JP2001318935A - Information processor, its method, recording medium recording information processing software, and relational database - Google Patents

Abstract

PROBLEM TO BE SOLVED: To attain effective structure retrieval for contents of a document expressed by a data description language having hierarchical structure by using a relational database management system(RDBMS). SOLUTION: A conversion/storage part 1 is a means for converting the format of an XML document D and storing the converted result in a relational database(RDB). Concretely tags and a text between the tags are successively extracted as a pause from the provided XML document D, Respective lines including a document number different in each XML document D, a pause number expressing the order of each pause and the contents of the pause are recorded in a table of the RDB. A retrieval processing part 2 is a means for executing the processing such as retrieval/change of the RDB table converted from the XML document D and restoration of the original XML document D.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an improvement in technology for processing the contents of a document expressed in a data description language having a hierarchical structure by an RDBMS. It is suitable.

[0002]

2. Description of the Related Art In recent years, with the development of information processing technology, a data description language for expressing data having a hierarchical structure in a clear format has been proposed, and XML has been proposed as a typical example thereof. XML is a data description language for general-purpose data representation, and converts a document created according to XML into XML.
It is called an L document. In each part of the XML document, meaningful information is described in a predetermined text format that is easy for humans to read and does not depend on application software.

In this XML, each data included in a document is described between predetermined identification character strings called tags. A tag is a pair of a start tag and a close tag that correspond to each other. Among these, a general format structure of the start tag is <tag name [attribute name = "value" attribute name = "value". . . ]
>. In the start tag, an attribute list can be added as an option. In the above example, [] is an attribute list including one or more attributes. This attribute list is composed of one or more attribute name = “value” pairs, and each pair can be described in the form of attribute name = “value”.

The closing tag uses the same tag name as the corresponding start tag, and its general format structure is </ tag name>.

[0005] Information about actual events often has a hierarchical structure, but in the above XML, it is possible to freely describe the hierarchical structure by nesting tags.

[0006] Since XML itself is a format for expressing data, some specific means is required to realize effective storage and retrieval of a large amount of XML documents on a digital medium. . Here, as a technology for handling a large amount of data, RDB (Re
lational Database: relational database)
And an RDBMS (Relati) that manages and operates this RDB
Onal Database Management System (relational database management system) is generally used, and therefore, a request to use an RDBMS for storing and retrieving XML documents arises.

Here, XML is suitable for expressing data including a hierarchical structure, but the data structure of RDB is based on a two-dimensional "table" model and has a different basic structure from XML. Requires some ingenuity in order to handle by the RDBMS.

As a conventional technique for storing an XML document in an RDB as described above, a technique of mapping a specific element in the XML document to a field of a specific table (also called a table) in the RDB is known. I have.

[0009]

However, in the above-mentioned prior art, there is a problem in that it is necessary to first individually perform mapping definition work for each type of XML document, which is complicated. .

[0010] In the prior art as described above, the RDB
When the XML document is stored, the hierarchical structure of the original XML document is already lost, and there is a problem that the structure cannot be searched. That is, in a system using XML, it is desired that a search utilizing a characteristic of XML called a hierarchical structure, specifically, “structure search” can be performed.

This structure search is a search in consideration of a hierarchical structure. For example, a search operation for extracting a group of elements having a specified hierarchical structure and a search for only a group of elements having a certain element in a lower layer. It is desired that this can be easily performed.

On the other hand, in the above-described prior art, since the contents of the XML document are mapped to a flat data structure in the RDB, the hierarchical structure is lost, and the structure cannot be searched.

The present invention has been proposed to solve the above-mentioned problems of the prior art. The purpose of the present invention is to provide a method for converting the contents of a document expressed in a data description language having a hierarchical structure into an RDBMS. An object of the present invention is to provide an information processing technology that enables an effective structure search, that is, an information processing apparatus and method, a recording medium on which information processing software is recorded, and a relational database. Another object of the present invention is to provide an information processing technique for realizing an effective structure search for an XML document.

[0014]

According to a first aspect of the present invention, there is provided an information processing apparatus comprising:
Means for sequentially taking out tags and text between tags as breaks, a table of a relational database, a document number which differs depending on the document, a break number indicating the order of each break, and the extracted breaks. And means for recording each line including the content. Claim 1
An information processing method according to claim 4 is a method in which the invention of claim 1 is grasped from the viewpoint of a method. In the information processing method, a tag and a text between the tags are sequentially extracted from a given document as a phrase, and a relational database is provided. Recording each line including a document number that differs depending on the document, a break number indicating the order of each break, and the extracted contents of the break. According to a twentieth aspect of the present invention, the inventions of the first and fourteenth aspects are viewed from the viewpoint of a machine-readable recording medium on which computer software is recorded, and information processing software for processing information using a computer is recorded. In the recorded recording medium, the software causes the computer to sequentially extract tags and text between the tags from a given document as a phrase, and stores in the table of the relational database a document number that differs for each document and the phrase. Each line including a cut-off number indicating the order of each cut and the content of the extracted cut-off is recorded. The invention of claim 21 is based on the viewpoint of a relational database of the invention of claims 1, 14, and 20, wherein tags sequentially extracted from a given document and text between the tags are cut off, and the document Each row including a document number that differs depending on the type, a break number indicating the order of each break, and the extracted contents of the break is recorded in a table. According to a second aspect of the present invention, in the information processing apparatus according to the first aspect, there is provided an information processing apparatus, further comprising means for performing processing in units of the document based on a designated document number. According to the invention of claims 1, 14, 20, 21, and 2, the content of a document represented by a data description language having a hierarchical structure, for example, an XML document is converted into a RD by using a tag and a text between the tags as a unit.
B. This makes it possible to perform processing such as data storage, retrieval, and deletion on the RDBMS in a unified procedure regardless of the structure of the original XML document. In particular, it is not necessary to define individual mappings in advance, and the hierarchical structure expressed by nesting tags can be changed to RD.
It is also possible to hold it on B and perform a structure search. In addition, since the XML document is decomposed for each phrase and stored in the RDB, the limitation on the length of a character string that can be stored in the RDB is reduced in meaning. Furthermore, by identifying the document that is the source of each punctuation by the document number, even if punctuations derived from multiple documents are accumulated in a single table, the punctuation can be extracted in units of documents. -Processing such as deletion and search becomes possible. Also, R
The DB does not guarantee that the order in which the stored rows are retrieved is the same as the order in which they are recorded, but it is possible to retrieve the breaks in the same order as in the original XML document by using the sequential break numbers.

According to a third aspect of the present invention, in the information processing apparatus according to the first or second aspect, each tag existing in a path from a root to each punctuation in the original hierarchical structure in the document is provided in each line. It is characterized by including a text path represented by. According to a fifteenth aspect of the present invention, the invention of the third aspect is grasped from the viewpoint of a method. In the information processing method according to the fourteenth aspect, each line is added to each line from the root in the hierarchical structure in the original document. It is characterized in that it includes a text path represented by each tag existing in the path leading to the punctuation. According to a fourth aspect of the present invention, in the information processing apparatus according to the third aspect,
The information processing apparatus further comprises means for performing a process for a punctuation at a specified level or lower based on the text path. According to the third, fifteenth, and fourth aspects of the present invention, the text path enables processing such as data search by freely designating the position or range in the hierarchical structure of the original document. It is of course possible to specify a document number in this search, and so on.

According to a fifth aspect of the present invention, in the information processing apparatus according to the third or fourth aspect, a comparison table of a tag name of the tag and a tag number for each tag name is used, and the text path is the tag number. It is characterized by being expressed using According to a sixteenth aspect of the present invention, the invention of the fifth aspect is grasped from the viewpoint of a method. In the information processing method according to the fourteenth or fifteenth aspect, the tag name of the tag is compared with a tag number for each tag name. Using a table, the text path is expressed using the tag number. According to the fifth and sixteenth aspects of the present invention, the text path indicating the position of the phrase in the hierarchical structure is represented by the tag number, so that even if the tag name is long, the data amount of the text path can be minimized.

According to a sixth aspect of the present invention, in the information processing apparatus according to the fifth aspect, the tag number is a predetermined number of digits. In the invention of claim 6, for example, 4
By fixing the tag number to a fixed length by supplementing “0” to the beginning of the number that does not reach the digit, the function of matching one character in the RDBMS search can be used, and freedom such as “anything for one layer is acceptable” is possible. A high-level search becomes easy.

According to a seventh aspect of the present invention, in the information processing apparatus according to any one of the first to sixth aspects, an attribute table for storing an attribute included in the tag is used. According to a seventeenth aspect of the present invention, in the information processing method according to any one of the fourteenth to sixteenth aspects, an attribute for storing an attribute included in the tag is provided. It is characterized by using a table. According to an eighth aspect of the present invention, in the information processing apparatus according to the seventh aspect, there is provided a means for searching the attribute table for each row relating to a specified attribute name or attribute value. According to the seventh, eighth, and seventh aspects of the present invention, the attribute included in the tag is stored in the attribute table, so that the number of characters and the data amount of the punctuation content can be minimized. It is possible to freely search for punctuation with.

According to a ninth aspect of the present invention, in the information processing apparatus according to any one of the first to eighth aspects, each of the rows is
It is characterized by including a punctuation type indicating the type of each punctuation. The invention of claim 18 captures the invention of claim 9 from the viewpoint of a method, and in the information processing method according to any one of claims 14 to 17, each of the lines is a type of each of the punctuation. Is included. According to a tenth aspect of the present invention, in the information processing apparatus according to the ninth aspect, there is provided means for performing a process based on the type of the punctuation based on the punctuation type.
According to the ninth, eighteenth, and tenth aspects, the type of punctuation is represented by the punctuation type. For this reason, it is not necessary to represent the type of the punctuation with a specific kind of parenthesis in the punctuation content, so that the number of characters and the data amount of the punctuation content can be minimized, and the complexity based on the type of the punctuation can be reduced. Implementation of simple search becomes easier.

According to an eleventh aspect of the present invention, in the information processing apparatus according to the ninth or tenth aspect, the attribute name is treated as a kind of tag name, and whether or not the tag name is an attribute is distinguished by the punctuation type. And According to the eleventh aspect, the attribute name can be described as a tag name at the end of the text path. This eliminates the need to describe the attribute name in the punctuation content, thereby minimizing the number of characters and the data amount of the punctuation content, and also facilitates a search using a complex condition of a text path and an attribute name.

According to a twelfth aspect of the present invention, in the information processing apparatus according to any one of the first to eleventh aspects, each of the lines is separated from a root in the hierarchical structure of the original document by a respective phrase. It is characterized by including a punctuation number path in which each punctuation present in the path leading to is indicated by a punctuation number. Claim 19
The invention according to claim 12, wherein the invention according to claim 12 is grasped from the viewpoint of a method. In the information processing method according to any one of claims 14 to 18, each of the lines corresponds to the hierarchy in the original document. It is characterized in that it includes a pruning number path in which each pruning existing in the path from the root to each pruning in the structure is represented by a pruning number. The invention of claim 13 is the invention of claim 12
The information processing apparatus according to claim 1, further comprising means for performing a process based on an absolute position in a hierarchical structure in the document using the punctuation number path. According to the twelfth, nineteenth, and thirteenth aspects, the position of each break in the hierarchical structure in the original document is absolutely determined by a string of break numbers in the path from the root to the break. Since the same punctuation is displayed at a plurality of places, it is possible to clearly distinguish different punctuations that exist under the punctuation. For this reason, various information processing becomes easy, for example, when a specific punctuation number is obtained as a result of the condition search, the parent punctuation is sequentially traced.

[0022]

Next, embodiments of the present invention (hereinafter, referred to as "embodiments") will be specifically described with reference to the drawings. Note that this embodiment is typically realized by controlling a computer with software. The software in this case realizes the operation and effect of the present invention by physically utilizing the hardware of the computer, and the conventional technology is applied to the common parts with the conventional technology.

However, in this case, the types and configurations of hardware and software, the range of processing by software, and the like can be variously changed. For example, a hard disk drive, a disk pack,
A recording medium such as a D-ROM alone is one embodiment of the present invention. Therefore, in the following description, virtual circuit blocks that realize the functions of the present invention and the embodiments are used. Further, each part included in the embodiment can adopt a network configuration such as a LAN.

[1. Configuration] This embodiment shows an information processing system which is an information processing apparatus according to the present invention and an information processing method executed on the information processing system, and can be grasped as a recording medium recording information processing software and a relational database. Can also.

First, in this embodiment, as shown in the functional block diagram of FIG. 1, a relational database (RDB)
), A relational database management system (represented as RDBMS) for the RDB, and a processing unit X for XML conversion and the like. Here, the XML conversion etc. processing unit X performs RDBMS on the given XML document D.
And performs processing such as storage in the RDB, search and restoration, and the like, and includes a conversion storage unit 1 and a search and processing unit 2.

The conversion storage unit 1 is a means for converting the format of the XML document D and storing it in the RDB. More specifically, the conversion storage unit 1 converts a given XML document D into a tag, a text between the tags, Are successively taken out, and the table of RDB is
This is a means for recording each line including a document number that differs depending on each XML document D, a break number indicating the order of each break, and the extracted contents of the break. The search processing unit 2 is a means for performing processing such as search / change and restoration of the original XML document D with respect to the RDB table converted from the XML document D.

More specifically, the conversion storage unit 1 includes a document number determination unit 11, a phrase extraction unit 12, and a phrase number determination unit 1.
3, text path creation unit 14, and punctuation type creation unit 1
5, a section number path creation unit 16, a comparison table creation unit 17, and an attribute table creation unit 18. Further, the search processing unit 2 includes a document unit processing unit 21, a text path processing unit 22, a punctuation type processing unit 23, a punctuation number path processing unit 24, and an attribute processing unit 25. .

The document number determination unit 1 of the conversion storage unit 1
Reference numeral 1 denotes a part for determining a different document number for each XML document D. In addition, the phrase extracting unit 12 is a part that sequentially extracts tags and text between the tags from the XML document D as a phrase. In addition, the punctuation number determination unit 13
Is a part for sequentially determining the punctuation numbers indicating the order of each punctuation by sequentially incrementing. Each set of the document number, the punctuation number, and the punctuation content (referred to as punctuation content) thus obtained is recorded as each row of the main body table T1 which is an RDB table. Further, the document unit processing unit 21 of the search processing unit 2 is a unit that performs processing in units of the XML document D based on the designated document number.

The text path creation unit 14 of the conversion storage unit 1 exists in each row of the main body table T1 in the RDB in a path from the root to each punctuation in the hierarchical structure in the original document. This is a means for adding a text path represented by each tag. Further, the text path processing unit 22 of the search etc. processing unit 2 is a means for performing processing for a punctuation below a specified hierarchy based on the text path.

Further, the phrase type creation unit 1 of the conversion storage unit 1
Reference numeral 5 denotes a unit for adding a punctuation type representing each punctuation type to each row of the main body table T1 in the RDB. In addition, the punctuation type processing unit 23 of the search processing unit 2 is a unit that performs a process based on the type of punctuation based on the punctuation type.

The break number path creating unit 16 of the conversion storage unit 1 stores each path of the main body table T1 in the RDB into a path from the root to each break in the hierarchical structure of the original XML document D. This is a means for adding a break number path in which each existing break is represented by a break number. Further, the phrase number path processing unit 24 of the search etc. processing unit 2 is means for performing processing based on the absolute position in the hierarchical structure in the XML document D based on the phrase number path.

The comparison table creation unit 17 of the conversion storage unit 1 stores a comparison table (referred to as a tag list table) T2 between the tag name of each tag and the tag number of each tag name in the RD.
This is a means for creating on the RDB through the BMS. When this tag list table T2 is used, the text path is expressed using the tag number.

The attribute table creation unit 18 of the conversion storage unit 1 is a means for creating an attribute table T3 for storing attributes included in each tag on the RDB through the RDBMS. The unit 25 is a means for searching each row related to the specified attribute name or attribute value based on the attribute table T3.

[2. Operation and Effect] Next, the operation of the present embodiment configured as described above will be described with reference to a plurality of storage models. Here, the following Expression 1 is an example of an XML document (sample XML format) used in the following description of each embodiment.
L document or sample document), which represents a part of a product list in a certain personal computer shop.

(Equation 1)

In the present embodiment, the most basic concept when handling an XML document is the idea of "decomposing an XML document into words." That is, the phrase break is obtained by decomposing the content of the XML document in units of a tag and text sandwiched between the tags.

[2-1. First Storage Model] The first storage model is a basic example based on the above concept,
Table 1 shows the meaning of each column in the main body table T1 on the RDBMS for storing an XML document in the first storage model.

[Table 1] Here, the “document number” is a key for identifying each document when storing a plurality of XML documents, and specifically, a different number is stored for each XML document. Also,
"Phrase number" indicates the order in which the punctuation appears from the beginning of the document. Generally, in RDBMS,
This column is necessary because the order in which the stored rows (also called records) are retrieved is not guaranteed. “Phrase content” stores a character string constituting a punctuation.

[2-1-1. Example of processing procedure for realizing first storage model] In addition, the main body table T1 in the first storage model includes, in the conversion storage unit 1, a document number determination unit 11, a phrase extraction unit 12, and a phrase number. Decision unit 13
Is created from the given XML document D based on the processing procedure illustrated in the flowchart of FIG.

That is, in this procedure, the document number determination section 11 determines the document number (step 101), and the break number determination section 13 resets the initial value of the break number to 0 (step 102). Then, until the end of the document is reached (step 106), the phrase extraction unit 12 outputs one phrase from the XML document D.
Each time one punctuation is extracted (step 103), the punctuation number determination unit 13 increments the punctuation number (step 104), and the document number, the punctuation number, and the punctuation content are added as one record in the main body table T1. (Step 10
5). Note that the content of each record is called punctuation information, and the same applies to the case where other columns are included as in other storage models described later.

[2-1-2. Example of first storage model]
The following Table 2 shows an example in which the sample document of Expression 1 is stored in the main table T1 of the RDB in the format of Table 1 in accordance with the processing procedure of FIG.

[Table 2] In this example, the document number is temporarily set to “1”, and the same applies hereinafter.

[2-1-3. Example of processing based on first storage model] By storing the XML document in the main body table T1 in the format described above, the following operation can be performed. -Retrieval of XML document with specified document number-Deletion of XML document with specified document number

Specifically, X having a designated document number
Extraction of the ML document D can be easily performed by the following procedure. That is, first, the record having the designated “document number” is extracted from the table while rearranging the records in the “punctuation number” column. The following Expression 2 shows an example of an SQL sentence used for this operation.

(Equation 2) The characters in italics are search keys, and so on.
Subsequently, all the extracted “phrase contents” are connected in the same order. As a result, the designated document number is changed to XM
The L document D can be taken out.

The deletion of the XML document D having the designated document number can be easily performed by deleting all the records having the designated "document number" from the main body table T1. An SQL sentence used for this operation is exemplified in the following Expression 3.

(Equation 3) The SQL sentence used for these processes may be input manually or may be passed from another application software or the like to the RDBMS. It may be automatically created based on the instruction content through a simple user interface, and the same applies to the following.

[2-1-4. Advantages of the first storage model]
As described above, in the first storage model, the content of a document represented by a data description language having a hierarchical structure, for example, an XML document, is expressed in RDB in units of tags and text between tags, as compared with the prior art. To be stored. Thereby, RD
Even on the BMS, processing such as data storage, retrieval, and deletion can be performed in a uniform procedure regardless of the structure of the original XML document. In particular, it is not necessary to define mapping in advance and individually, and it is also possible to hold a hierarchical structure expressed by nesting tags on the RDB and perform a structure search. In addition, when storing a character string in the RDB, there is usually a limit to the length of a character string that can be stored.
Since the XML document is decomposed for each phrase and stored in the RDB, the limit on the length of a character string that can be stored in the RDB is relaxed semantically.

Further, by identifying the document that is the source of each punctuation by a document number, even if punctuations originating from a plurality of documents are accumulated in a single table, a phrase in units of a document can be obtained. Processing such as removal, deletion, and retrieval of cuts can be performed. Also, RDB
Although it is not guaranteed that the order in which the stored rows are retrieved is the same as the order in which they are recorded, it is possible to retrieve the breaks in the same order as in the original XML document by using the sequential break numbers.

[2-2. Second Storage Model] The second storage model introduces a text path to realize a structure search. The text path is a column indicating to which layer each of the breaks stored in the table on the RDBMS belongs in the semantic hierarchical structure in the original XML document. Table 3 below shows the meaning of each column when the text path is introduced in the main body table T1.

[Table 3] Here, the root (root) in the hierarchical structure is indicated by “/”, and the notation that separates each layer by “/” is used. Such a delimiter is temporarily called a delimiter. The specific type of the delimiter can be freely defined. Other columns are the same as in Table 1 shown for the first storage model.

[2-2-1. Example of processing procedure for realizing second storage model] In addition, the main body table T1 in the second storage model includes a document number determination unit 11, a phrase extraction unit 12, and a phrase separation number determination in the conversion storage unit 1. In addition to the operation of the unit 13, the operation of the text path creating unit 14 in particular, based on the processing procedure illustrated in the flowchart of FIG.
It is realized from a given XML document D.

That is, in this procedure, the phrase extracting section 12
, The tag name is pushed to the text path acquisition stack (step 203), while in the case of the start tag (step 202),
If the tag is a closing tag (step 206), the tag is popped from the text path acquisition stack (step 207). Then, a text path is acquired from the text path acquisition stack for each punctuation (step 204), and punctuation information including the text path is added to the main body table T1 (step 205).

In order to obtain a text path from the text path acquisition stack at each time, the tag names included in the copy of the text path acquisition stack at that time are connected by a delimiter in the order in which they were pushed. do it.

[2-2-2. Example of second storage model]
Table 4 below shows an example in which the sample document of the above formula 1 is stored in the main table T1 of the RDB in the format of the above table 3 according to the processing procedure of FIG.

[Table 4]

[2-2-3. Example of processing based on second storage model] In the second storage model, by introducing the concept of a text path as described above, the “specified hierarchical structure in a document that spans all / or a specified document” A basic structure search such as "retrieve the following" can be performed.

Here, at the time of structure search, in order to designate a hierarchical structure given as a search key or a search condition from the outside, similarly, '/' is used as a delimiter, and this is expressed as "hierarchical structure name". Shall be. In this case, records satisfying any of the following conditions can be taken out by rearranging them using the document number and the break number as keys, and by connecting them, the result of the structure search can be obtained. (1) The text path string is the same as the hierarchical structure name. (2) The text path string starts with the character string composed of the hierarchical structure name + '/'.

If, for example, only a document having a designated document number is to be subjected to such a structure search, a condition for a document number is added to the above conditions.

Here, in the second storage model, RD
Specific SQL statements used as an instruction to the BMS are shown in Expressions 4 and 5 below. (1) When searching across all documents:

(Equation 4) (2) When searching for a document with the specified number:

(Equation 5)

Next, a more specific example of such a structure search will be described. For example, in the case of a structure search of “query the hierarchical structure name“ / product list / product / product name ”for the document of document number 1”, first, for example, the following operation is performed on the RDBMS by the operation of the text path processing unit 22. SQL
Issue a statement.

(Equation 6) In this case, the next result set is returned from the RDBMS.

[Table 5] By combining all of the result sets described above and forming one character string, a search result such as "<product name> A4 paper </ product name><productname> thin printer </ product name>" is obtained. can get.

The above search results can be semantically separated into the following two cases. First case: "<Product name> A4 paper </ Product name>" 2 Subject: "<Product name> Thin printer </ Product name>"

Here, FIG. 4 is a flowchart showing a procedure for obtaining a result in such a separated form instead of concatenating all the result sets into one character string. In this procedure, for example, the text path processing unit 22
However, after clearing the work area (step 301), one line is read from the result set (step 302), and if the read punctuation content is the start tag (step 303), the content is pushed onto the stack (step 304). Further, the read data is added to the end of the work area (step 30).
5).

If the read punctuation content is a closing tag (step 306), it is popped from the stack (step 307), and if the stack becomes empty (step 308).
The work area is added to the list and the work area is cleared (step 309). As a result of repeating such processing, when the end of the result set is reached (step 310), the processing is terminated. In such a processing procedure, each is stored as one item in a “list” (array) based on the contents of the result set.

[2-2-4. Modification Example Based on Second Storage Model] As a modification of the second storage model as described above, a unique tag name number is assigned to the tag name by the operation of the comparison table creating unit 17, and the tag name and the tag name are assigned. The correspondence between the numbers may be stored in a table (referred to as a tag list table) T2 different from the main body table T1, and the text path column of the main body table T1 may be expressed using this tag name number. Here, the meaning of each column included in the tag list table T2 is exemplified in Table 6 below.

[Table 6]

In this case, a specific example of the tag list table T2 based on the sample document and a specific example of the corresponding main body table T1 are shown in Tables 7 and 8, respectively.

[Table 7]

[Table 8]

Further, when storing the tag name number in the text path string, for example, if the number of digits of the tag name number does not reach four digits in the case of four digits, '0' may be added to the head of each. . Table 9 below shows a part of the head of the storage example in the main body table T1 in this case.

[Table 9]

According to such a structure, since the number of characters occupied by one layer in the text path string is determined in advance, the function of one character matching in the RDBMS search is used to say, "Anything in one layer is acceptable." A meaning search can be easily performed.

In such a structure, in the name of the hierarchical structure, the expression “applies to all of one hierarchy” is represented by a wildcard “*”, and “document number 1”
Inquiry about hierarchical structure name "/ E / * / W" for document "" is performed. As a precondition in this case, one tag name number in the text path is always stored in four digits, and if it is less than four digits, a leading zero is added.

Such a search is realized by a process including the following steps. (1) The tag name numbers of the tags E and W are obtained from the tag list table T2. (2) As an example, when the tag name number of the tag E is obtained as 17 and the tag name number of the tag W is obtained as 23, the following SQL statement is issued. In this example, a wildcard '_' that matches one character is used.

(Equation 7) (3) The processing after the result set is obtained from the RDBMS is the same as the search example in Table 4 described above.

As described above, in the second storage model, it is possible to freely specify the position or range in the hierarchical structure in the original document by using the text path and perform processing such as data search. Also, since the text path indicating the position of the phrase in the hierarchical structure is represented by the tag number, the data amount of the text path can be minimized even when the tag name is long. Also, by making the tag number a fixed length by supplementing “0” at the beginning of the number that does not reach four digits, the function of matching one character in the search of the RDBMS or the like is used. ”Can be easily searched.

[2-3. Third Storage Model] Up to the second storage model, the attribute has always been treated as a part of the start tag, but by introducing the attribute table T3, it is possible to realize an easy search for the attribute content. it can. [2-3-1. Example of Third Storage Model and Procedure for Realization] First, the meaning of each column of the attribute table T3 is shown in Table 10 below.

[Table 10]

As shown in FIG. 5, in addition to the processing procedure shown in FIG. 3, the processing procedure at the time of document storage for realizing the third storage model is performed by the operation of the attribute table creating unit 18. If an attribute list is included in the start tag (step 403), a process of storing those attributes in the attribute table (step 404) is added.

According to the processing procedure shown in FIG.
Attribute table T3 created based on the sample document
The following is an example.

[Table 11] Note that the main body table T1 in this case is the same as each example described above.

[2-3-2. Example of Retrieval in Third Storage Model] In the above-described third storage model, the operation of the attribute processing unit 25 causes a phrase having a specific attribute name or attribute value to be processed according to the following procedure. Can be searched.

(1) First, a record of an attribute to be searched is searched from the attribute table T2. An example of an SQL sentence used by the attribute processing unit 25 for this search is shown below.

(Equation 8)

(2) Next, based on the obtained document number and punctuation number, the entire element including the element is extracted from the main body table. Here, FIG. 6 shows a processing procedure for extracting the entire element including the document number and the punctuation number when the document number and the punctuation number are specified, and the result is entered in the “work area”. That is, in this procedure, after clearing the work area (step 501), the content of the punctuation is extracted from the main body table T1 using the document number and the element number as keys (step 50).
2). At this time, if the extracted punctuation content is a start tag (step 503), it is pushed onto the stack (step 504), and an attribute list included in the start tag is obtained (step 505). The original punctuation content is restored by the acquired attribute list.

The punctuation content is combined with the end of the work area (step 506). If the punctuation content is a closed tag (step 507), it is popped from the stack (step 5).
08) As a result, if the stack becomes empty, the procedure ends (step 509). If the tag is not a closing tag or the stack is not empty, the element number is increased by 1 (step 510), and the process returns to step 502.

Next, a specific example of a search using an attribute name and an attribute value will be described. For example, in the sample document shown in Expression 1, it is assumed that an element whose attribute name is “classification” and whose attribute value is “peripheral device” is searched. In this case, (1) First, the following SQL statement is issued to the attribute table.

(Equation 9) This results in the following result set:

[Table 12] (2) Then, according to the flowchart of FIG. 6, the entire element is extracted based on the obtained document number and punctuation number.

As a result, the following final result is obtained. <Product category = "Peripheral equipment"> <Product name> Thin printer </ Product name> <Price unit = "US $"> 980 </ Price> </ Product>

Further, as described above, not only the search based on the attribute name and the attribute value but also the search linked to the document structure, that is, the search for the hierarchical structure name having the specified attribute name and the attribute value can be performed. Is possible, and such a search
For example, it can be realized by an SQL statement as exemplified below.

(Equation 10) Thereby, the document number and the punctuation number of the corresponding punctuation can be obtained.

In the above structure, information about attributes is stored in both the main body table and the attribute table. However, a format in which the attributes are separated from the main body table to eliminate the duplication can be adopted. A processing procedure for storing a document in that case is illustrated in a flowchart of FIG.

In this procedure, if the extracted phrase is a start tag (step 601), the attribute list is added to the attribute table T3 (step 604) if the attribute list is present (step 603) if the attribute list is present (step 603). The attribute list is deleted from the punctuation (step 605). In addition, the tag name is pushed to the text path acquisition stack for the extracted punctuation (step 606), and the text path is acquired (step 607).
1 (step 608). If the extracted phrase is a closing tag (step 609), the phrase is popped from the text path acquisition stack (step 610), and the above procedure is repeated until the end of the document (step 611).

According to the processing procedure of FIG. 7,
An example in which the sample document of Expression 1 is stored is shown below.

[Table 13]

As described above, in the third storage model, by storing the attributes included in the tag in the attribute table,
The number of characters and the amount of data in the punctuation content are minimal, and
It is possible to freely search for a phrase having a desired attribute or attribute value from the attribute table.

[2-4. Fourth storage model] The fourth storage model introduces a “phrase type” column representing the type of each punctuation into the main body table T1 by the operation of the punctuation type creation section 15, thereby obtaining the main storage table T1. Only the name of each punctuation can be stored as punctuation content, and the implementation for implementing a complicated search is made easier.

First, Table 14 shows the definition contents of the punctuation type.
Table 15 shows an example in which a sample document of Expression 1 is stored according to the definition. Here, “attribute” is also defined as a punctuation type, and all are stored in the main body table without using an attribute table separately.

[Table 14]

[Table 15] In this case, the phrase separation type processing unit 23 of the search etc. processing unit 2
The type of each punctuation can be determined based on the punctuation type.

Further, the attribute can be treated as a kind of tag name. That is, the main body table T1 shown in Table 15
As a modification of, there is a method in which attribute names are treated as a kind of tag name and stored. In this case, the attribute can be distinguished by the punctuation type. A storage example of the main body table T1 in this case is shown below.

[Table 16]

In such a fourth storage model, the type of punctuation is represented by the punctuation type. For this reason, it is not necessary to represent the type of the punctuation with a specific kind of parenthesis in the punctuation content, so that the number of characters and the data amount of the punctuation content are minimized, and complicated Implementation of simple search becomes easier.

[2-5. Fifth Storage Model] The fifth storage model is an example in which a punctuation number path is introduced. That is, the “text path” in the second storage model does not completely define the hierarchical relationship between individual breaks in the document. That is, in the XML document D, there is a possibility that the same punctuation appears in a plurality of places. Therefore, for example, when attention is paid to a certain punctuation in the main body table T1, which layer is higher (parent punctuation) is determined. It is impossible to make an unambiguous judgment. In order to make such a determination possible, by introducing a “phrase number path” by the operation of the phrasing type creation unit 15, it becomes possible to completely store information relating to the hierarchical structure of the document.

That is, in the present invention, a document is decomposed into phrases, and a phrase number is assigned to each document. The phrase number path referred to here is used for each phrase using this phrase number. This shows the hierarchical structure from the document root of the document.

For this phrase number path, a means such as a symbol for separating the phrase numbers can be freely selected, but here, the phrase numbers are stored separated by '/'.
The meaning of each column constituting the main body table T1 including such a pass numbering path is shown below.

[Table 17]

Further, according to such a format, the following equation 1 is obtained.
The following is an example of the main body table T1 storing the sample document. In this example, the attributes are stored in the attribute table.

[Table 18]

In the fifth storage model, the position of each break in the hierarchical structure in the original document exists in the path from the root to the break by using the break number path. Since the phrase is absolutely displayed by the column of the phrase number, even if the same phrase is present in a plurality of places, it is possible to clearly distinguish different phrases present under the respective phrases. For this reason, various information processing is facilitated, for example, by the operation of the phrase number pass processing unit 24, when a specific phrase number is obtained as a result of the condition search, the parent phrase is sequentially traced. become. [3. Other embodiments]

The present invention is not limited to the above embodiments, but includes other embodiments as exemplified below. For example, the RDBM in the above embodiment
Although S is assumed to be a database management system that supports queries in a standard SQL language, the method of the present invention can be applied to other types of database management systems.

As for the assignment of the tag name number and the punctuation number, a unique number is assigned to the number throughout the entire document instead of always starting from 1 at the beginning of the document (global uniqueness of the number). It is also possible to make the implementation easier, for example, by distinguishing the same punctuation that exists in a plurality of documents without referring to the document number. For the closing tag, RDBM
The storage in the table on S may be omitted, and the document may be restored at the time of document retrieval.

The individual features described for each of the storage models described above can be combined infinitely, but can be implemented in any combination.

[0091]

As described above, according to the present invention,
An excellent information processing technique, that is, an information processing apparatus and method, which enables an effective structure search on an RDBMS with respect to the content of a document expressed in a data description language having a hierarchical structure,
It is possible to provide a recording medium on which information processing software is recorded and a relational database.

[Brief description of the drawings]

FIG. 1 is a functional block diagram showing a configuration of an embodiment of the present invention.

FIG. 2 is a flowchart showing a processing procedure for creating a main body table from an XML document in a first storage model in the embodiment of the present invention.

FIG. 3 is a flowchart showing a processing procedure for creating a main body table from an XML document in a second storage model in the embodiment of the present invention.

FIG. 4 is a flowchart showing a processing procedure for obtaining a result in a separated format in a second storage model in the embodiment of the present invention.

FIG. 5 is a flowchart showing a processing procedure for storing a document in an RDB in a third storage model according to the embodiment of the present invention.

FIG. 6 is a flowchart showing a processing procedure for extracting an entire element including a document number and a break number when a document number and a break number are specified in the third storage model in the embodiment of the present invention.

FIG. 7 illustrates a third storage model according to the embodiment of the present invention in which a document is stored in a format in which attributes are separated from a main body table.
9 is a flowchart showing a processing procedure to be stored in a DB.

[Explanation of symbols]

 D: XML document X: XML conversion processing unit RDB: Relational database T1: Body table T2: Tag list table T3: Attribute table RDBMS: Relational database management system 1: Conversion storage unit 11: Document number determination unit 12: Phrase extraction Unit 13: Punctuation number determination unit 14: Text path creation unit 15 ... Punctuation type creation unit 16 ... Punctuation number path creation unit 17 ... Contrast table creation unit 18 ... Attribute table creation unit 2 ... Search processing unit 21 ... Document Unit processing unit 22 Text path processing unit 23 Punctuation type processing unit 24 Punctuation number path processing unit 25 Attribute processing unit

Claims (21)

[Claims] 1. A means for sequentially taking out tags and text between tags as a break from a given document, a relational database table showing a document number which differs for each document, and an order for each break. The punctuation number,
Means for recording each line including the extracted contents of the punctuation, and an information processing apparatus comprising: 2. The information processing apparatus according to claim 1, further comprising means for performing processing in units of the document based on a designated document number. 3. The method according to claim 1, wherein each line includes a text path represented by each tag existing in a path from a root to each punctuation in the hierarchical structure in the original document. An information processing apparatus according to claim 1. 4. The information processing apparatus according to claim 3, further comprising means for performing a process for a punctuation at a specified hierarchy or lower based on the text path. 5. The text path is expressed using the tag number using a comparison table of the tag name of the tag and a tag number for each tag name. Information processing device. 6. The information processing apparatus according to claim 5, wherein the tag number has a predetermined number of digits. 7. The information processing apparatus according to claim 1, wherein an attribute table storing attributes included in the tag is used. 8. The information processing apparatus according to claim 7, further comprising means for searching the attribute table for each row related to a specified attribute name or attribute value. 9. The information processing apparatus according to claim 1, wherein each row includes a punctuation type indicating a type of each punctuation. 10. The information processing apparatus according to claim 9, further comprising means for performing processing based on the type of the punctuation based on the punctuation type. 11. The information processing apparatus according to claim 9, wherein the attribute name is treated as a kind of tag name, and whether or not the tag name is an attribute is distinguished by the punctuation type. 12. The method according to claim 1, wherein each line includes a break number path in which each break present in a path from a root to each break in the hierarchical structure in the original document is represented by a break number. The information processing apparatus according to any one of claims 1 to 11, wherein 13. The information processing apparatus according to claim 12, further comprising means for performing processing based on an absolute position in a hierarchical structure in the document using the break number path. 14. A step of sequentially extracting a tag and a text between tags as a break from a given document; and displaying a document number which differs depending on the document and an order for each break in a table of a relational database. The punctuation number,
Recording each line including the extracted contents of the punctuation. 15. The method according to claim 14, wherein each line includes a text path represented by each tag present in a path from a root to each punctuation in the hierarchical structure in the original document. Information processing method. 16. The text path is expressed using the tag number, using a comparison table of the tag name of the tag and a tag number for each tag name. Information processing method. 17. An apparatus according to claim 14, wherein an attribute table storing attributes included in said tag is used.
7. The information processing method according to any one of 6. 18. The method according to claim 14, wherein each of the lines includes a punctuation type indicating the type of each punctuation.
The information processing method according to any one of the above. 19. The method according to claim 19, wherein each line includes a break number path in which each break present in a path from the root to each break in the hierarchical structure in the original document is represented by a break number. The information processing method according to any one of claims 14 to 18, wherein: 20. A recording medium on which information processing software for processing information by using a computer is recorded, wherein the software separates a tag and a text between the tags from a given document. In the relational database table, a document number that differs depending on the document, a punctuation number indicating the order of each punctuation,
A recording medium on which information processing software is recorded, wherein each line including the extracted content of the punctuation is recorded. 21. A tag which is sequentially extracted from a given document and a text between the tags are cut off, and a document number which differs depending on the document, a cut-off number indicating the order of each cut-off, A relational database, characterized in that each row containing the contents of the punctuation is recorded in a table.