JP4410005B2 - Structured document structure conversion device and program - Google Patents

Description

  The present invention relates to a method, an apparatus, and the like for performing structure conversion / inverse conversion from an XML document to an XML document.

  In recent years, various types of systems such as individuals, companies, and local governments are connected to each other through the Internet so that they can communicate with each other, and these systems cooperate with each other to provide Web services or EDI (Electronic Data Interchange). EC (Electronic Commerce) is being carried out. For this reason, a wide range of information exchange is required.

  Under such circumstances, XML (eXtensible Markup Language) has a flexible expression ability to structure data and is suitable for processing by a computer. Therefore, data exchange between the above systems and data in each system are possible. It is attracting attention as a common platform format for processing.

  XML was established in February 1998 by W3C (World Wide Web Consortium) in order to facilitate the use of SGML (Standard Generalized Markup Language) standardized by ISO in 1986 on the Internet. It has been done.

  HTML (HyperText Markup Language), a Web page creation language that has been used in the past, has a fixed tag and is specialized for display, and meets the requirement to process information on a computer based on tag information There was a problem that could not be done.

  On the other hand, XML has a language structure in which a user can freely define a tag and can give meaning to a character string in a document. When a document is described in such XML, the document can be processed by a computer based on tag information.

Note that XML documents are roughly classified into the following two types according to their characteristics.
・ Data type XML documents: slips, schedules, etc. with many tags and short element contents ・ Document type XML documents: articles with long element contents such as magazines, manuals, dictionaries, etc. Suppose that the target is a type XML document.

  Here, terms used in the following description will be described based on the XML standard. It is well known that the character string surrounded by a pair of “<” and “>” is “tag”, “<character string>” is “start tag”, “</ character string” is "End tag", the entire character string from the start tag to the end tag is "element", the character string between the start tag and the end tag is "element content", and the name of the element described in the tag is "tag name" "(Or" element name "), additional information for the element is called" attribute ".

  In a structured document, the data structure is described in a form in which tags are embedded in the document. In this way, by adopting a structure in which the data structure is embedded in the document as a tag, flexibility and extensibility can be obtained for adding, deleting, and changing data items, and the tag name is read and meaningful by humans. By giving a certain name, the data can be made visible.

  By the way, in order to improve the processing capability for XML documents by increasing the processing speed for XML documents, reducing the memory usage, etc., it is generally the mainstream to improve the performance of the base software. Yes. However, in addition to such a method, it is possible to improve the processing performance for an XML document by processing the XML document itself in advance. The present invention relates to the latter method (a method for improving the processing performance by processing an XML document), and here, a conventional technique related to the latter method will be described.

  For example, Non-Patent Document 1 discloses a case where a problem occurs that processing speed is slow when XML is introduced and the problem is dealt with by changing the data structure. For example, in the example of Sumitomo Electric Systems (see p.64-65 of the same magazine), the same type of data is described in one CSV (Comma Separated Value) format, and the combined data is stored in an XML document. It is disclosed to embed in one tag. In other words, “the CSV data is embedded in the XML data”. For example, the definition information of the XML data is changed, and the XML data for one month is grouped by separating them with commas in order of date.

In particular,
<KOUSU day = ”01”> 8.0 </ KOUSU><KOUSU day = ”02”> 5.5 </ KOUSU>… <KOUSU day = ”31”> 12.8 </ KOUSU>
As such, data on daily performance that was described in separate tags,
<KOUSU day = ”01, 02,…, 31” data = ”8.0, 5.5,…, 12.8”></KOUSU>
In this format, the original document is rewritten so that it can be collected monthly.

  As a result of this change, when referring to the data for one month, the database server can be inquired only once, and the XML definition information is transmitted only once, so the data capacity is also reduced. It is said that it has decreased to 1/10.

  Also, Non-Patent Document 2 aims to reduce the amount of data. An XML document in a record format is converted to a CSV format for all elements in a record by using XSL conversion while maintaining the XML document standard. Conversion to an XML document connected in a format is disclosed. In order to reduce the load of data processing, it is intended to handle a document in which all elements in a record are collected in one CSV format by a dedicated API.

  Specifically, XML documents before and after conversion by the method of Non-Patent Document 2 are as shown in FIGS. 21A and 21B, for example. FIG. 21A is an original XML document before conversion, and FIG. 21B is an XML document after conversion.

  As shown in the figure, the converted XML document is divided into two parts. One is a part describing each tag name of the original XML document, and the other is a part describing the contents (1, 2, 3, 4, etc.) of each element connected in CSV format. However, this corresponds to a standard XML document, and an atypical XML document was not included.

  By the way, in an XML document that is a typical structured document, in order to allow some application software to handle the XML document (to perform operations such as search, update, and deletion), DOM (Document Object Model) and Two standard interface (API: Application Programming Interface) standards called SAX (Simple API for XML) are defined.

  SAX is characterized by low memory consumption and generally high speed, but is suitable for simple processing with only time series output and reference. On the other hand, DOM is generally low speed and consumes a large amount of memory. However, since DOM elements are expanded into a hierarchical tree structure, a program can be easily assembled even with complicated processing contents.

  In general, when an operation such as search / update / deletion is performed on an XML document, the operation target XML document is expanded into a DOM tree using a standard API (DOM), and then the operation is performed. However, when an XML document is expanded into a DOM tree, an operation memory capacity that is six times the original data amount is required, and items that are not used (items that are not subject to operation) are also expanded. For this reason, the development process takes a long time (the processing speed and the memory consumption are proportional to the number of elements of the XML document).

The reason for improving the processing performance by processing the XML document as in Non-Patent Documents 1 and 2 is because of such circumstances.
However, the non-patent documents 1 and 2 have the following problems.

  First, the method described in Non-Patent Document 1 is a data-dependent individual method, and is not an organized general-purpose method. In other words, the method described in Non-Patent Document 1 is a method of combining the same kind of data used for data processing into one, and is applied to specific data having the same kind of data, and the effect of the improvement depends on the data. To do. In other words, it is not a general purpose method.

  In the method described in Non-Patent Document 2, the data amount can be reduced by removing the tag of the XML document, but the data processing load of the existing application software cannot be reduced by this method.

  Non-Patent Document 2 assumes that special API software that can handle converted documents is created to reduce the load of data processing. This means that software having the same function as existing DOM software must be created separately. For this reason, this work requires a lot of man-hours. Therefore, it is unlikely to be used in the same way as existing DOM.

The method described in Non-Patent Document 2 assumes only a standard (table format) XML document.
In contrast to such a conventional technique, the applicant of the present application has proposed a method described in Patent Document 1.

  In the invention of Patent Document 1, elements in a record in an XML document in a record format are divided into target items (key elements) and non-target items (non-key elements) for application software data processing. Proposes converting the contents of non-key elements into an XML document as a new element (referred to as a CSV element) while keeping the key elements as they are. If the XML document is atypical, the element name of the element grouped in the new element in the CSV format is attached to the attribute. This conversion (hereinafter referred to as CSV compression conversion) is executed as XSL conversion.

In this CSV compression conversion, the key elements that are data processing target items are not converted to the CSV format but are left as they are, and therefore can be applied with only slight modifications to the application software. In addition, by deleting the tags of non-key elements and combining the contents of the elements into one new element, the memory usage of XML document processing can be reduced and the memory expanded according to the number of elements in which the tags of the original document are reduced. Time and processing time can be shortened.
JP 2003-203667 A "The Truth of Universal Illusions That Has Been Visible Overturns XML's" common sense "", Nikkei Computer Magazine 2001.3.12, p52-p71 “Building an XML Bloat Buster using ZXMLXML Compression Mehod” by Alain Trotter; [Search February 18, 2002], Internet <URL: http://www.ASPToday.com/> Or as an overview, <URL: http: //www.XML.com/pub/r/904>

As described above, the invention of Patent Document 1 proposes a superior technique especially for application software to process a converted XML document as compared to the prior art.
However, the invention of Patent Document 1 still has room for improvement. In particular, when the XML document to be converted is an atypical XML document (the number of elements in the record and the tag name is variable), there is a large room for improvement.

In this regard, the applicant of the present application has further proposed the invention of Japanese Patent Application No. 2003-165735 (hereinafter referred to as the prior application).
In the prior application, in order to convert the structure of the standard XML document, a plurality of new elements in the converted XML document are defined, and each element in the record in the XML document to be converted is a key element that is a target of data processing. And whether to assign each non-key element, which is an element other than the key element, to which of the plurality of new elements, the definition defines the elements in the order in which they appear in the record. The above key elements are described in the converted XML document as they are, and for each non-key element, the contents of the element are summarized in CSV format for each corresponding new element, and the converted XML document is used as the element contents of each new element. Output to.

In the prior application, the following three methods have been proposed in order to convert the structure of an atypical XML document.
In the first method, when an atypical XML document is handled, the converted XML document has, as its header, a new element name (CSV element name) and element names of non-key elements to be collected (possibly appearing) Additional information enumerating all of them is attached. Then, for each record, the element content of the non-key element that appears is connected to the CSV element in the CSV format and stored. At that time, atypical elements that did not appear are connected as empty elements. On the other hand, empty elements are not output during reverse conversion. FIG. 22 shows an example of an original document, and FIG. 23 shows an example of a converted document.

  Here, conversion specifications are not shown, but “book number” and “book name” are defined as key elements, “author name”, “registration date”, and “last update date” are defined as non-key elements. To do. Furthermore, it is assumed that “author name” may appear up to four people. The tag name of the new element (CSV element) is defined as “information”. Accordingly, as shown in FIG. 23, in the additional information header, the tag name of the non-key element (author name [1] to author name [4], registration date, last update date) regarding the CSV element “information”. Are connected and described in CSV format. In the first record, since there is only one author, empty elements (indicated in the figure,,,) are described for the author name [2] to author name [4]. In the second record, since there are three authors, an empty element (, shown) is described for the author name [4].

  In the second method, only the difference from the first method will be described. That is, in the converted XML document, empty elements are described for non-key elements that did not appear in the first technique, whereas in the second technique, the additional information of the appearing non-key elements is described. Is described as the attribute of the tag of the CSV element (this is also described by being linked in the CSV format).

  When the original document is FIG. 22, the XML document after conversion by the second method is as shown in FIG. In the example shown in the figure, the attribute tag of the CSV element “information” tag describes the order of the added non-key elements in the additional information in a CSV format. That is, in the additional information header, the author name [1], the author name [2], the author name [3], the author name [4], the registration date, and the last update date are described as in the first method. Therefore, the author name [1] is first and the last update date is sixth. In the first record, since there is only one author, the non-key elements that appear are the first, fifth, and sixth elements, and 1, 5 and 6 are described in the attribute tag as shown. The

  In the third method, structure conversion can be performed even for an atypical XML document in which items in a record are switched according to conditions in the record. An example of this type of atypical XML document is shown in FIG. 25, and a converted XML document obtained by converting the structure using the conversion specification of FIG. 27 or FIG. 28 is shown in FIG. This structural conversion is performed based on the conversion specification XML document shown in FIG. In order to specify the CSV compression format, a conversion specification that specifies an atypical CSV element is used in FIG. 27, and a conversion specification that specifies a conditional branch is used in FIG.

  In the example shown in FIG. 25, the type of record to be converted is one type (component). Depending on whether the attribute value of the attribute (type) is CPU, hard disk, or memory, Data items are different from each other. Element <product name> and element <model number> are common to all records, but the other elements are different from each other. That is, when type = “CPU”, element <CPU>, <clock>, <cache capacity>, and when type = “hard disk”, element <disk capacity>, <transfer speed>, <number of revolutions> >, When type = “memory”, the elements are <memory capacity>, <base clock>, and <power supply voltage>.

  The XML document having such a data structure can be structurally converted by the first and second methods. However, empty elements are output unnecessarily, and a conversion result as shown in FIG. 26 cannot be obtained. Further, the processing load becomes heavy.

  On the other hand, in the third method (part 1), for example, the conversion specification shown in FIG. 27 is used. In this conversion specification, the attribute format = “unfixed” is specified even in <csv-tag> that specifies the element name (CSV element name) of a new element that collects a plurality of non-key elements in the converted XML document, and the structure If all the non-key elements corresponding to each CSV element are collected in the conversion process and are all empty elements, the CSV elements are not output to the conversion XML document. This eliminates unnecessary output of empty elements, but it cannot cope with the problem of heavy processing load.

  On the other hand, in the third method (part 2), for example, the conversion specification shown in FIG. 28 is used. In this conversion specification, the conversion specification of the element corresponding to the attribute value is described separately for each attribute value. Then, the conversion specification to be used is selected according to the attribute value by specifying the switching condition (branch condition) by the when attribute of the <items> element. When an XSLT stylesheet is generated from this conversion specification, it can be switched according to the condition of <choose>-<when> <otherwise>.

  In the above-mentioned prior application, an atypical XML document can be handled, but the conversion can be performed at a time only when there is only one type of structure conversion target record. Further, the third method can cope with the case where data items are switched depending on conditions (partially different from each other), but this is a case where there is only one type of record but the data item (element) changes depending on the attribute. It corresponded to.

  However, XML can adopt a complicated atypical data structure such as nesting records in a non-table format or having multiple types of records. The method of the prior application cannot cope with an XML document having such a complicated structure across a plurality of types of records.

  An object of the present invention is to enable structure conversion such as CSV compression even for an XML document having a data structure having a record nesting or a plurality of types of records in a non-table format, for example. An object of the present invention is to provide a structured document structure conversion apparatus, structured document conversion method, program, and the like that can easily process an XML document having such a structure with application software.

  A first structure conversion apparatus of the present invention is a structure conversion apparatus for converting a structured document, and when the structured document has a data structure having a plurality of types of records in the same hierarchy, for each record type Correspondingly, a conversion specification definition means for defining a conversion operation for compressing and converting an element appearing in a record of the record type and defining a calling relationship between the plurality of conversion operations, and the respective definitions by the conversion specification definition means And a structure conversion means for converting the structure of a structured document having a data structure having a plurality of types of records in the same hierarchy.

  The structured document is, for example, an XML document and may have a data structure having a plurality of types of records in the same hierarchy. In the prior art (Patent Document 1) and the prior application, structural conversion can be performed without any problem for a structured document having only one type of record. In contrast, the first structure conversion apparatus of the present invention first creates a conversion operation definition (template) and its call definition for each record type. Each conversion operation definition itself may be basically the same as that of the prior art or the prior application. In the structure conversion process, the structure conversion is executed by calling a conversion operation definition corresponding to the record type for each record in the structured document subject to the structure conversion process.

  The second structure conversion apparatus of the present invention is a structure conversion apparatus for converting the structure of a structured document, wherein the structured document has a plurality of types of records and any record to be converted includes a child record. In the case of having a data structure, a conversion operation for compressing and converting an element appearing in a record of the record type corresponding to each record type is defined, and corresponding to a record type of an arbitrary record having the child record During the conversion operation definition, a conversion specification defining means for defining a calling relationship between conversion operations corresponding to the record type of the child record, and a plurality of types in the same hierarchy based on the respective definitions by the conversion specification defining means Structure conversion means for converting the structure of a structured document having a data structure having a record is provided.

  The above structured document is, for example, an XML document, and when there is a record nesting, when a record has a child record when structurally transforming an arbitrary record, the conventional technique (Patent Document 1) or a prior application can cope with it. There wasn't. On the other hand, the second structure conversion device of the present invention creates a conversion operation definition (template) for each record type as in the case of the first structure conversion device, but further includes records having child records. During the conversion operation definition corresponding to the record type, a call relationship between the conversion operations corresponding to the record types of the child records is defined. Thus, in the structure conversion process, even when the record to be processed has a child record, the structure conversion including the structure conversion of the child record can be performed based on the conversion operation definition of each of the process target record and the child record.

  The third structure conversion apparatus of the present invention is a structure conversion apparatus for converting the structure of a structured document, wherein the structured document has a data structure in which elements in the record change depending on conditions in each record. A conversion-like definition means for defining a conversion operation for compressing and converting an element appearing in the record corresponding to each condition, and defining a conditional branch to which of the plurality of conversion operation definitions is applied, and Structure conversion means for converting the structure of the structured document having the data structure based on the respective definitions by the conversion specification defining means is provided.

  The structured document is, for example, an XML document, and even for the same type of record, the elements appearing in the record may be variable for each record depending on some condition. On the other hand, in the third structural conversion device, a conversion operation definition is made corresponding to each condition, and a conditional branch for which one is applied is defined. Thus, in the structure conversion process, it is possible to refer to the condition in the record for each record and perform the structure conversion by the conversion operation definition corresponding to the condition.

  For example, the structured document is an XML document, the conversion specification definition is also defined in the XML document, and an XSLT stylesheet in which the conversion specification definition is described in the XSLT grammar is generated from the XML document defining the conversion specification. Style sheet generation means for performing the structure conversion using the XSLT style sheet.

  For example, the conversion specification definition is described in a notation system closely linked to the XSLT stylesheet grammar. Accordingly, an XSLT stylesheet can be automatically generated easily from an XML document in which the conversion specification is defined without having to describe the XSLT stylesheet one by one. Further, this enables the structure conversion process to be executed by a standard XSLT processor, and the structure conversion process can be applied to almost all XML document systems.

  Further, the above-mentioned program can be executed by causing the computer to read and execute the program from a computer-readable storage medium storing a program that causes the computer to perform the same control as the function performed by each configuration of the present invention described above. Can solve the problem. That is, the present invention can be configured as such a program itself, or can be configured as a recording medium (particularly a portable recording medium) on which the program is recorded.

  According to the structured document structure converting apparatus, structured document converting method, program, etc. of the present invention, for example, an XML document having a data structure having a record nesting or a plurality of types of records in a non-tabular format is also used. Structural conversion such as compression can be performed. Therefore, an XML document having such a structure can be easily processed by application software. In addition, the conversion specification for performing this structural conversion is expressed in a notation system closely linked to the XSLT stylesheet grammar. As a result, the process of generating the XSLT stylesheet from the conversion specification can be performed easily and smoothly.

Embodiments of the present invention will be described below with reference to the drawings.
FIG. 1 is a diagram showing a schematic flow and a configuration of an entire process for executing the structured document conversion method of this example by a computer or the like.

  In FIG. 1, a data structure conversion / inverse conversion mechanism 10 is a computer that realizes the structured document conversion method of this example, and includes a structure conversion unit 11, an inverse conversion unit 12, and an XSL conversion unit 13. The data structure conversion / inverse conversion mechanism 10 converts the input XML document 21 into a converted XML document 23 based on the input conversion specification XML document 22, and conversely selects from the plurality of converted XML documents 23. The extracted XML document 24, which is the converted XML document 23, is inversely converted into an XML document (result XML document 25) having the original structure.

The input XML document 21 is an XML document to be converted. The XML document is an example of a structured document and is not limited to this.
The conversion specification XML document 22 is an XML document that provides conversion specifications for conversion / inverse conversion. That is, it is extremely troublesome and time-consuming to create style sheets corresponding to each XML document, that is, XSL (Extensible Stylesheet Language) sheets (also referred to as XSLT style sheets) for various types of XML documents. It is. Therefore, in order to save this trouble, in this example, an XML document describing the specification for converting the data structure of the XML document (similar to the prior application etc.), that is, the conversion specification XML document 22 is created.

  Based on the conversion specification given by the conversion specification XML document 22, the structure conversion unit 11 converts the structure of the input XML document 21 into the conversion XML document 23, and the inverse conversion unit 12 converts the extracted XML document 24 into the result XML document 25. And reverse transform. In addition, a method that directly executes conversion / inverse conversion processing based on the conversion specification may be used as described above, but in particular, when converting a large amount of data, it is necessary to read and determine the conversion specification for each record. It becomes.

  On the other hand, the XSL conversion unit 13 instructs the conversion execution procedure based on the conversion specification XML document 22 and the converted XSL sheet generation XSL sheet 14 (automatic conversion style sheet in Patent Document 1). (Data structure conversion style sheet) and an inverse conversion XSL sheet 16 (inverse conversion style sheet) for instructing an inverse conversion execution procedure may be generated. Strictly speaking, the conversion XSL sheet generation XSL sheet 14 includes a conversion XSL sheet 15 generation and a reverse conversion XSL sheet 16 generation. . However, not limited to this example, the conversion / inverse conversion XSL sheets 15 and 16 may be created from the beginning.

  Then, the structure conversion unit 11 or the inverse conversion unit 12 may execute the conversion process or the inverse conversion process using the generated XSL sheets 15 or 16. Once the XSL sheets 15 and 16 are generated and then converted / inverted, an operation for reading and determining the conversion specifications for each record when converting a large amount of data is unnecessary, so the processing is executed at high speed. Will be able to.

  If the execution procedure of conversion / inverse conversion is given in the XSLT stylesheet in this way, the conversion / inverse conversion can be executed by a standard XSLT processor. In almost all kinds of XML document systems, this example Conversion / inverse conversion processing can be executed. In this case, the data structure conversion / inverse conversion mechanism 10 (structure conversion unit 11, inverse conversion unit 12) is actually realized by, for example, one standard XSLT processor (structured document conversion processor).

  The application software 30 is an application program that performs some kind of processing (eg, tag search), and the converted XML document 23 is structurally converted to have a data structure that can be easily processed by the application software 30.

  Here, the conversion specification XML document (XSL sheet) of the prior application had two types of templates. That is, the number of types of records subject to structure conversion processing is one, and consists of (i) a template for copying outside the target record and (ii) a template for performing structure conversion processing on the inside of the target record.

  On the other hand, in this method, a template (conversion operation definition) for performing structure conversion processing corresponding to a plurality of types of records is described in the conversion specification XML document 22, and an XSL sheet (XSLT style sheet) is written. Describe the relationship between these multiple templates according to the grammar. That is, the citation relationship between records is also described in the conversion specification XML document 22, thereby associating the templates. Further, as described above, an XSL sheet may be automatically generated from the conversion specification XML document 22. This XSL sheet has a plurality of templates, and these are associated with each other.

In this method, structure conversion (XML-XML conversion) can be performed even for an input XML document 21 having at least one data structure among the following three types of data structures (1) to (3).
(1) Data structure in which multiple element name (plural types) records appear in the upper hierarchy For the input XML document 21 having the above data structure, the following (a), As in (b).
(A) A template is generated in which a plurality of record element names to be processed are written together, the records are set as push processing, and the outside of the records is copied. Even if the same type of record is repeated any number of times by the push process, it can be dealt with by one designation.

For example, a plurality of record element names are combined in the conversion specification XML document 22 as follows.
<record> Target record element name 1 | Target record element name 2 | Target record element name 3 |… </ record>
Here, the push process is to write a template rule for each type of node that may appear, call an arbitrary template with <xsl: apply-templates>, and execute the process using that template. Yes (determine which template to call by pattern matching). It's called push processing because it looks like the processor is pushing someone out of the door while saying, "Who wants to handle this?"
(B) A template for processing various records is generated as a template for matching processing.

  That is, the template called by the push process is set as follows, for example, in the conversion XSL sheet 15 generated from the conversion specification XML document 22, for example.

<xsl: template match = “Target record element name”>
Convert operation
</ xsl: template>
Note that the conversion operation (CSV compression) itself may use the above-described conventional technique or the technique of the prior application, and no specific example is shown here.

By doing so, the processing template corresponding to the record element name is automatically applied to each processing target record.
(2) Data structure in which a record has child records For an input XML document 21 having the above data structure, a conversion XSL sheet 15 generated from a conversion specification XML document 22 created by the second method of this example is For example, it has the following contents.

In the template that processes the various records in (b) above,
<xsl: apply-template select = ”Target child record element name” />
The child record processing template is called by push processing.

Of course, a processing template for the corresponding child record element name is also created. By performing the push process, it is possible to deal with a child record that is repeated an arbitrary number of times by one designation.
(3) Data structure in which most of the data items of the record are replaced (conditional branch)
For the input XML document 21 having the above data structure, the converted XSL sheet 15 generated from the conversion specification XML document 22 created by the third method of this example has the following contents, for example.

In the record processing template, the pull processing for quoting a template that matches the condition only once is performed as follows.
<xsl: call-template name = ”Quoted template name” />
Of course, a corresponding named template (above (b)) is also created.

  Here, the pull process is a process that specifies a name and calls a specific template.Instead of pushing the node out of the door (push) and picking it up by another template, it pulls the node (pull) itself. Since it is processed, it is called pull processing.

The push process and the pull process are introduced on, for example, the following IBM website.
http://www-6.ibm.com/jp/developerworks/xml/020726/j_x-xdpshpul.html
This homepage introduces push processing and pull processing as push style and pull style, and discloses that the push style is used when the appearance order and the number of repetitions are unknown.

  As described with reference to FIG. 22, the non-standard XML document to be converted in the prior application has non-key elements that may appear in the record. It was uncertain what non-key elements appeared.

  On the other hand, the XML document (input XML document 21) to be converted by this method has one or more data structures of (1) to (3) above. For example, FIG. 2 shows an example of an XML document having the data structures (1) and (2). FIG. 6 shows an example of an XML document having the data structure (3).

Although the XML document having these data structures cannot be converted by the method of the prior application, the structure can be converted by this method.
Hereinafter, an example will be described with reference to FIG.

  The input XML document 21 in the example of FIG. 2 can be subjected to structure conversion (XML-XML conversion) according to the structure conversion definition described in the conversion specification XML document 22 shown in FIG. 3, for example. For example, a converted XML document 23 shown in FIG. 5 is obtained. Alternatively, the conversion XSL sheet 15 shown in FIG. 4 may be automatically generated from the conversion specification XML document 22 shown in FIG. 3, and the above conversion may be performed using this. Although not shown here, the reverse conversion process and the reverse conversion XSL sheet 16 can be similarly performed.

  The input XML document 21 shown in FIG. 2 has the data structures (1) and (2) described above, that is, the record element <research> has child record elements <patent>, and a plurality of types of record elements <research> in the same hierarchy. , <References>.

  The feature of this method is the relationship definition between records and the branching, calling, etc., and the conversion process for each record is fixed in each record in the examples shown in FIGS. Conventional techniques (Patent Document 1, etc.) may be used, but not illustrated, but if each record is atypical, use any one of the first to third techniques of the prior application. That's fine.

As described above, in the following description, the conversion process in each record is not particularly described, but the characteristics of the present method are described.
First, as shown in FIG. 27, in the conversion XML document of the prior application, there is one <items> element, and only one record conversion operation can be specified (that is, <record> part </ record> only).

  In contrast, in the conversion specification XML document 22 of this method shown in FIG. 3, first, a plurality of types of record element names “investigation” and “reference document” are specified together with <record> elements (as shown in the figure). Survey | references), the conversion process of each record is described as an <items record = “survey” or “references”> element.

Further, regarding the child record element <patent> of the record element <survey>, it is described that the child record element <patent> appears at the end of the conversion process description relating to the record element <survey>. In other words, in the example shown,
<items record = ”Survey”>
<item ctag = ”-RECORD”> Patent </ item>
Is specified, the ctag attribute of the item element specifies branching to a child record, and the element name of the child record is specified by element content. The conversion process of the child record that receives this is described as an <items record = “patent”> element.

  The <items record = “search” or “patent” or “reference”> element described above is a conversion processing template corresponding to each record type, and the contents of this template itself are described above. As described above, any method of the prior application may be used. In addition, csv_tag and ctag in FIG. 3 etc. are merging_tag and mtag in the prior application.

Here, in the XML document conversion process or the XSL sheet generation process, the process corresponding to the above description is performed in advance as in the prior application.
That is, for example, when the structure conversion unit 11 performs structure conversion (XML-XML conversion) using the conversion specification XML document 22 directly, it recognizes record types that may appear due to the combination of the record names, Thereafter, the input XML document 21 is sequentially read, and each time a target record name is encountered, a corresponding conversion processing template is selected, and conversion processing is performed using this template. In addition, according to <item ctag = “-RECORD”> record name </ item> in the conversion specification, a record corresponding to the record type of the record name is used to convert the record with the record name.

  On the other hand, regarding the XSL sheet generation processing, the XSL conversion unit 13 has a processing function for generating the conversion XSL sheet shown in FIG. 4 (and the reverse conversion XSL sheet not shown here) from the conversion specification XML document shown in FIG. Have.

In other words, the XSL conversion unit 13 uses the conversion specification XML document 22 to
<record> Survey | References </ record>
If there is, this is shown in FIG.
<xsl: when test = ”descendant-or-self :: * {Research | References}”>
Are replaced with a series of descriptions (for example, part B in FIG. 12; described later).

  According to this series of descriptions, when a part other than the <investigation> and <references> elements is copied and the <investigation> and <references> elements are found, a process for calling the template by push processing is performed.

Also, the XSL conversion unit 13 replaces the <items record = “record name”> element in the conversion specification XML document 22 with <xsl: template match = “record name”>.
As a result, the record element <investigation> and <reference material> are processed with the template of <xsl: template match = “investigation” or “reference material”>, respectively. Since the contents of each template (CSV compression operation) can be converted into the XSL sheet format of the prior art or the prior application as described above, a complete version that is omitted here but not shown later is shown. Shall.

  As described above, the push process branch is the <record> element of the conversion specification, the template of each record process is the element in <items record = “investigation” or “reference material”>, and the conversion XSL sheet 15 in association with it. Is generated.

  Also, the above <item ctag = ”-RECORD”> patent </ item> in the conversion specification XML document 22 is included in the <xsl: template match = “search”> template as shown in FIG. : apply-template select = ”patent”> element.

As a result, when a child record element <patent> is found in the input XML document 21, the process branches by a push process and is processed with a template of <xsl: template match = “patent”>.
By performing the structure conversion of the XML document using the conversion specification of FIG. 3 described above, the input XML document 21 shown in FIG. 2 is converted into the conversion XML document 23 shown in FIG. Note that the content of the converted XML document is not particularly relevant to the description of this method, and is therefore omitted.

  Next, processing in the case where the input XML document 21 has the data structure (3) described above, that is, in the case of a data structure in which elements in a record are switched depending on conditions in the record will be described.

FIG. 6 shows an example of the input XML document 21 having the data structure (3).
Further, for the input XML document 21 in the example of FIG. 6, for example, the conversion specification XML document 22 shown in FIG. 7 can be used to convert the structure of the XML document. As a result, the converted XML document shown in FIG. 23 is obtained. Alternatively, the conversion XSL sheet 15 (also the reverse conversion XSL sheet 16 not shown) shown in FIG. 8 may be automatically generated from the conversion specification XML document 22 shown in FIG. 7, and the conversion may be performed using this.

  The input XML document 21 shown in FIG. 6 has one type of record (<part> only), but the element in the record changes for each record according to the element content of the element <type>. A type of XML document. Here, it is assumed that there are a CPU, a memory, an HDD, and the like as <type> element contents, and the element arrangement in the record changes depending on which of them is. In the illustrated example, if the element content of <type> is CPU, the element list in the record is <model number> <chip> <clock>, and if it is memory, <model number> <capacity> < In the case of an HDD, <model number> <capacity> <transfer rate>.

Thus, in the conversion specification shown in FIG. 7, first, a conditional branch is specified depending on what the element content of the element <type> is. For example, taking a CPU as an example,
<item ctag = ”-BRANCH” case = “Type = 'CPU'”> process-a </ item>
Specifies a conditional branch with ctag, gives a condition with the case attribute, and if the judgment result is true, specifies to branch to a template with the name process-a. As the condition of the case attribute, it is checked whether or not the element content of the element with the element name “kind” is “CPU”. The same applies to the memory and HDD.

  The conversion process that receives this conditional branch is described as a template of <items name = “process-a”> element. As already described, the contents of each template are not specifically described.

  The XSL conversion unit 13 generates the converted XSL sheet shown in FIG. 8 from the conversion specification XML document shown in FIG. In this generation process, the conditional branch specification in the above conversion specification is the conditional branch <xsl: choose> <xsl: when test = in <xsl: template match = ”component”> which is a template for processing <component> It is generated as “XX”> statement.

For example, it is an example on the above conversion specification
<item ctag = ”-BRANCH” case = “Type = 'CPU'”> process-a </ item>
In FIG.
<xsl: when test = “type =“ CPU ””><xsl: call-template name = ”process-a” />
And written in the XSL sheet.

  This is such that the condition of the case attribute in the conversion specification is directly included in the test condition of the when statement of the XSL sheet. Therefore, even complicated conditions can be dealt with widely in the XSL transformation grammar rules, and the implementation of the generation of the transformation XSL sheet can be facilitated. Furthermore, since conditional branching is a one-time process, it is called with a call-template statement as a pull process. The conversion process that receives this conditional branch is described in a sequence of <items name = “process-a”> elements.

  In the prior patent (No. 3), when corresponding to a plurality of types of records, for example, a plurality of input XML documents each having a record of a different record type are prepared, and each conversion specification XML document corresponding to each of the input XML documents is prepared. Will be created. Then, for each of these conversion specification XML documents, processing corresponding to each condition is created internally (that is, conditional branching and processing corresponding to each condition are embedded in one template). On the other hand, in this method, since a structure conversion processing template corresponding to each condition is prepared and the branch processing is performed as the template pull processing, for example, a combination of the conversion specifications shown in FIG. Even an XML document having a data structure having a kind of record can be handled. Also, after creating a structure conversion processing template according to each of the above conditions for a certain type of record, and when describing the structure conversion processing for another type of record, if there is the same condition processing, the already created structure Since the conversion processing template can be cited, the labor of creating the conversion specification can be reduced and the creation can be made flexibly.

Hereinafter, a process in which the XSL conversion unit 13 generates the conversion XSL sheet 15 and the reverse conversion XSL sheet 16 from the conversion specification XML document 22 will be described.
FIG. 10 shows a flowchart of processing for generating a conversion XSL sheet based on the conversion specification. FIG. 11 is a detailed flowchart of step S16 in FIG. The processes shown in FIGS. 10 and 11 mainly show the process of associating templates, and are common to conversion and inverse conversion.

  4 and 8 show the outline of the XSL sheet. The complete version is shown in FIGS. 12, 15 and 16. FIG. Further, the respective inverse conversion sheets are shown in FIGS. 13 and 14, and FIGS. 17 and 18. That is, FIG. 12 shows a conversion XSL sheet 15 generated based on the conversion specification of FIG. 3, and FIGS. 13 and 14 show an inverse conversion XSL sheet 16 generated based on the conversion specification of FIG. FIGS. 13 and 14 only show one inverse transform XSL sheet divided into two. Similarly, FIGS. 15 and 16 are converted XSL sheets 15 generated based on the conversion specifications of FIG. 7, and FIGS. 17 and 18 are inverse converted XSL sheets 16 generated based on the conversion specifications of FIG. . FIGS. 15 and 16, and FIGS. 17 and 18 each show only one reverse conversion XSL sheet / inverse conversion XSL sheet in two parts.

Hereinafter, the processing illustrated in FIGS. 10 and 11 will be described with reference to FIGS. 12 to 18 as necessary.
In FIG. 10, the XSL conversion unit 13 reads the conversion specification XML document 22 (step S11), and first generates fixed parts such as an XML declaration sentence and a start tag of the conversion XSL sheet 15 (step S12). Thereby, the part A shown in FIGS. 12, 13, 15, and 17 is created.

  Next, the element content of the <record> element in the conversion specification is set as a target record group (step S13), a part other than the target record element group is copied, and a template for pushing the target record group is created (step S14). ). If there are multiple types of record elements in the upper hierarchy, they are written together in the <record> element as shown in Fig. 3. By using this as a template condition, the template that pushes the record group can do.

  By the process of step S14, the part B shown in FIGS. 12 and 13 is created for the conversion specification of FIG. Since this has the same contents for both conversion and inverse conversion, the same symbol (B) is given. Further, by the process of step S14, a portion C shown in FIGS. 15 and 17 is created for the conversion specification of FIG. Note that B and C differ only in that a plurality of records are written together in B ([Survey | References]), and only [Survey] is in C.

  It should be noted that most of A and B and C may be held in advance as a template (converted XSL sheet generation XSL sheet 14). That is, only the contents (record names) of [] in B and C may be changed according to the conversion specification.

  Then, the process of FIG. 11 is repeatedly performed for each <items> element in step S16, and when the process of FIG. 11 is performed for all <items> elements (step S15, YES), the conversion / reverse is finally performed. An end tag is described in the converted XSL sheet (step S17), and the XSL sheet generation process is terminated.

Here, the parts A and B will be described. Here, A and B will be described with reference to FIG.
First, the description of A ′ in the portion A means that “all defined templates are applied to the root node”. All of the templates are a total of four templates: B template and templates (H, I, and E parts in the figure) related to each of research, patent, and reference.

Next, the part B will be described. First, as is well known, there is a notation composed of an axis, a node test, and a predicate in the XPath specification. For example, a notation called “axis :: node test [expression]” is known. According to this notation, “descendant-or-self :: * [investigation | references]” in Part B
The part of is described.

  The descendant-or-self, which is a kind of axis notation, is the current node, the child node of the current node, and the child node of the child node (grandchild node; hereinafter referred to collectively as the descendant node). All the descendant nodes are selected recursively. *, Which is a kind of node test notation, indicates all child nodes of the axis. And the OR condition of the survey and the reference is described as an expression.

Accordingly, the processing by the portion B is as follows.
(a) If there is <investigation> or <reference> in its own node or its descendant nodes, after copying the current element with attributes, all the above templates are applied to all child nodes.

(b) Otherwise, the local node and its descendant nodes are copied.
When the processing of part B is applied to the example of FIG. 2, the target node is first <Technical Survey>, which is not the <Survey> nor <Reference>, but its child node <Survey> and <References> fall under the above (a), and after copying the current element with attributes, each of its child nodes <Theme>, <Survey>, and <References> Apply all templates recursively. Note that this is applied to all records in all hierarchies, regardless of the number of hierarchies.

  First, if you try to apply all templates B, H, I, E to <theme>, it will not match the application conditions of templates H, I, E, so only template B will be applied. Since it corresponds to (b) above, the local node and its descendant nodes are copied (however, since <theme> has no descendants, only the local node).

  Next, if all templates B, H, I, and E are applied to <Survey>, it matches the application condition of template H, and if template B is applied, it corresponds to the above (a). The former is a specific element name designation, whereas the latter is a wild card designation, so the template application priority is higher in the former. Therefore, only the template H is applied.

At that time, since there is a push process to the template I in the template H, the template I is applied to each child node <patent>.
Finally, with respect to <reference document>, only template E is applied for the same reason as the above <investigation>.

Hereinafter, the process illustrated in FIG. 11 will be described.
The process of FIG. 11 starts with <xsl: template match = “record attribute value”> or <xsl: template name = “name attribute value”>, depending on whether the attribute of the <items> element is record or name. A starting template is created (steps S21 to S24). These correspond to record push processing and conditional branch pull processing. Although not shown, in the case of an <items> element having no attribute, a template starting with <xsl: template match = “element content of the record element”> is created.

Next, each <item> element in the <items> is scanned (step S25), and for each <item> element,
According to any of the conditions of steps S27, S29, S31, S33, and S34, the process of any of steps S28, S30, S32, S35, and S36 is executed.

Here, the function of the <item> element is identified by the ctag attribute value.
Thus, if the ctag attribute value is “-ORG” (step S27, YES), it is processed as a key element (step S28). The first and second <item> elements in the first <items> element in FIG. 7 correspond to this, and the portion D shown in FIGS. 16 and 17 is generated.

  Alternatively, if the ctag attribute value matches the element content (CSV element name) of <csv_tag> (step S29, YES), it is processed as a non-key element (step S30). The processing of the non-key element or the processing of the key element is the same as that of the prior application, and thus description thereof is omitted here.

  As described above, the processes of FIGS. 10 and 11 are common to the conversion and the reverse conversion, but the process of the non-key element changes in the conversion and the reverse conversion. This is also explained in the prior application, but for simplicity, the key element is copied as it is in the original document in the case of conversion, and the same element as the original document is reversed in the converted document in the case of reverse conversion. Just copy it to the converted document. On the other hand, in the case of a non-key element, in the case of conversion, an operation is performed to combine the contents of the <item> element in the ctag attribute value (CSV element) in CSV format. In the case of reverse conversion, the element content of the CSV element of the ctag attribute value Will be inserted into the content of the <item> element.

For example, taking the conversion specification of record C in FIG. 3 as an example, the part E shown in FIG. 12 is generated for conversion. That is, the element content of each element of the tag title name, source, and remarks in the input XML document 21 (XML grammar, XML commentary company, issued by 200306) is associated with the new element whose CSV element name is “document csv”. The following statements that are connected in the CSV format and output to the converted XML document 23 are generated.
<Reference csv><xsl: value-of select = ”concat (title, ',' source, ','remarks)"/></ reference csv>
On the other hand, with respect to the inverse transformation, as shown by F in FIG. 14, an operation for separating the contents of each element connected in the CSV format is performed using select = “substring-before / after... A statement will be generated.

  If the ctag attribute value is “-RECORD” (step S31), since a child record is specified, <xsl: apply-templates select = ”<item> element content” / Generate>. That is, a command statement for calling the processing template of the child record by the push processing is generated (step S32).

  Alternatively, if the ctag attribute value is “-BRANCH” (step S33, YES), a conditional branch is designated. Therefore, the case attribute value of this <item> element is subsequently examined. If the case attribute value is not “otherwise” (step S34, NO), the process of step S35 is executed.

  In the process of step S35, first, it is checked whether or not the <item> element to be processed this time is the first <item> element in a series of conditional branch <item> elements. Only in the first case, <xsl: Choose> statement. Then, regardless of whether it is the first or not, the following conditional judgment and pull processing statements are generated.

<xsl: when test = ”case attribute value”>
<xsl: call-template name = ”<item> element content” /></ xsl: when>
On the other hand, if the case attribute value is “otherwise” (step S34, YES),
<xsl: otherwise>
<xsl: call-template name = ”<item> element content” />
</ xsl: otherwise>
Then, </ xsl: choose> is continued and a series of conditional branch statements is terminated.

As a result, a series of four conditional branch statements shown in the conversion specification of FIG. 7 is generated as an <xsl: choose> statement of the G portion shown in FIGS.
When the above processing is executed for each <item> element and processing of all <item> elements in <items> is completed (step S26, YES), a template closing tag </ xsl: template> is described. The process in FIG. 11 ends.

  As described above, according to this method, it is possible to perform structure conversion such as CSV compression even for an XML document having a data structure having a record nesting or a plurality of types of records in a non-table format, for example. become. Therefore, an XML document having such a structure can be easily processed by application software. In addition, the conversion specification for performing this structural conversion is expressed in a notation system closely linked to the XSLT stylesheet grammar. As a result, the process of generating the XSLT stylesheet from the conversion specification can be performed easily and smoothly.

FIG. 19 is a diagram illustrating an example of a hardware configuration of a computer that implements the structured document conversion method according to the present embodiment.
A computer 100 shown in the figure includes a CPU 101, a memory 102, an external storage device 105, a medium driving device 106, and the like, which are connected to a bus 108. Furthermore, the structure which has the network connection apparatus 107 may be sufficient. The configuration shown in the figure is an example, and the present invention is not limited to this.

The CPU 101 is a central processing unit that controls the entire computer 100.
The memory 102 is a memory such as a RAM that temporarily stores a program or data stored in the external storage device 105 (or the portable recording medium 109) during program execution, data update, or the like. The CPU 101 uses the program / data read out to the memory 102, and functions of the structure conversion unit 11, the inverse conversion unit 12, and the XSL conversion unit 13 in FIG. 2 described above (with regard to the XSL conversion unit 13, FIGS. The flowchart processing shown in FIG.

  The external storage device 105 is, for example, a magnetic disk device, an optical disk device, a magneto-optical disk device, or the like, and stores programs / data and the like for realizing the various functions of the present invention. The data includes, for example, the converted XSL sheet generation XSL sheet 14, the converted XSL sheet 15, and the inverse converted XSL sheet 16 shown in FIG. Further, when data such as the input XML document 21, conversion specification XML document 22, and extracted XML document 24 are input from the outside, these are temporarily stored.

  The medium driving device 106 reads a program / data stored in the portable recording medium 109. The portable recording medium 109 is, for example, an FD (flexible disk), a CD-ROM, a DVD, a magneto-optical disk, or the like.

The network connection device 107 is configured to be connected to a network and to transmit / receive programs / data and the like to / from an external information processing device.
FIG. 20 is a diagram illustrating an example of a recording medium in which the above-described program is recorded and download.

  As shown in the figure, the information may be read from the portable recording medium 109 storing the program / data for realizing the functions of the present invention to the information processing apparatus 100 side, stored in the memory 102, and executed. The program / data downloads the program / data stored in the storage unit 111 of the external server 110 via a network (such as the Internet) connected by the network connection device 107. Also good.

  The present invention is not limited to the apparatus / method, and can be configured as a recording medium (such as the portable recording medium 109) storing the program / data, or as the program itself.

(Supplementary note 1) A structure conversion device for converting the structure of a structured document,
When the structured document has a data structure having a plurality of types of records in the same hierarchy, a conversion operation for compressing and converting an element appearing in the record of the record type is defined corresponding to each record type and the plurality A conversion specification defining means for defining the calling relationship between the conversion operations of
Structure conversion means for converting the structure of a structured document having a data structure having a plurality of types of records in the same hierarchy, based on each definition by the conversion specification definition means;
A structure conversion device comprising:

(Supplementary note 2) A structure conversion device for converting a structured document into a structure,
When the structured document has a plurality of types of records and an arbitrary record to be converted has a data structure having a child record, the elements appearing in the record type corresponding to each record type are compressed. A conversion specification definition that defines a conversion operation to be converted and defines a calling relationship between conversion operations corresponding to the record type of the child record during the conversion operation definition corresponding to the record type of the arbitrary record having the child record Means,
Structure conversion means for converting the structure of a structured document having a data structure having a plurality of types of records in the same hierarchy, based on each definition by the conversion specification definition means;
A structure conversion device comprising:

(Supplementary note 3) A structure conversion device for converting the structure of a structured document,
When the structured document has a data structure in which elements in the record change depending on conditions in each record, a conversion operation for compressing and converting elements appearing in the record corresponding to the conditions is defined and the plural A conversion specification defining means for defining a conditional branch of which one of the conversion operation definitions is applied,
A structure conversion means for converting the structure of the structured document having the data structure based on the respective definitions by the conversion specification definition means;
A structure conversion device comprising:

(Additional remark 4) The definition of the conversion operation which compresses and converts the element in the record is a key element which defines a plurality of new elements in the converted structured document and is subject to data processing for each element in the record. Defining whether or not each non-key element that is an element other than the key element is allocated to the plurality of new elements,
According to this definition, the structure conversion means describes each element in the converted structured document as it is in the order in which the elements appear in the record. For each non-key element, the element content is described. The structure conversion apparatus according to any one of appendix 1 to appendix 3, wherein a summary of the corresponding new elements in CSV format is described in the structured document after conversion as element contents of each new element .

(Supplementary Note 5) The structured document is an XML document, the conversion specification definition is also defined in the XML document,
Style sheet generating means for generating an XSLT stylesheet in which the conversion specification definition is described in an XSLT grammar from an XML document defining the conversion specification;
5. The structure conversion apparatus according to any one of appendix 1 to appendix 4, wherein the structure conversion means is an XSLT processor that performs the structure conversion using the XSLT stylesheet.

  (Supplementary Note 6) When the structured document has a data structure having a plurality of types of records in the same hierarchy, the call definition of the plurality of conversion operation definitions is to any of the plurality of conversion operation definitions in the XSLT stylesheet. 6. The structure conversion device according to appendix 5, wherein the structure conversion device is described as a push process.

  (Supplementary note 7) When the structured document has a plurality of types of records and any record to be converted has a data structure having a child record, the call definition of the conversion operation definition corresponding to the record type of the child record is The structure conversion apparatus according to appendix 5, wherein the structure conversion apparatus is described as a push process to the conversion operation definition of the child record in the XSLT stylesheet.

  (Supplementary note 8) The conditional branch definition in the case where the structured document has a data structure in which elements in the record change depending on conditions in each record is described as pull processing in the XSLT stylesheet. The structure conversion apparatus according to appendix 5.

(Supplementary note 9) A structure conversion method for converting the structure of a structured document,
When the structured document has a data structure having a plurality of types of records in the same hierarchy, a conversion operation for compressing and converting an element appearing in the record of the record type is defined corresponding to each record type and the plurality A structure conversion characterized in that a structured document having a data structure having a plurality of types of records in the same hierarchy is structurally converted based on a conversion specification XML document or XSLT stylesheet defining a calling relationship between conversion operations of Method.

(Supplementary Note 10) A structure conversion method for converting the structure of a structured document,
When the structured document has a plurality of types of records and an arbitrary record to be converted has a data structure having a child record, the elements appearing in the record type corresponding to each record type are compressed. A conversion specification XML defining a conversion operation to be converted and defining a calling relationship between conversion operations corresponding to the record type of the child record during the conversion operation definition corresponding to the record type of the arbitrary record having the child record A structure conversion method, comprising: converting a structured document having a data structure having a plurality of types of records in the same hierarchy based on a document or an XSLT stylesheet.

(Supplementary note 11) A structure conversion method for converting the structure of a structured document,
When the structured document has a data structure in which elements in the record change depending on conditions in each record, a conversion operation for compressing and converting elements appearing in the record corresponding to the conditions is defined and the plural A structure conversion method characterized by structurally converting a structured document having the data structure based on a conversion specification XML document or an XSLT stylesheet that defines a conditional branch of which one of the conversion operation definitions is applied.

(Supplementary note 12)
If the structured document to be converted has a data structure with multiple types of records in the same hierarchy, define a conversion operation that compresses and converts elements that appear in records of that record type for each record type And a structural conversion of a structured document having a data structure having a plurality of types of records in the same hierarchy, based on a conversion specification XML document or XSLT stylesheet that defines a calling relationship between the plurality of conversion operations.
A program to realize

(Supplementary note 13)
If the structured document to be converted has multiple types of records and any record to be converted has a data structure with child records, it will appear in the record type corresponding to each record type. In addition to defining a conversion operation that compresses and converts elements, a call relationship between conversion operations corresponding to the record type of the child record was defined during the conversion operation definition corresponding to the record type of the arbitrary record having the child record. A function of structurally converting a structured document having a data structure having a plurality of types of records in the same hierarchy, based on a conversion specification XML document or an XSLT stylesheet;
A program to realize

(Supplementary note 14)
When the structured document to be converted has a data structure in which the elements in the record change depending on the conditions in each record, a conversion operation for compressing and converting elements appearing in the record corresponding to the conditions is defined. And a function for structural conversion of the structured document having the data structure based on a conversion specification XML document or XSLT stylesheet that defines a conditional branch of which of the plurality of conversion operation definitions is applied,
A program to realize

(Supplementary note 15)
If the structured document to be converted has a data structure with multiple types of records in the same hierarchy, define a conversion operation that compresses and converts elements that appear in records of that record type for each record type And a structural conversion of a structured document having a data structure having a plurality of types of records in the same hierarchy, based on a conversion specification XML document or XSLT stylesheet that defines a calling relationship between the plurality of conversion operations.
The computer-readable recording medium which recorded the program which implement | achieves.

(Supplementary note 16)
If the structured document to be converted has multiple types of records and any record to be converted has a data structure with child records, it will appear in the record type corresponding to each record type. In addition to defining a conversion operation that compresses and converts elements, a call relationship between conversion operations corresponding to the record type of the child record was defined during the conversion operation definition corresponding to the record type of the arbitrary record having the child record. A function of structurally converting a structured document having a data structure having a plurality of types of records in the same hierarchy, based on a conversion specification XML document or an XSLT stylesheet;
The computer-readable recording medium which recorded the program which implement | achieves.

(Supplementary note 17)
When the structured document to be converted has a data structure in which the elements in the record change depending on the conditions in each record, a conversion operation for compressing and converting elements appearing in the record corresponding to the conditions is defined. And a function for structural conversion of the structured document having the data structure based on a conversion specification XML document or XSLT stylesheet that defines a conditional branch of which of the plurality of conversion operation definitions is applied,
The computer-readable recording medium which recorded the program which implement | achieves.

It is a figure which shows the schematic flow of the whole process which performs the structured document conversion method of this example by a computer etc., and its structure. It is an example (part 1) of an atypical XML document to be converted by this method. It is an example of the conversion specification XML document corresponding to the XML document of FIG. FIG. 4 is an outline of a conversion XSL sheet generated from the conversion specification XML document of FIG. 3. FIG. 3 is an example of a converted XML document obtained by structurally converting the XML document in FIG. 2. It is an example (the 2) of the atypical XML document made into the conversion object by this method. It is an example of the conversion specification XML document corresponding to the XML document of FIG. FIG. 8 is an outline of a conversion XSL sheet generated from the conversion specification XML document of FIG. 7. FIG. 7 is an example of a converted XML document obtained by structurally converting the XML document in FIG. 6. It is a process flowchart figure of an XSL conversion part. It is a detailed flowchart figure of the process of step S16 of FIG. It is an example of the conversion XSL sheet | seat produced | generated based on the conversion specification of FIG. It is an inverse transformation XSL sheet (1/2) produced | generated based on the conversion specification of FIG. It is an inverse transformation XSL sheet (2/2) produced | generated based on the conversion specification of FIG. It is the conversion XSL sheet | seat (1/2) produced | generated based on the conversion specification of FIG. It is the conversion XSL sheet | seat (2/2) produced | generated based on the conversion specification of FIG. It is an inverse transformation XSL sheet (1/2) produced | generated based on the conversion specification of FIG. 8 is an inversely converted XSL sheet (2/2) generated based on the conversion specification of FIG. It is a computer hardware block diagram. It is a figure which shows the recording medium which recorded the program, and the download of a program. It is an example of conventional structure conversion, (a) is a conversion source, (b) is an XML document after conversion. It is an example (the 1) of the XML document of the structure conversion object in a prior application. 23 is an example of a converted XML document obtained by structurally converting the XML document in FIG. 22 by the first method of the prior application. 23 is an example of a converted XML document obtained by structurally converting the XML document of FIG. 22 by the second method of the prior application. It is an example (the 2) of the XML document of the structure conversion object in a prior application. FIG. 26 is an example of a converted XML document obtained by structurally converting the XML document of FIG. 25. FIG. It is the conversion specification XML document by the 3rd method (the 1) of a prior application. It is the conversion XML document by the 3rd method (the 2) of a prior application.

Explanation of symbols

10 Data structure conversion / inverse conversion mechanism 11 Structure conversion unit 12 Inverse conversion unit 13 XSL conversion unit 14 Conversion XSL sheet generation XSL sheet 15 Conversion XSL sheet 16 Reverse conversion XSL sheet 21 Input XML document 22 Conversion specification XML document 23 Conversion XML document 100 Computer 101 CPU
102 Memory 105 External Storage Device 106 Medium Drive Device 107 Network Connection Device 108 Bus 109 Portable Recording Medium 110 External Server 111 Storage Unit

Claims (4)

A structure conversion device for converting the structure of a structured document,
When the structured document has a data structure having a plurality of types of records in the same hierarchy, a template for performing structure conversion processing is described for each record type and a citation relationship between records is described. Storage means for storing the conversion specification document,
Template a structured document having a data structure having a plurality of types of records before Symbol same hierarchy, copy to the portion other than the record element of the structure converted, for target record group, described in the conversion specification document A template corresponding to each record of the target record group is determined by pattern matching, and using the determined template, structure conversion means for converting the structure,
A structure conversion device comprising: A structure conversion device for converting the structure of a structured document,
When the structured document has a plurality of types of records and an arbitrary record to be converted has a data structure having a child record, a template for performing each structure conversion process is described corresponding to each record type. Storage means for storing a conversion specification document in which a citation relationship between records corresponding to the record type of the child record is described in a template corresponding to the record type of an arbitrary record having the child record;
Template a structured document having a data structure having a plurality of types of records before Symbol same hierarchy, copy to the portion other than the record element of the structure converted, for target record group, described in the conversion specification document A template corresponding to each record of the target record group is determined by pattern matching, and using the determined template, structure conversion means for converting the structure,
A structure conversion device comprising: The structured document and the conversion specification document are XML documents,
Style sheet generating means for generating an XSLT stylesheet describing the conversion specification document in XSLT grammar from the conversion specification document;
The structure conversion apparatus according to claim 1, wherein the structure conversion unit is an XSLT processor that performs the structure conversion using the XSLT stylesheet. On the computer,
If the structured document to be converted has a data structure with multiple types of records in the same hierarchy, templates for performing structure conversion processing are described for each record type and citation relationships between records A structured document having a data structure having a plurality of types of records in the same hierarchy, based on a conversion specification XML document or XSLT stylesheet in which is described, for portions other than the record element to be converted, For the target record group, a template corresponding to the record written by the target record group is determined from the templates described in the conversion specification XML document or the XSLT stylesheet, and the determined template is used. Ability to convert structures,
A program to realize

Images