in the first character string corresponding to semantically matched part verification by verification means for checking whether there is a portion to semantically match a partial A data retrieving apparatus comprising: a first extracting unit that extracts a string as a searched character string. 2. A first instructing means for instructing a third character string in which the second character string is to be replaced, and the character string extracted by the first extracting means, The data retrieving apparatus according to claim 1, further comprising: a first replacing unit that replaces the third character string designated by the unit. 3. The first extracting means includes a position of a character string in the first character string corresponding to a semantically matching portion in the matching by the first matching means, and the first replacing means. And the first replacement unit performs the replacement with the third character string based on the position of the character string notified from the first extraction unit. Data retrieval device described. 4. The first analyzing means associates the analysis result of the first character string with information for performing replacement with the third character string instructed by the instructing means, The said 1st replacement | exchange means replaces with the said 3rd character string based on the information matched with the analysis result of the said 1st character string, It is characterized by the above-mentioned. Data retrieval device described. 5. The first and second analysis means divide the first and second character strings into words, respectively, and associate the divided words with a predetermined data structure to obtain respective analysis results. The first collating means compares the analysis results by the first and second analyzing means in terms of words included in each, and collates whether or not there is a semantically matching portion. The data search device according to any one of claims 1 to 4. 6. The first and second character strings include a plurality of words described in a hierarchical structure, and the first and second analysis means include a plurality of words described in a hierarchical structure, The upper side or the lower side words are sequentially connected in a list to form respective analysis results, and the first collating means traces the analysis results by the first and second analyzing means, respectively, 6. The data search device according to claim 5, wherein whether or not there is a semantically matching portion is checked. 7. The first analysis means stores the highest or lowest word of the hierarchical structure in a predetermined storage position according to a hash value obtained by a hash method, and the first collation means. Is the hash value from the highest or lowest word of the words described by the hierarchical structure in the second character string, and in the first character string stored at the position of the hash value. 7. The data search device according to claim 6, wherein whether or not there is a semantically matching portion is checked while tracing the list in order from the word. 8. The first character string is described in a predetermined programming language, and the first and second analysis means are based on specifications of the predetermined programming language. 8. The data search device according to claim 1, wherein each of the two character strings is parsed. 9. The first character string is a data definition in the predetermined programming language, and an operand analysis unit that divides a fourth character string, which is a process description in the predetermined programming language, into operands. Third parsing means for parsing the operands divided by the operand parsing means, and a portion semantically matching the parsing result of the operand by the third parsing means with the parsing result of the first parsing means. A second collating means for collating whether or not there is a second collating means for extracting a character string in the fourth character string corresponding to a portion semantically matched by the collation by the first collating means as a searched character string; 9. The data search device according to claim 8, further comprising two extracting means. 10. The data according to claim 9, wherein the third analysis means targets operands other than reserved words among the operands divided by the operand analysis means as objects of syntax analysis. Search device. 11. A fifth to-be-replaced fourth character string.
Second instructing means for instructing the character string of, and a second replacing means for replacing the character string extracted by the second extracting means with the fifth character string instructed by the second instructing means. The data search device according to claim 9 or 10, further comprising: 12. A first parsing step of parsing a first character string in which a hierarchical relationship is clearly described according to a predetermined rule based on the predetermined rule, and a second character to be searched. A second analysis step of parsing a string based on the predetermined rule; and an analysis result of the first analysis step, and an object of analysis is the same as the analysis result of the second analysis step. Recognizable
Collating step for collating whether there is a semantically matching portion that is an effective portion, and searching for a character string in the first character string corresponding to the semantically matching portion in the matching in the collating step. And a extracting step of extracting the extracted data as a character string. 13. A third to-be-replaced second character string.
Further comprising an instructing step of instructing the character string of No. 1 and a substituting step of substituting the character string extracted in the extracting step with the third character string instructed in the first instructing step. The data search method according to claim 12. 14. A first parsing step of parsing a first character string having a hierarchical relationship clearly described in a computer according to a predetermined rule, based on the predetermined rule, and a first object to be searched. The second analysis step of parsing the second character string based on the predetermined rule and the analysis result of the first analysis step have the same object as the analysis result of the second analysis step. Recognizable as
Collating step for collating whether there is a semantically matching portion that is an effective portion, and searching for a character string in the first character string corresponding to the semantically matching portion in the matching in the collating step. A computer-readable recording medium for recording a program for executing an extracting step of extracting the extracted character string. 15. An instructing step for instructing a computer to a third character string to replace the second character string, and instructing the character string extracted in the extracting step, in the first instructing step. 15. The computer-readable recording medium according to claim 14, further comprising a program for executing a substituting step of substituting the generated third character string.