JP2008084341A - Structured document compressing method, compressing device, and computer-readable recording medium recording structured document compressing program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To improve compressibility of a structured document by compressing a tag part without impairing the characteristics of the structured document. <P>SOLUTION: A structured document compressing device comprises a document achievement value analyzing part 20 analyzing description in a tag of a document achievement value forming the structured document; a tag dictionary preparing part 80 preparing a tag dictionary which associates a character string described in the tag of the document achievement value, with a shortened character string shorter than the character string and allowing the character string to be specified according to an analyzed result by the document achievement value analyzing part 20; and a document achievement value character string substituting part 41 substituting the shortened character string corresponding to the character string, for the character strings described in the tag of the document achievement value using the tag dictionary prepared by the tag dictionary preparing part 80. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a method and apparatus for compressing structured documents such as HTML (Hyper Text Markup Language), SGML (Standard Generalized Markup Language), and XML (Extensible Markup Language), and a computer-readable recording program. The present invention relates to various recording media.

  In recent years, with the spread of computers, the Internet, and intranets, electronic documents including various types of data such as documents, software, numerical values, and image data are increasing. When the data amount of the electronic document is large, the redundant portion is omitted from the electronic document data and the data is compressed to reduce the data amount of the electronic document stored in the memory, or the transmission time of the electronic document. Or shorten it.

  In the following description, in accordance with the name used in information theory, one word unit of data is referred to as “character”, and an arbitrary number of connected characters is referred to as “character string”. In addition, so-called universal encoding technology capable of compressing any kind of data on the assumption that data consists of characters has been widely studied and spread.

  On the other hand, in electronic documents, structured documents with tags are widely used. When such a structured document is used, processing is performed separately for “character data” and “markup”. Here, “mark” is a general term for tags and other structured information. Specifically, a start tag, an end tag, an empty element tag, an entity reference, a character reference, a comment, a CDATA section delimiter, a document Points to type declarations, processing instructions, etc.

  Typical examples of structured documents include SGML (Standard Generalized Markup Language) for large-scale storage databases, HTML (Hyper Text Markup Language) with a simple configuration for WWW (World Wide Web), and SGML for the Internet. There is a simplified XML (eXtensible Markup Language).

SGML is used for application to viewers of large-scale stored document databases in public offices and companies, database search, parallel development of products, publication (CAD, e-book, database) and data exchange It is also used as an intermediate language for In addition, HTML is widespread worldwide with WWW.
XML has recently received particular attention as a supplement to HTML. This XML is being used not only for handling documents on the Internet but also as an intermediary for communication of various information devices such as mobile phones and car navigation systems.

  An XML document is roughly divided into three parts (components) of an XML declaration, a document type definition (DTD), and a document realization value (XML instance). Also, from the viewpoint of processing, XML documents are classified into two types, well-formed and validated.

  Here, the XML declaration is completely different from the SGML declaration, and is simply an XML version declaration, a character code declaration, etc. DTD, like SGML, has elements and attributes that appear in a tagged document. , An entity definition part, and an XML instance is a part where an actual tagged document is written.

  A well-formed XML document has a correspondence between a start tag and an end tag, and an XML instance is written in accordance with a tagging rule defined in XML. A verified XML document is a DTD. Are tagged according to the element type declaration, element hierarchy defined by the attribute list declaration, attribute type, etc. It is a prerequisite that this verified XML document naturally follows the tagging rules as a well-formed XML document.

  Table 1 shows the correspondence between the component and the XML document classified as described above, and the correspondence between the component and the SGML document or HTML document (whether it is essential).

  As shown in Table 1, in the well-formed XML document, only the document realization value is an essential component, and in the verified XML document, the document type definition and the document realization value are essential components. In addition, in the SGML document, all components are essential, and in the HTML document, components other than the HTML declaration are indispensable components.

  In XML, a document is regarded as a set of elements having a hierarchical structure, and a mark used to identify each element is a tag. A tag indicating that an element of a document has started is called a start tag, a tag indicating that the element has ended is called an end tag, and the portion sandwiched between these two tags is the content of the element.

  Basic tagging of elements is as shown in FIG. The example shown in FIG. 38A is tagging when expressing one element, but there are actually several elements in the XML instance, and these elements have a hierarchical structure. That is, another group of elements may exist under a certain element, and in order to express such a hierarchical structure, tags are nested as shown in FIG. As shown in FIG. 38 (B), the element b and the element c which are lower than the element a are located in the same column in the hierarchical structure, and have a sibling relationship with the parent element “a”. The element having such a sibling relationship is written with the start tag of the brother element c immediately after the end tag of the brother element b. In addition, as a content of an element, a lower element can be written in a mixed form with plain text (text). In that case, tagging as shown in FIG.

Furthermore, the tag can not only represent the structure of the element, but can also give some attached information (attribute) to the element. That is, an attribute as attached information may be given to an element for reasons such as “I want to distinguish between element types” or “I want to refer to another element with a unique identifier”. As shown in FIG. 39, this attribute is represented by a pair of attribute name and attribute value, and is written in the start tag.
The processing of the XML document is further classified into five types as shown in Table 2 (patterns (1) to (5); numbers 1 to 5 in circles in Table 2) depending on the configuration.

  As shown in Table 2, the XML document has a DTD (pattern (1)), a DTD including an entity declaration (entity declaration) (pattern (2)), and an entity declaration to the outside. With DTD (pattern (3)), internal DTD described (pattern (4); common with SGML), and external file DTD (pattern (5); SGML, HTML) And common).

The XML documents corresponding to the patterns (1) to (3) are called “formatted XML documents”, and tags can be set without requiring verification by DTD. The XML documents corresponding to the patterns (4) and (5) are called “verified XML documents”.
In XML and SGML corresponding to patterns (1) to (4), a user can freely set tags. In pattern (4), both XML and SGML can define tags with their own DTDs in the document. On the other hand, HTML corresponding only to the pattern (5) depends only on the DTD (issued by the World Wide Web Consortium (W3C)) of the external file, and the user cannot freely set the tag.

  Pattern (1) is the simplest well-formed XML document that has only an XML instance and does not check as a verified XML document, thus completely removing the DTD. Even if there is no DTD, the XML instance The contents can be interpreted.

  The pattern (2) has a DTD including a replacement character string definition (entity declaration), and in order to use the shortened character string in the XML instance, the well-formed XML in which those shortened character strings are declared in the entity declaration in the DTD. It is a document. In this pattern (2), correspondence between a short character string and a long character string is defined in the DTD in order to replace a long character string in the content of the XML instance with a short character string using an entity reference.

  The pattern (3) is a well-formed XML document in which these files are declared with entity declarations in DTD to create an XML document with a plurality of files. In this pattern (3), an entity reference is used to quote an external file in the contents of the XML instance. In DTD, a short character string used in the XML instance, information specifying an actual file, and Is defined.

The pattern (4) defines an element type declaration and an attribute list declaration necessary for checking as a verified XML document in a DTD (DOCTYPE declaration) attached to an XML instance.
Pattern (5) defines element type declarations and attribute list declarations with DTDs existing in external files. Information specifying the external files is described in DTDs (DOCTYPE declarations) attached to XML instances. ing.

  Whether an XML document is interpreted as a well-formed XML document or an verified XML document depends on an XML processor (XML parser) that is software that interprets the XML document. As shown in FIG. 40, the XML processor analyzes the XML document, performs a check as a well-formed XML document and a check as a verified XML document, and then finishes the check (expressed as a tree structure). It fulfills the function of passing a document to other application software such as a browser.

  When compressing a structured document as described above, as described above, a universal encoding technique is used in which the structured document is compressed as a collection of character data.

Conventional compression methods for structured documents are roughly divided into the following two methods (a) and (b).
(A) A method of compressing only a plaintext portion sandwiched between tags without moving the tag from the original position.
(B) A technique in which only plain tags are compressed together by moving only the tags to the head of the document realization value (instance).

In the method (a), the tag itself is not compressed. Usually, the tag alone occupies a capacity of about 30% of the structured document. Therefore, unless the tag is compressed, the compression rate of the structured document is lowered.
In the method (b), in order to return the tag to the original position when decompressing and restoring the compressed document, a 2-byte identification code must be attached to the original position of the tag. The rate will drop.

  For example, when compressing an HTML document using the universal encoding technique, a method of compressing contents below a predetermined depth in the tree structure of the HTML document, or a more concise expression by detecting redundant parts of the tag expression of the HTML document. However, in the former method, the document structure below a predetermined depth in the tree structure is not known unless it is decompressed or restored. In the latter method, the compression target document is an HTML document. Element names and attribute names cannot be compressed.

  As described above, tags are necessary for searching in structured documents. However, if only tags are stored as they are as in method (a), the compression rate of structured documents deteriorates, and the method described above is used. If only the tag is compressed as in (b), the compressed file search function is lost and the search cannot be performed in the compressed state.

  Normally, a compressed file can be decompressed at about 200 KB / s. For example, an SGML / XML document having a size of 4 to 6 KB for a Japanese industrial standard A4 size is about 0.02 to 0.02. It will be expanded and viewable in 03 seconds. On the other hand, in the database search method, the search is performed in a processing time of about 0.08 seconds.

When a compression method is applied to a structured document database, depending on the size of the unit to be compressed, if a search after decompression is performed, the time until the search may exceed 0.1 seconds. For this reason, the selection of the compression method varies depending on whether or not the search time may exceed 0.1 seconds.
As described above, if the tag is not compressed, the compression rate of the structured document is lowered and the storage efficiency of the document data is lowered, which is not preferable in a system that handles a large-scale database.

  On the other hand, in a parts list or price list described in XML, a redundant expression such as a pair of a start tag and an end tag sandwiching a short word (content) [see FIG. 2 (A) and FIG. 2 (B)]. ] Appear frequently, but in such a case, it is desired to be able to compress the portion of the tag while maintaining the searchable state.

  In many cases, the data length of “content” sandwiched between tags in actual XML document data is short. Specifically, it is about 20 bytes and about 10 characters in Japanese. Usually, short data is difficult to compress. Moreover, if you leave a part of “content” as a search keyword, if the “content” data is short, it will leave the “content” as it is without being compressed, eventually reducing the compression rate of structured documents. Resulting in.

  The present invention has been devised in view of such problems, and enables compression of a tag portion without impairing the characteristics of the structured document, and the compression of the structured document in which the compression rate of the structured document is improved. It is an object of the present invention to provide a method, a compression apparatus, and a computer-readable recording medium on which a structured document compression program is recorded.

  To achieve the above object, a structured document compression method as a related technique of the present invention includes a document realization value analysis step for analyzing a tree structure of elements in a document realization value forming a structured document, and the document realization value. A document realization value structure that moves information about an element that is a leaf of the tree structure (hereinafter referred to as a leaf element) into the start tag of the parent element as an attribute of the parent element of the leaf element according to the analysis result in the analysis step And a changing step.

  In the document realization value configuration change step, the start tag, end tag, and content for the leaf element are deleted from the document realization value, and the element name and content, which are information about the leaf element, are attributed to the parent element, respectively. It may be added as a name and attribute value in the start tag of the parent element. At this time, if an attribute which is information about the leaf element is described in the start tag of the leaf element, the attribute is applied. The attribute name and attribute value may be added in the start tag of the parent element as the attribute name and attribute value of the parent element, respectively. In the document actual value configuration changing step, the end tag of the parent element may be deleted and the start tag of the parent element may be changed to an empty element tag.

  Further, a document type definition analyzing step for analyzing a tree structure of elements in the document type definition constituting the structured document, and an element (hereinafter referred to as a leaf) of the tree structure according to an analysis result in the document type definition analyzing step. A document type definition configuration change step of deleting information about the element) from the document type definition and redefining it as an attribute of the parent element of the leaf element in the document type definition. At this time, in the document type definition configuration change step, the element type declaration of the leaf element is deleted from the document type definition and the description of the leaf element is deleted from the element type declaration of the parent element. The information related to the element type declaration may be redefined as the attribute of the parent element in the attribute list declaration of the parent element. Further, in the document type definition, the attribute of the leaf element is determined by the attribute list declaration of the leaf element. If defined, the attribute list declaration of the leaf element may be deleted from the document type definition, and the attribute of the leaf element may be redefined as the attribute of the parent element in the attribute list declaration of the parent element.

  In order to achieve the above object, a structured document compression method according to the present invention (claim 1) includes a document realization value analyzing step for analyzing a description in a tag of a document realization value forming a structured document, and the document realization value. A tag dictionary that creates a tag dictionary that associates a character string described in a tag of the document realization value with a shortened character string that is shorter than the character string and that can identify the character string, according to the analysis result in the analysis step Using the tag dictionary created in the creation step and the tag dictionary creation step, a document actual value character that replaces the character string described in the tag of the document actual value with a shortened character string corresponding to the character string And a column replacement step.

  Further, the structured document compression method of the present invention (claim 2) includes a document realization value analyzing step for analyzing a description in a tag of a document realization value forming a structured document, and a document type definition forming the structured document. A document type definition analyzing step for analyzing the description, a character string described in the tag of the document realized value and in the document type definition according to the analysis result in the document realized value analyzing step and the document type definition analyzing step, and the document type definition Using the tag dictionary created in the tag dictionary creating step to create a tag dictionary that associates a shortened character string that is shorter than the character string and that can specify the character string, and realizing the document using the tag dictionary created in the tag dictionary creating step The document realization value character string replacement step of replacing the character string described in the value tag with the shortened character string corresponding to the character string, and the tag dictionary created in the tag dictionary creation step, the document The character string described in the definition, is characterized by having a document type definition string substitution step of replacing the shortening character string corresponding to the character string.

At this time, an element name and an attribute name described in the tag or in the document type definition may be treated as the character string, and the element name and the attribute name may be replaced with the abbreviated character string.
Further, by using a word dictionary that associates a word character string with a shortened character string that is shorter than the word character string and that can identify the word character string, the word character string included in the content of the document realization value is A word character string replacement step for replacing with a shortened character string corresponding to the word character string may be provided (Claim 4), and a character string described in the tag or in the document type definition is replaced with the shortened character string. In addition, after the word character string is replaced with the shortened character string, a variable length encoding step of compressing these character strings by variable length encoding may be provided.

  On the other hand, a structured document compression apparatus as a related technique of the present invention is a document realization value analysis unit that analyzes a tree structure of elements in a document realization value that forms a structured document, and an analysis result by the document realization value analysis unit, A document realization value configuration change unit that transfers information about an element that is a leaf of the tree structure (hereinafter referred to as a leaf element) as an attribute of the parent element of the leaf element into a start tag of the parent element. It is characterized by that.

  At this time, the document realization value configuration unit deletes the start tag, the end tag, and the content for the leaf element from the document realization value, and the element name and content, which are information about the leaf element, are respectively replaced with the parent element. The attribute name and attribute value of the parent element may be added to the start tag of the parent element, and when an attribute that is information about the leaf element is described in the start tag of the leaf element, Such attribute name and attribute value may be added to the start tag of the parent element as the attribute name and attribute value of the parent element, respectively, or the end tag of the parent element may be deleted and the start of the parent element The tag may be changed to an empty element tag.

  In addition, a document type definition analysis unit that analyzes a tree structure of elements in the document type definition that forms the structured document, and an element that becomes a leaf of the tree structure (hereinafter referred to as a leaf element) according to the analysis result by the document type definition analysis unit The document type definition configuration change unit may be further provided that deletes the information on the document type definition from the document type definition and redefines the leaf element as an attribute of the parent element of the leaf element.

  The structured document compression apparatus according to the present invention (Claim 6) includes a document realization value analysis unit that analyzes a description in a tag of a document realization value that forms a structured document, and the analysis result by the document realization value analysis unit, A tag dictionary creating unit that creates a tag dictionary that associates a character string described in a tag of a document realization value with a shortened character string that is shorter than the character string and that can identify the character string; and the tag dictionary creating unit And a document realization value character string replacement unit that replaces the character string described in the tag of the document realization value with a shortened character string corresponding to the character string. It is characterized by that.

The structured document compression apparatus according to the present invention (Claim 7) includes a document realization value analysis unit that analyzes a description in a tag of a document realization value that constitutes a structured document, and a description of a document type definition that constitutes the structured document. According to the analysis result of the document type definition analysis unit to be analyzed, the document realization value analysis unit, and the document type definition analysis unit, the character string described in the document realization value tag and the document type definition and the character string A tag dictionary creating unit that creates a tag dictionary that is associated with a shortened character string that is short and can specify the character string, and using the tag dictionary created by the tag dictionary creating unit, the tag of the document realization value The document realization value character string replacement unit that replaces the character string described in the character string with a shortened character string corresponding to the character string and the tag dictionary created by the tag dictionary creation unit, Corresponds to the character string described Is characterized in that shortened constructed and a document type definition string substitution unit to substitute the string that.
At this time, an element name and an attribute name described in the tag or in the document type definition may be treated as the character string, and the element name and the attribute name may be replaced with the abbreviated character string.

  Furthermore, a recording medium as a related technique of the present invention is a computer-readable recording medium storing a structured document compression program for realizing a function of compressing a structured document by a computer, and the structured document compression program The document realization value analysis unit that analyzes the tree structure of the elements in the document realization value that forms the structured document, and the element that becomes the leaf of the tree structure (hereinafter referred to as leaf) according to the analysis result by the document realization value analysis unit The computer is caused to function as a document realization value configuration changing unit that transfers information on the element) into the start tag of the parent element as an attribute of the parent element of the leaf element.

  At this time, the structured document compression program deletes the start tag, end tag, and contents of the leaf element from the document actual value by the document actual value configuration unit, and an element name that is information about the leaf element And the contents may be added to the start tag of the parent element as the attribute name and attribute value of the parent element, respectively, and the leaf element is further included in the start tag of the leaf element. When the attribute that is information about the attribute is described, the document actual value configuration changing unit uses the attribute name and attribute value relating to the attribute as the attribute name and attribute value of the parent element, respectively, and the start tag of the parent element The computer may function to be added to the document, or the document realization value configuration changing unit deletes the end tag of the parent element and changes the start tag of the parent element to an empty element tag. So that, it may function the computer.

  Further, the structured document compression program analyzes a tree structure of elements in the document type definition constituting the structured document, and the analysis result of the tree structure according to the analysis result by the document type definition analyzing unit. As a document type definition configuration change unit that deletes information about a leaf element (hereinafter referred to as a leaf element) from the document type definition and redefines it as an attribute of the parent element of the leaf element in the document type definition. The computer may function.

  The recording medium of the present invention (Claim 9) is a computer-readable recording medium storing a structured document compression program for realizing the function of compressing a structured document by a computer. A document realization value analysis unit for analyzing a description in the tag of the document realization value forming the structured document, and a character string described in the tag of the document realization value according to the analysis result by the document realization value analysis unit A tag dictionary creating unit that creates a tag dictionary that is associated with a shortened character string that is shorter than a character string and that can specify the character string; and the document using the tag dictionary created by the tag dictionary creating unit, The computer is caused to function as a document actual value character string replacement unit that replaces a character string described in a tag of an actual value with a shortened character string corresponding to the character string.

The recording medium of the present invention (Claim 10) is a computer-readable recording medium storing a structured document compression program for realizing the function of compressing a structured document by a computer, and the structured document compression program is stored in the recording medium. A document realization value analysis unit that analyzes a description in a tag of a document realization value that forms the structured document, a document type definition analysis unit that analyzes a description of a document type definition that forms the structured document, and a document realization value analysis unit And a character string described in the tag of the document realization value and in the document type definition and a shortened character string shorter than the character string and capable of specifying the character string according to the analysis result by the document type definition analysis unit A tag dictionary creating unit that creates a corresponding tag dictionary, and using the tag dictionary created by the tag dictionary creating unit, a character string described in a tag of the document realization value is shortened corresponding to the character string Using the tag dictionary created by the document realization value character string substitution unit that replaces the character string and the tag dictionary creation unit, the character string described in the document type definition is converted to a shortened character corresponding to the character string. The computer is made to function as a document type definition character string replacement unit for replacement with a column.
At this time, the structured document compression program treats the element name and attribute name described in the tag or in the document type definition as the character string, and causes the computer to convert the element name and attribute name to the short character. A column may be substituted (claim 11).

  According to the above-described method and apparatus for compressing a structured document and a computer-readable recording medium on which a structured document compression program is recorded, the following effects or advantages can be obtained.

  (1) By analyzing the tree structure of the element in the document realization value and moving the information about the leaf element in the document realization value as the attribute of the parent element in the start tag of the parent element according to the analysis result, more specific Specifically, the start tag, end tag, and content for the leaf element are deleted from the document realization value, and the element name and content that is information about the leaf element are used as the parent element attribute name and attribute value, respectively. By adding it in the tag, the description of the leaf element can be handled as an attribute of the parent element, so there is no need to describe the start tag and end tag of the leaf element, and the characteristics of the structured document are not impaired. The description of the tag relating to the leaf element is omitted / compressed while maintaining the searchable state.

  Therefore, the compression rate of the structured document can be greatly increased, and the storage efficiency of document data can be greatly increased in a system that handles a large-scale database. In particular, when a bill of materials or a price list with a lot of short words is described in a structured document, the paired expression of the start tag and end tag with a short word (content) sandwiched can be omitted. The compression rate can be greatly increased.

(2) When an attribute is described in the start tag of the leaf element, the attribute name and attribute value relating to the attribute are added to the start tag of the parent element as the attribute name and attribute value of the parent element, respectively. The description concerning the attribute of the leaf element is also handled as the attribute of the parent element, and the compression rate of the structured document can be further increased.
(3) By deleting the end tag of the parent element and changing the start tag of the parent element to an empty element tag, the end tag of the parent element can be further deleted from the description of the structured document. The compression rate can be further increased.

  (4) Analyzing the tree structure of elements in the document type definition, and deleting the leaf element information from the document type definition and redefining it in the document type definition as an attribute of the parent element according to the analysis result. Specifically, the element type declaration of the leaf element is deleted from the document type definition, the description related to the leaf element is deleted from the element type declaration of the parent element, and the information related to the element type declaration of the leaf element is used as the attribute of the parent element. By redefining (Claim 6), the compression processing corresponding to the compression performed on the document realization value is also performed on the document type definition, and the description of the leaf element is handled as the attribute of the parent element. Can do. Therefore, the description of the element type declaration for the leaf element is omitted without compromising the characteristics of the structured document, and the document type definition is compressed while maintaining the searchable state, and the compression rate of the structured document is increased. Can be increased.

  (5) When the attribute of a leaf element is defined by the attribute list declaration of the leaf element in the document type definition, the attribute list declaration of the leaf element is deleted from the document type definition, and the attribute of the leaf element is set as the attribute of the parent element By redefining, the description about the attribute of the leaf element can be handled as the attribute of the parent element. Therefore, the description of the attribute list declaration for the leaf element is omitted without losing the characteristics of the structured document, while maintaining the searchable state, and the document type definition is further compressed. Can be further enhanced.

  (6) Analyzing the description in the tag of the document actual value, creating a tag dictionary according to the analysis result, and using the tag dictionary, the character string described in the tag of the document actual value is If the character string in the tag is compressed without impairing the characteristics and structure of the structured document, the character string in the tag is compressed without replacing the character string. The document compression rate can be greatly increased, and consequently the storage efficiency of document data can be significantly increased in a system that handles a large-scale database.

  (7) Analyze the description of the document realization value tag and document type definition, create a tag dictionary according to the analysis result, and use the tag dictionary to describe in the document realization value tag and document type definition By replacing a character string with a shortened character string shorter than the character string and specifying the character (claims 2, 7, and 10), the structured document has a document type definition. However, since the character string of the document type definition is compressed without losing the characteristics and structure of the structured document, the compression rate of the structured document can be greatly increased. Data storage efficiency can be greatly increased.

  (8) By replacing element names and attribute names described in tags and document type definitions with abbreviated character strings (claims 3, 8, and 11), tag portions and document type definitions are maintained in a searchable state. Can be compressed. In other words, element names and attribute names are replaced, and attribute values are retained in their original form, so that document data can be searched and document structure can be grasped without being decompressed. ing. Therefore, when searching after analyzing the structure of the document after compressing the document, it is not necessary to decompress the compressed document. Even if the document data is compressed and stored in a large-scale database, the document data search processing is performed. Etc. can be performed in a short time.

(9) By using a word dictionary, the word character string included in the content of the document realization value is replaced with a shortened character string that is shorter than the word character string and can identify the word character string (claim 4). Since the plain text portion (content of the document realization value) of the structured document is compressed, the compression rate of the structured document can be further increased.
(10) The compression rate of the structured document can be further increased by further compressing the character string after the replacement processing by variable length coding (claim 5).

Embodiments of the present invention will be described below with reference to the drawings.
[1] Description of First Embodiment First, the compression principle of a structured document in the first embodiment of the present invention will be described with reference to FIGS. 2A to 2D and FIGS. 3A to 3C are diagrams for explaining the compression principle of the structured document (document realization value) in the first embodiment. FIGS. 4A and 4B are diagrams for explaining the compression principle of the structured document (DTD) in the first embodiment. In the following description of the first embodiment of the present invention, the structured document is an XML document.

As described above, in the parts list and price list written in XML, as shown in FIG. 2A and FIG. 2B, a pair of a start tag and an end tag sandwiching a short word (plain text) as content A redundant expression such as appears frequently.
Here, FIG. 2A shows a general description of XML in the case where two child elements (element 2 and element 3 having no child element) exist under a certain parent element (element 1). An example is shown. FIG. 2B shows a specific description example corresponding to the general description example shown in FIG. Hereinafter, an element having no child element may be referred to as a leaf element, a leaf element, or simply a leaf.

  In the general description example shown in FIG. 2A, element 1 is designated with element name 1 and attribute information (attribute name 1 and attribute value 1), and element 2 is a child element of element 1. Has an element name 2 and content 2, and element 3, which is a child element of element 1 (similar to element 2) like element 2, is specified with element name 3 and attribute information (attribute name) 3 and attribute value 3) are specified and content 3 is included.

In the specific description example shown in FIG. 2B, the element name 1 is “book”, the attribute name 1 is “field”, the attribute value 1 is “book”, the element name 2 is “title”, and the content 2 Is “Introduction to XML”, the element name 3 is “author”, the attribute name 3 is “year”, the attribute value 3 is “1955”, and the content 3 is “Moto Sato”.
The tree structure of the description example shown in FIGS. 2A and 2B is shown in FIG. Further, FIG. 2D shows a leaf element list for the examples shown in FIGS. 2A to 2C.

  In the description example shown in FIG. 2B, the description <book field = "book"> on the first line is the start tag of the element "book", and the description </ book> on the fifth line is the element "book". The description on the 2nd to 4th lines enclosed by these tags indicates the content of the element “book”. The description “field =“ book ”” in the start tag on the first line indicates the attribute information (the attribute name is “field” and the attribute value is “book”) of the element “book”.

In the second line, the description <title> is the start tag of the element “title”, the description </ title> is the end tag of the element “title”, and the description “Introduction to XML” between these tags is the element “ The content of “title”.
Similarly, the description <author year = "1955"> on the third line is the start tag for the element "author" and the description </ author> on the fourth line is the end tag for the element "author". Is the content of the element “author”. The description “year =“ 1955 ”” in the start tag on the third line indicates the attribute information of the element “author” (attribute name is “year” and attribute value is “1955”).

  In the first embodiment of the present invention, as shown in FIGS. 2 (A) to 2 (D), elements that have no child elements and are arranged as leaves of a tree structure (hereinafter referred to as leaf elements, leaf elements or simply 2), and a file about the leaf element list as shown in FIG. 2D is output. Then, based on the file, as shown in FIGS. 3A and 3B, the leaf element name and contents are replaced with the attribute of the upper element 1 (parent element) and the leaf element is deleted. At the same time, the start tag of element 1 is changed to an empty element tag. At this time, the attribute information (attribute name 3 and attribute value 3) of the element 3 is also arranged as an attribute of the element 1 that is equivalent to the element name 3 and the content 3 of the element 3.

Here, as shown in FIGS. 3A and 3B, a tag described by enclosing an element name or attribute information between “<” and “/>” is an empty element tag having no content. It is. At this time, the attribute information does not necessarily have to be described / specified.
FIG. 3A shows a description obtained as a result of compressing the general description example shown in FIG. 2A by the compression method of the first embodiment, and FIG. A description obtained as a result of compressing the specific description example shown in (B) by the compression method of the first embodiment is shown.

As shown in FIGS. 3A and 3B, in the start tag (empty element tag) of element 1, element name 1 and attribute information (attribute name 1 and attribute value 1) are designated. In addition, element name 2 and content 2 are designated as the second attribute name and attribute value of element 1, respectively, element name 3 and content 3 are designated as the third attribute name and attribute value of element 1, respectively. Attribute name 3 and attribute value 3 are designated as the fourth attribute name and attribute value of element 1, respectively.
FIG. 3C is a schematic representation of the structure of the element 1 having the start tag (empty element tag) shown in FIGS. 3A and 3B.

On the other hand, when the structured document subjected to the compression processing as described above with reference to FIGS. 2 and 3 has DTD, FIG. 4A and FIG. As shown in B), DTD is changed (compressed).
4A shows the DTD before change (before compression) that defines the XML document shown in FIG. 2B. In the DTD shown in FIG. 4A, the description on the first line is a document type declaration (DOCTYPE declaration). Here, the document type name of this document, that is, the element name of the top-level element is “book”. Is declared.

The description between the “[” at the end of the first line and “]>” on the seventh line (the description on the second to sixth lines) defines the structure of this document.
The element name of the element type declaration on the second line needs to match the document type name. Here, “book” is designated as the element name of the top-level element according to the rule. In addition, the fact that there are two child elements with the element names “title” and “author” side by side under the parent element “book” is the content model description “(title, author) ”. That is, the element “book” is declared to be composed of two child elements “title” and “author”.

  Furthermore, the element type declaration on the third line specifies “title” as the element name of the element and declares that the content of the element “title” is character data (#PCDATA). Similarly, the element type declaration on the fourth line specifies “author” as the element name of the element and declares that the content of the element “author” is character data (#PCDATA). In these element type declarations, the existence of child elements under each element is not declared. In other words, there are no child elements having these elements as parents, and these elements are elements that form leaves of a tree structure.

In the attribute list declaration on the fifth line, the attribute name “field”, the three types of candidate “book”, “magazine”, “booklet”, and default values for the attribute “book” are included. "Book" is declared.
Further, in the attribute list declaration on the sixth line, as an attribute accompanying the element “author”, an attribute name “year” indicating the author's birth year and a data type (CDATA) of the attribute value are declared.
The description “]>” on the seventh line indicates that the internal subset description part of the document type declaration on the first line is completed.

  FIG. 4B shows the DTD after modification (after compression) that defines the XML document shown in FIG. 3B. In the DTD shown in FIG. The element type declarations for the element “title” and the element “author” that existed in the DTD shown in FIG. 4A disappear, and the child element definition in the element type declaration of the parent element “book” on the second line (Content model) disappears. In addition, the attribute “year” associated with the element “author” has also disappeared.

  Instead, attribute names “title”, “author”, and “year” are newly added as attributes of the parent element “book” in the attribute list declaration on the third to sixth lines. Then, “#PCDATA” is declared as a candidate attribute value of these attribute names “title” and “author”, and “CDATA” is declared as a candidate attribute value of the attribute name “year”.

  Here, “#PCDATA” includes tags and marks called entity references in addition to character data. In entity reference, when a structured document is described, a specific character string is represented by a predetermined mark in character data, and when character data is passed to application software, the original character string is added to the mark portion. It is a function to substitute. However, when an entity reference is made in an attribute value, it is allowed to refer to a character string in the same document as that of the attribute value, but it is not allowed to refer to a character string in an external file. Absent. As an attribute value representing the year of birth, CDATA, which is pure character data, is declared.

  In the structured document compression method of the first embodiment, the DTD is also changed / compressed in accordance with the compression processing of the document realization value as described above. Therefore, the XML processor described above with reference to FIG. Thus, a new XML document as shown in FIGS. 3A and 3B can be properly considered.

  Further, as described above, conventionally, information expressed as “element” is converted into “attribute” of the parent element, and DTD is also changed in accordance with the structure change of the XML document accompanying the conversion. As a result, information conventionally expressed as “element” is detected as “attribute” of the parent element. Therefore, the XML document expression is compressed and saved by the number of “element name bytes” + “3 bytes”.

Hereinafter, the first embodiment of the present invention will be described in more detail and specifically with reference to FIGS.
FIG. 1 is a block diagram showing the functional configuration of a structured document compression apparatus according to the first embodiment of the present invention. As shown in FIG. 1, the compression apparatus of the first embodiment includes a document storage unit 10, The document realization value analysis unit 20, the DTD analysis unit 30, the document realization value configuration change unit 40, the DTD configuration change unit 50, the new DTD file creation unit 60, and the old and new DTD correspondence table output unit 70 are configured.

Here, the compression apparatus of the present embodiment is realized by a computer system such as a personal computer, which is configured by connecting a CPU, a RAM, a ROM, and the like through a bus line.
That is, in addition to the RAM and ROM serving as the document storage unit 10, the RAM includes a document realization value analysis unit 20, a DTD analysis unit 30, a document realization value configuration change unit 40, a DTD configuration change unit 50, and a new DTD file. Application programs for realizing the creation unit 60 and the old and new DTD correspondence table output unit 70 are stored.

  When the CPU executes the application program, the document realization value analysis unit 20, the DTD analysis unit 30, the document realization value configuration change unit 40, the DTD configuration change unit 50, the new DTD file creation unit 60, and the old and new DTD support A function (details of which will be described later) as the table output unit 70 is realized, and the structured document compression apparatus of the first embodiment is realized.

  A program for realizing the compression apparatus of the first embodiment is provided in a form recorded on a computer-readable recording medium such as a flexible disk or a CD-ROM. Then, the computer reads the program from the recording medium, transfers it to the internal storage device or the external storage device, and uses it. The program may be recorded in a storage device (recording medium) such as a magnetic disk, an optical disk, or a magneto-optical disk, and provided from the storage device to the computer via a communication path.

  When the function of the compression device according to the first embodiment is realized by a computer, the program stored in the internal storage device (for example, RAM) is executed by the microprocessor (for example, CPU) of the computer. At this time, the program recorded on the recording medium may be directly read and executed by the microprocessor.

  In the present embodiment, the computer is a concept including hardware and an operating system, and means hardware that operates under the control of the operating system. Further, when an operating system is unnecessary and hardware is operated by an application program alone, the hardware itself corresponds to a computer. The hardware includes at least a microprocessor such as a CPU and means for reading a computer program recorded on a recording medium.

  The application program is stored in such a computer on the document realization value analysis unit 20, the DTD analysis unit 30, the document realization value configuration change unit 40, the DTD configuration change unit 50, the new DTD file creation unit 60, and the old and new DTD correspondence table output unit. The program code for realizing the function 70 is included. Some of the functions may be realized by an operating system instead of an application program.

  In addition to the flexible disk, CD-ROM, magnetic disk, optical disk, and magneto-optical disk described above, the recording medium in this embodiment includes an IC card, a ROM cartridge, a magnetic tape, a punch card, and an internal storage device (RAM) of the computer. In addition, various computer-readable media such as an external storage device or a printed matter on which a code such as a barcode is printed can be used.

  In the compression apparatus of the first embodiment shown in FIG. 1, the document storage unit 10 stores an XML document that is a structured document. In this embodiment, both XML documents before and after compression are stored. In addition to storing, a new and old DTD correspondence table described later is also stored.

  The document realization value analysis unit 20 analyzes the document realization value forming the XML document and analyzes the element tree structure (parent-child relationship) in the document realization value and the description of the document realization value. This will be described later with reference to the flowchart shown in FIG. Then, the document realization value analysis unit 20 of this embodiment outputs, for example, a leaf element list (file) as shown in FIG. 2D as the analysis result of the tree structure (parent-child relationship) of the elements.

  The DTD analysis unit (document type definition analysis unit) 30 analyzes the DTD when the XML document is a verified XML document (that is, when the XML document is the pattern (4) or (5)). It analyzes the tree structure (parent-child relationship) of elements and the description of DTD, and the analysis procedure will be described later with reference to the flowchart shown in FIG. The DTD analysis unit 30 of the present embodiment also outputs a leaf element list (file) as shown in FIG. 2D, for example, as an analysis result of the element tree structure (parent-child relationship). However, when the XML document is the pattern (4), the DTD is read from the document storage unit 10, but when the XML document is the pattern (5), the DTD is read from the external file 100.

At this time, the leaf element list output from the document realization value analysis unit 20 and the DTD analysis unit 30 detects the elements arranged as leaves of the tree structure without any child elements, as described above. It clearly specifies the correspondence between leaf elements and parent elements.
In order to simplify the expression of the document realization value, the document realization value configuration change unit 40 obtains information about the leaf element in the document realization value according to the analysis result (leaf element list) by the document realization value analysis unit 20. It is moved within the start tag of the parent element as an attribute of the parent element of the leaf element, and the following configuration change processing (a1) to (a3) is executed.

(A1) The start tag, end tag, and content for the leaf element are deleted from the document realization value, and the element name and content, which is information about the leaf element, are used as the parent element attribute name and attribute value, respectively. Append inside.
(A2) When an attribute that is information about a leaf element is described in the start tag of the leaf element, the attribute name and attribute value related to the attribute are used as the attribute name and attribute value of the parent element, respectively. Append inside.

(A3) The end tag of the parent element is deleted, and the start tag of the parent element is changed to an empty element tag.
The configuration change procedure by the document realization value configuration change unit 40 will be described later with reference to the flowchart shown in FIG. In the present embodiment, the configuration change result (compressed document actual value) by the document actual value configuration changing unit 40 is output / stored in the document storage unit 10, but is output / stored in another recording medium or the like. May be.

  The DTD configuration change unit (document type definition configuration change unit) 50, when the XML document is a verified XML document (when the XML document is the pattern (4) or (5)), the document realization value configuration change unit 40 In order to simplify the representation of the DTD in accordance with the change in the configuration, the information about the leaf element in the DTD is deleted from the DTD according to the analysis result (leaf element list) by the DTD analysis unit 30, and the parent of the leaf element The element is redefined as an element attribute by DTD, and the following configuration change processing (b1) and (b2) are executed.

  (B1) The element type declaration of the leaf element is deleted from the DTD, the description related to the leaf element (content model description) is deleted from the element type declaration of the parent element, and the information related to the element type declaration of the leaf element (the deleted part) Is redefined in the attribute list declaration of the parent element as the attribute of the parent element. At this time, the element name and content (data type) of the leaf element are declared and redefined as the attribute name and attribute value of the parent element, respectively.

  (B2) When the attribute of the leaf element is defined in the DTD by the attribute list declaration of the leaf element, the attribute list declaration of the leaf element is deleted from the DTD, and the attribute of the leaf element is set as the attribute of the parent element of the leaf element Redefine in parent element attribute list declaration. At this time, the attribute name and attribute value for the leaf element are declared and redefined as the attribute name and attribute value of the parent element, respectively.

  The configuration change procedure by the DTD configuration change unit 50 will be described later with reference to the flowchart shown in FIG. In the present embodiment, when the XML document is the pattern (4), the configuration change result (compressed DTD) by the DTD configuration change unit 50 is changed to the configuration change result (compressed value) by the document realization value configuration change unit 40. The document realization value) is output / stored in the document storage unit 10, but may be output / stored together with the compressed document realization value in another recording medium or the like.

When the DTD exists in the external file 100 (when the XML document is the pattern (5)), the new DTD file creation unit 60 selects a file (new DTD file) for the DTD that has been changed by the DTD configuration change unit 50. It is created and output to the external file 100.
When the DTD exists in the external file 100 (when the XML document is the pattern (5)), the old and new DTD correspondence table output unit 70 clarifies the correspondence between the DTD before the configuration change and the new DTD after the configuration change. An old and new DTD correspondence table is created and output to the document storage unit 10.

  Note that the XML document compressed by the compression method (element-to-attribute conversion) of the first embodiment does not impair the characteristics of the XML document at all, and remains in the compressed state (without decompression). Therefore, there is no need to discuss the decompression of the compressed XML document. Therefore, in the first embodiment, only the compression method will be described.

Next, the operation of the compression apparatus according to the first embodiment will be described with reference to FIGS.
First, a structured document (XML document) compression procedure according to the first embodiment will be described with reference to the flowchart shown in FIG. 5 (steps S11 to S29). In the compression method of the first embodiment, as described above, by converting the element to the attribute, the leaf element disappears and the parent element of the leaf becomes an empty element.

  In the compression apparatus of the first embodiment, although not shown in FIG. 1, the XML document stored in the document storage unit 10 is one of the patterns (1) to (5) (see Table 2). It has a pattern recognition function for recognizing it. The processing by this pattern recognition function corresponds to the processing by steps S12 to S14 shown in FIG.

  When an XML document to be compressed is input and stored in the document storage unit 10 (step S11), it is determined whether or not “<! DOCTYPE” is described in the XML document (step S12) and is not described. If so (NO route of step S12), the XML document is recognized as a well-formed XML document having no DTD, that is, an XML document of pattern (1), and steps S15, S16, and S29 are executed as described later. .

When “<! DOCTYPE” is described in the XML document (YES route in step S12), it is determined whether or not “[” is described thereafter (step S13).
When “<! DOCTYPE” is described but “[” is not described (NO route of step S13), the XML document is a verified XML document having the DTD as the external file 100, that is, the pattern (5). As described later, steps S21 to S29 are executed.

If “[” is described (YES route in step S13), it is determined whether “<! ELEMENT” (or “<! ATTLIST”) is described (step S14).
When “<! DOCTYPE” and “[” are described but “<! ELEMENT” is not described (NO route of step S14), a well-formed XML document having a DTD including an entity declaration, that is, a pattern (2 ) And the steps S15, S16 and S29 are executed as in the case of the pattern (1).

When “<! DOCTYPE”, “[”, and “<! ELEMENT” are all described (YES route in step S14), a verified XML document having a DTD in the XML document, that is, the XML of the pattern (4) The document is recognized and steps S17 to S20 and S29 are executed.
Hereinafter, compression processing for each pattern (1) to (5) will be described with reference to specific examples (first to fourth examples) shown in FIGS.

FIGS. 10A and 10B are diagrams for explaining specific compression processing (first example) of an XML document according to the first embodiment.
The XML document before compression shown in FIG. 10A is the XML document of the pattern (1) described above, and an XML declaration indicating that this document is a version 1.0 XML document is described in the first line. In the second to fifth lines, document realization values are described. The document realization value described here is the description of the XML document shown in FIG. 2B except that the attribute information (field = “book”) is not described in the start tag of the element “book”. Same as example.

  Since “<! DOCTYPE” is not described in the XML document (pattern (1)) shown in FIG. 10A, the process proceeds from the NO route of step S12 to step S15, and the document realization value analysis unit 20 performs the process. The document realization value is analyzed. This detects where the leaf element is described in the document realization value, and the correspondence between the element name of the leaf element and the element name of the parent element is registered as a leaf element list. Is output. In the case of the XML document shown in FIG. 10A, a leaf element list similar to that shown in FIG. 2D is obtained.

  Then, according to the leaf element list (analysis result) obtained in step S15, the document realization value configuration changing unit 40 converts the element name and content of the leaf element into the attribute name and attribute value of the parent element, respectively. The start tag of the parent element is changed to an empty element tag (step S16). At this time, the attribute “year” attached to the leaf element “author” replaces the attribute of the parent element “book”.

Through steps S15 and S16, for example, the XML document shown in FIG. 10A is changed and compressed into an XML document as shown in FIG. 10B, and then output and stored as a compressed document in the document storage unit 10 and the like. (Step S29).
In the XML document shown in FIG. 10B, the description of the XML declaration in the first line is the same as before compression, but the descriptions in the second to fifth lines in FIG. is described. That is, in the XML document shown in FIG. 10B, as in the example shown in FIG. 3B, all the information relating to the two child elements “title” and “author” is the start of the parent element “book”. The tag (empty element tag) is described as an attribute of the parent element “book”.

FIG. 11A and FIG. 11B are diagrams for explaining a specific compression process (second example) of an XML document according to the first embodiment.
The XML document before compression shown in FIG. 11A is the XML document of the pattern (2) described above, and an XML declaration indicating that this document is a version 1.0 XML document is described in the first line. In the 2nd to 4th lines, the DTD including the replacement character string definition (substance declaration) is described, and in the 5th to 8th lines, the document realization value is described.

In the DTDs in the 2nd to 4th lines, the entity of the replacement character string “XML” used in the document realization value (XML instance) is “Extensible Markup Language” by the entity declaration (line 3) included in the document type declaration. It is defined that there is.
The document realization values described in the 5th to 8th lines are almost the same as the description example in the 2nd to 5th lines of the XML document shown in FIG. 10 (A), but the example shown in FIG. 11 (A). Then, “Introduction to XML (abbreviation of &XML;)” is described as the content of the element “title” on the sixth line.

Here, “&XML;” is a description instructing to refer to the entity of the replacement character string “XML”, and is expressed as “Extensible Markup Language” in a document that is actually output by display or printing. Will be.
In the XML document (pattern (2)) shown in FIG. 11A, both “<! DOCTYPE” and “[” are described, but “<! ELEMENT” and “<! ATTLIST” are described. Therefore, the process proceeds from the NO route of step S14 to step S15, and the same process as the XML document of the pattern (1) described above is executed. At this time, the description of the entity reference (“&XML;” in FIG. 11A) in the content of the document realization value is handled as it is as the attribute value of the parent element.

Thus, for example, the XML document shown in FIG. 11A is changed and compressed into an XML document as shown in FIG. 11B, and then output and stored as a compressed document in the document storage unit 10 or the like (step) S29).
In the XML document shown in FIG. 11B, the descriptions in the first to fourth lines are the same as before compression, but in the fifth line, the descriptions in the fifth to eighth lines in FIG. Has been. That is, also in the XML document shown in FIG. 11B, as in the example shown in FIG. 10B, all the information related to the two child elements “title” and “author” is the start of the parent element “book”. The tag (empty element tag) is described as an attribute of the parent element “book”. However, in FIG. 11B, “Introduction to XML (&XML;)” is described as it is as the attribute value for the attribute name “title”.

  By the way, the XML document of the pattern (3) [see, for example, FIG. 22A] uses an entity reference to quote an external file in the content of the document realization value (XML instance) as described above. In the first embodiment, the content of the document realization value is handled as the attribute value of the parent element. However, since the entity reference to the outside cannot be used as the attribute value due to the specification of the XML document, the XML document of the pattern (3) In addition, the compression method of the first embodiment is not applied.

FIGS. 12A and 12B are diagrams for explaining specific compression processing (third example) of an XML document according to the first embodiment.
The XML document before compression shown in FIG. 12A is an XML document of the pattern (4) described above, and an XML declaration indicating that this document is a version 1.0 XML document is described on the first line. The DTD is described in the 2nd to 8th lines, and the document realization value is described in the 9th to 12th lines. Here, the DTD described in the 2nd to 8th lines is the same as the DTD shown in FIG. 4A, and the document realization value described in the 9th to 12th lines is shown in FIG. The description is omitted because it is the same as the description example of the document actual value.

  In the XML document (pattern (4)) shown in FIG. 12A, “<! DOCTYPE” and “[” are described and “<! ELEMENT” or “<! ATTLIST” is also described. The process proceeds from the YES route in step S14 to step S17, where the document actual value analysis unit 20 analyzes the document realization value and the DTD analysis unit 30 analyzes the DTD (step S18). As a result, it is detected where the leaf element is described in the document realization value or DTD, and the correspondence between the element name of the leaf element and the element name of the parent element is the leaf element list. Registered and output as In the case of the XML document shown in FIG. 12A, a leaf element list similar to that shown in FIG. 2D is obtained.

  Then, according to the leaf element list (analysis result) obtained in step S17, the document actual value configuration changing unit 40 converts the element name and content of the leaf element into the attribute name and attribute value of the parent element, respectively. The start tag of the parent element is changed to an empty element tag (step S19). At this time, the attribute “year” attached to the leaf element “author” replaces the attribute of the parent element “book”.

  Also, since the XML document shown in FIG. 12A is the pattern (4) (that is, a verified XML document in which DTD is described), the configuration of the document actual value by the document actual value configuration change unit 40 is shown. In accordance with the change, the DTD configuration is changed by the DTD configuration changing unit 50 as follows (step S20).

  In other words, in the document realization value after the configuration change, the parent element “book” has no child elements “title” and “author”, so the child element is declared in the element type declaration for the parent element “book”. The content model (title, author) is deleted. In the document realization value after the configuration change, the leaf elements “title” and “author” are deleted and changed to the attributes of the parent element “book”, and the attribute “year” of the element “author” is also changed to the parent element “book”. ", The element type declaration of the elements" title "and" author "and the attribute list declaration of the element" author "are also deleted. On the other hand, since the parent element “book” will newly have “title”, “author”, and “year” as attributes, the attribute name and attribute of the new attribute in the attribute list declaration of the parent element “book” Enumerate values.

The XML document shown in FIG. 12A is changed / compressed into the XML document shown in FIG. 12B by steps S17 to S20 described above, and then output and stored as a compressed document in the document storage unit 10 or the like. (Step S29).
In the XML document shown in FIG. 12B, the descriptions on the first and second lines are the same as before compression, but the description of the content model is deleted from the element type declaration of the element “book” on the third line. Yes. Further, the descriptions on the 4th to 7th lines are the descriptions of the 4th to 7th lines in FIG. 12 (A) combined in the attribute list declaration of the element “book”.

  Further, as in the example shown in FIG. 3B, the descriptions of the 9th to 12th lines in FIG. 12A are collectively described in the 9th line. That is, even in the XML document shown in FIG. 12B, as in the example shown in FIG. 3B, all information related to the two child elements (leaf elements) “title” and “author” In the start tag (empty element tag) of “book”, it is described as an attribute of the parent element “book”.

FIGS. 13A to 13D are diagrams for explaining specific compression processing (fourth example) of an XML document according to the first embodiment.
The XML document before compression shown in FIG. 13A is an XML document of the pattern (5) described above, and an XML declaration indicating that this document is a version 1.0 XML document is described on the first line. In the second line, the DTD including information (system identifier) for designating the DTD of the external file 100 is described, and the document realization value is described in the third to sixth lines. Here, the document realization values described in the 3rd to 6th lines are the same as the description example of the document realization values shown in FIG.

In the DTD document type declaration on the second line, it is declared and defined that the DTD (file name “.. \ book.dtd”) held in the external file 100 is used by the system identifier “SYSTEM”.
The DTD of the file name “.. \ book.dtd” is described as shown in FIG. 13B corresponding to the configuration of the document realization value in FIG. The DTD (first to fifth lines) shown in FIG. 13B is the same as the description example of the second to sixth lines in the DTD shown in FIG.

  In the XML document (pattern (5)) shown in FIG. 13A, “<! DOCTYPE” is described, but after that “[” is not described, the DTD in the external file 100 is specified. Since the system identifier to be processed is described, the process proceeds from the NO route of step S13 to step S21, the document actual value is analyzed by the document real value analysis unit 20, and the external is performed by the DTD analysis unit 30 according to the system identifier. The DTD (file name “.. \ book.dtd”) read from the file 100 is analyzed (step S22). Thereby, in the document realization value or DTD, it is detected where the leaf element is described, and the correspondence between the leaf element name and the parent element name is represented as a leaf element list. Registered / output. In the case of the XML document shown in FIGS. 13A and 13B, a leaf element list similar to that shown in FIG. 2D is obtained.

  At this time, a new file name (for example, “.. \ book2.dtd”) different from the original file name is set for the new DTD obtained by changing and compressing the DTD shown in FIG. By filling in the document realization value, the file name specified by the system identifier “SYSTEM” of the document type declaration in the document realization value is changed from the old file name “.. \ book.dtd” to the new file name “. Rewrite as ". \ book2.dtd".

  Thereafter, according to the leaf element list (analysis result) obtained in step S21, the document realization value configuration changing unit 40 changes the element name and content of the leaf element to the attribute name and attribute of the parent element, respectively. The value is moved / changed to a value, and the start tag of the parent element is changed to an empty element tag (step S24). At this time, the attribute “year” attached to the leaf element “author” replaces the attribute of the parent element “book”.

  As a result, the XML document shown in FIG. 13A is changed and compressed into an XML document as shown in FIG. In the XML document shown in FIG. 13C, the description on the first line is the same as before compression, but the file name specified by the system identifier “SYSTEM” on the second line is the new file name “.. \ book2.dtd”. In the third line, the descriptions of the third to sixth lines in FIG. 13A are collectively described. That is, in the XML document shown in FIG. 13C, as in the example shown in FIG. 3B, all the information related to the two child elements (leaf elements) “title” and “author” In the start tag (empty element tag) of “book”, it is described as an attribute of the parent element “book”.

  Thereafter, the new DTD file creation unit 60 creates a new DTD file, copies the contents of the uncompressed DTD file read from the external file 100 to the DTD file (step S25), and then realizes the document realization value. In accordance with the configuration change of the document realization value by the configuration change unit 40, the DTD configuration in the new file is changed by the DTD configuration change unit 50 in the same manner as step S20 described above (step S26).

  As a result, the DTD shown in FIG. 13B is changed and compressed into a DTD as shown in FIG. In the DTD shown in FIG. 13D, the description of the content model is deleted from the element type declaration of the element “book” on the first line. The description on the 2nd to 5th lines is a summary of the description on the 2nd to 5th lines in FIG. 13B in the attribute list declaration of the element “book”.

  Then, the DTD file (new DTD file) changed / compressed by the DTD configuration changing unit 50 is given a new file name “.. \ book2.dtd”, and the new DTD file creating unit 60 sends the external file 100. Is output and stored (step S27).

  Also, the old and new DTD correspondence table output unit 70 creates a new and old DTD correspondence table in which the correspondence between the old DTD and the new DTD (specifically, the correspondence between the old file name and the new file name) is created and stored in the document storage. The XML document output / stored in the unit 10 or the like (step S28), and the XML document changed / compressed in step S24 is output / stored as a compressed document in the document storage unit 10 or the like (step S29). At that time, the old and new DTD correspondence table may be added to the compressed document in the form of an annotation instead of an independent file.

  In the first embodiment, since it is not necessary to restore (decompress) the compressed XML document to the original state, it is not always necessary to save the original DTD or create the old and new DTD correspondence table. That is, in the first embodiment, the processing of steps S25 and S27 by the new DTD file creation unit 60 and the new DTD file creation unit 60, and the steps by the old and new DTD correspondence table output unit 70 and the old and new DTD correspondence table output unit 70. It is also possible to omit the process of S28.

  However, the original DTD and the old and new DTD correspondence tables are required when restoring (expanding) an XML document compressed as described later in the second embodiment. The compression apparatus of the first embodiment has the original DTD storage function and the old and new DTD correspondence table creation function as described above, and the tag dictionary creation function described later with reference to FIG. The compression method according to the second embodiment is also configured to be realized.

Next, operations of the respective units 20, 30, 40, and 50 constituting the compression device of the first embodiment will be described with reference to FIGS.
First, according to the flowchart (steps S31 to S43) shown in FIG. 6, the analysis procedure by the document actual value analysis unit 20 according to the first embodiment will be described. The document real value analysis unit 20 sets the document real value to be compressed to the end. While determining whether or not the document has been scanned (step S31), the document realization value is scanned (step S32), and the description of the document realization value is sequentially recognized from the top, and it is checked whether or not “<” is described. (Step S33). Note that “<” is not described in the content of the document realization value in the XML specification.

  When “<” is detected as the description of the document realization value (YES route in step S33), whether the tag starting with “<” is a start tag or an end tag is determined based on the description of 1 byte following “<”. Determination is made (step S34). The determination is made based on whether or not the description following “<” is “/”. That is, if the description following “<” is “/”, the tag is determined to be an end tag, and if the description following “<” is not “/”, the tag is determined to be a start tag. Is done.

  In the case of a start tag (YES route in step S34), the element name and attribute name described in the start tag are detected and registered in the element name list and attribute name list, respectively (steps S35 and S36). . At that time, the appearance frequency of element names and attribute names is also totaled. This appearance frequency is used when creating a tag dictionary required in the second embodiment. In some cases, the attribute name is not described in the start tag. In this case, the attribute name is not detected, and thus the process of step S36 is omitted. After completing the registration to the list, the process returns to step S31.

On the other hand, in the case of an end tag (NO route in step S34), the element name described in the end tag is detected (step S37), and the element name is the last element name registered in the element name list. (Step S38).
At this time, if a description of a child element exists in the content of the document realization value of the element enclosed by the end tag (hereinafter referred to as the element of interest), the element name in the end tag and the last of the element name list Does not match element name. If there is no child element description in the document realization value of the element of interest, that is, if the element of interest is a leaf element, the element name in the end tag and the last element name in the element name list are: Match.

Therefore, if it is determined in step S38 that the element names do not match (NO route), the element of interest has a child element and is not a leaf element, the process returns directly to step S31.
On the other hand, if it is determined in step S38 that the element names match (YES route), it is possible to determine that the element of interest is a leaf element that does not have any child elements. It is determined whether or not an entity reference for an external file is described (step S39).

The content of the leaf element is handled as the attribute of the parent element by the compression conversion of the first embodiment. However, as described above, the entity reference for the external file cannot be used in the attribute value according to the XML specification.
Therefore, when it is determined in step S39 that the entity reference is described (YES route), the process directly returns to step S31. That is, a leaf element having an entity reference to an external file is not registered in the “leaf element list”.

  On the other hand, if it is confirmed in step S39 that no entity reference is described in the content of the element of interest (NO route), the element name of the element of interest is displayed as the leaf element name in the “leaf element list”. While registering / adding, the element name of the parent of the leaf is also registered / added (step S40). Thereafter, the process returns to step S31.

  If it is determined in step S31 that the document realization value to be compressed has been scanned to the end (YES route), the appearance frequency is determined based on the appearance frequency of the element name and attribute name that appeared during the scan of the document realization value. A tag dictionary (see reference numeral 90 in FIG. 14) that associates a high element name or attribute name with a shorter character string (for example, 1 byte; abbreviated character string) is generated and output (steps S41 and S42), and finally The obtained “leaf element list” [see, for example, FIG. 2D] is output (step S43), and the process ends.

  In the first embodiment, a tag dictionary is created by the document realization value analysis unit 20, but this tag dictionary is not used when executing the compression method of the first embodiment, and is described later. It is used in the embodiment. Therefore, in the first embodiment, steps S41 and S42 may be omitted. In the second embodiment, the processes in steps S41 and S42 are described as being performed by the tag dictionary creation unit 80 (see FIG. 14), not the document actual value analysis unit 20.

  Next, the analysis procedure performed by the DTD analysis unit 30 according to the first embodiment will be described with reference to the flowchart shown in FIG. 7 (steps S51 to S58). The DTD analysis unit 30 has scanned the DTD to be changed to the end. (Step S51), the DTD is scanned (step S52), the DTD descriptions are sequentially recognized from the top, and it is checked whether or not “<! ELEMENT” is described (step S53).

  For example, as shown in the second line of FIG. 4A, in the element type declaration, an element name and a content model (element name of a child element) are described after “! <ELEMENT”. In the content model, if only a reserved word such as “#PCDATA” is described and no unique child element name is registered, the element type declaration is for a leaf element. .

  Therefore, when “! <ELEMENT” is detected as a description of DTD in step S53 (YES route), an element character string (element name) following “! <ELEMENT” is detected (step S54), and thereafter Next, the description of the content model described is checked (step S55), and it is determined whether or not a child element name is described in the content model (step S56).

  If the element name is not described in the content model (NO route of step S56), the element type declaration of interest is determined to be related to the leaf element, and the element name in the element type declaration is determined. After registering (detected in step S54) together with the parent element name in the “leaf element list” (step S57), the process returns to step S51.

When “! <ELEMENT” is not detected as the DTD description in step S53 (NO route), or when it is determined in step S56 that the child element name is described in the content model (YES route). Returns to step S51 without performing registration processing in the “leaf element list”.
If it is determined in step S51 that the configuration change target DTD has been scanned to the end (YES route), the finally obtained “leaf element list” (see, for example, FIG. 2D) is output. (Step S58), the process is terminated.

  However, even if the content type (for example, #PCDATA) is defined by the content model in the DTD, it cannot be recognized from the description of the DTD whether the content includes an entity reference. That is, unlike the document realization value analysis unit 20 described above, the DTD analysis unit 30 recognizes whether or not an entity reference is included in the content of the document realization value described according to the DTD only by analyzing the DTD. Of course, it is not possible to distinguish whether the entity reference is for a document or an external file.

  The configuration change procedure performed by the document actual value configuration changing unit 40 according to the first embodiment will be described with reference to the flowchart shown in FIG. 8 (steps S61 to S72). Or after inputting the “leaf element list” obtained by the DTD analysis unit 30 (step S61), it is determined whether or not the document realization value to be compressed has been scanned to the end (step S62). The actual value is scanned (step S63).

At that time, the description of the document actual value is sequentially recognized from the top while comparing the description of the document actual value with the leaf element name registered in the “leaf element list”. It is determined whether or not it is a leaf element registered in "" (step S64).
When a leaf element registered in the “leaf element list” is detected in the document realization value (YES route), it is determined whether or not the leaf element has an attribute (step S65). (YES route), the attribute, that is, the character string of the attribute name and attribute value is registered in the attribute name list and attribute value list, respectively (step S66).

If the attribute information is not included (NO route in step S65), or after the registration process in step S66 ends, the element name and content character string of the leaf are registered in the attribute name list and attribute value list, respectively. (Step S67).
Then, the start tag, contents, and end tag for the leaf element for which the registration process in steps S66 and S67 has been completed are deleted from the document realization value (step S68), and the process returns to step S62.

If no leaf element is detected in step S64 (NO route), the process returns to step S62 without performing steps S65 to S68.
If it is determined in step S62 that the document realization value to be compressed has been scanned to the end (YES route), the leaf parent element is detected from the “leaf element list” (step S69), and the start tag of the parent element is detected. The attribute name and attribute value respectively registered in the attribute name list and the attribute value list are newly added (step S70).

  Thereafter, the end tag “<parent element name />” of the parent element is deleted (step S71), and “/” is entered before “>” described at the end of the parent start tag. As a result, the start tag of the parent element of the leaf is changed to an empty element tag (step S72), and the process ends.

  The configuration change procedure by the DTD configuration change unit 50 according to the first embodiment will be described with reference to the flowchart shown in FIG. 9 (steps S81 to S89). First, the DTD configuration change unit 50 includes the document realization value analysis unit 20 and the DTD analysis unit. After inputting the “leaf element list” obtained in 30 (step S81), the DTD is scanned while determining whether or not the DTD to be changed is scanned to the end (step S82) (step S82). S83).

At this time, it is determined whether or not an element type declaration having a leaf element name registered in the “leaf element list”, that is, “<! ELEMENT leaf element name” is described (step S84).
If such a leaf element type declaration is described (YES route of step S84), after deleting the leaf element type declaration from the DTD (step S85), an attribute list declaration having the leaf element name; That is, it is determined whether or not “<! ATTLIST leaf element name attribute name” is described (step S86).

  If such a leaf attribute list declaration is described (YES route of step S86), after deleting the leaf attribute list declaration from the DTD (step S87), in the element type declaration of the parent element for that leaf The description of the leaf (child element) is deleted from the description of the content model (step S88).

  Then, in the attribute list declaration for the leaf parent element, the element name and content for the leaf element deleted in step S85 are added as new attribute names and attribute values, respectively, and the leaf element deleted in step S87. The attribute name and the attribute value are added as new attribute names and attribute values (step S89), and the process returns to step S82. At this time, if the attribute list declaration for the parent element does not exist before the configuration change, a new attribute list declaration is created.

If it is determined in step S84 that the leaf element type declaration is not described (NO route), the process returns to step S82. If it is determined in step S86 that the leaf attribute list declaration is not described (NO route), the process proceeds to step S88.
If the NO route in step S86 or the description of the leaf has already been deleted from the content model when the process proceeds from step S87 to step S88, the process proceeds to step S89 without performing any processing in step S88.

  Further, when the process proceeds to step S88 via the NO route of step S86, in the attribute list declaration for the parent element of the leaf, the element name and content for the leaf element deleted in step S85 are newly set. After adding the attribute name and attribute value, the process returns to step S82. At this time, if the attribute list declaration for the parent element does not exist before the configuration change, a new attribute list declaration is created.

  As described above, according to the first embodiment of the present invention, the tree structure of the element in the document realization value is analyzed, and the start tag, end tag, and content for the leaf element are deleted from the document realization value according to the analysis result. By adding the element name, content, attribute name, and attribute value of the leaf element as the parent element attribute in the start tag of the parent element, the description of the leaf element can be handled as the attribute of the parent element. There is no need to describe the start tag and end tag of the element, and the description of the tag relating to the leaf element is omitted / compressed while maintaining the searchable state without impairing the characteristics of the XML document.

  Therefore, the compression rate of the XML document can be significantly increased, and consequently, the storage efficiency of the document data can be greatly increased in a system that handles a large-scale database. In particular, when a parts list, price list, or the like having a large number of short phrases is described in an XML document, the paired expression between the start tag and the end tag with the short phrase (content) sandwiched can be omitted. The compression rate can be greatly increased.

At this time, even after compression, plain text with a short data length can be handled as a search target, and there is no need to restore (decompress) the XML document when performing a search.
In addition, since the characteristics of the XML document are not impaired, it is possible to easily match with application software such as a browser.
Furthermore, by deleting the end tag of the parent element and changing the start tag of the parent element to an empty element tag, the compression rate of the XML document can be further increased.

  Similarly, the tree structure of the element in DTD is analyzed, and the element type declaration and attribute list declaration of the leaf element are deleted from the DTD according to the analysis result, and the description related to the leaf element is changed to the element type declaration of the parent element (content model) , And redefining the information related to the element type declaration and attribute list declaration of the leaf element as the attribute of the parent element, the compression processing corresponding to the compression performed on the document realization value is applied to the DTD. The leaf element is described as an attribute of the parent element. Therefore, since the description of the element type declaration and the attribute list declaration concerning the leaf element is omitted and the DTD is compressed while maintaining the searchable state without losing the characteristics of the XML document, the XML document is compressed. The rate can be increased further.

[2] Description of Second Embodiment Next, a second embodiment of the present invention will be described.
First, the compression principle of a structured document in the second embodiment of the present invention will be described with reference to FIGS. 15 (A) to 15 (D). In the second embodiment of the present invention, the case where the structured document is an XML document will be described.

In the second embodiment, each character string of the element name and attribute name in the tag of the document realization value or in the DTD is replaced with a 1- or 2-byte character string, and the correspondence relationship is replaced with a tag dictionary (see reference numeral 90 in FIG. ).
Usually, a character string (element name or attribute name) described in a tag is defined by DTD using a character string having a length of several bytes or more so that a person can read and understand the meaning.

However, the first character of element names and attribute names is limited to 1-byte alphabetic characters (AZ, az) in SGML. On the other hand, in XML, the first character is limited to 1-byte alphabet, 2-byte hiragana or katakana, and 1-byte “#” or “:”.
In general, only the tag portion of the document realization value occupies 60% to 80% of the total document amount. For this reason, it is possible to greatly increase the compression rate of the XML document only by converting the character string into a 1- or 2-byte character string at the expense of the readability of the character string in the tag.

  Therefore, in the second embodiment of the present invention, as shown in FIGS. 15A to 15D, existing names of element names and attribute names and newly defined 1- or 2-byte shortened character strings are provided. Is recorded in the tag dictionary, and the corresponding character string in the tag of the document realization value and in the DTD is replaced with a shorter shortened character string based on the tag dictionary. This shortened character string is naturally shorter than the character string of the existing name and must be able to specify the character string.

FIG. 15A is a diagram showing a specific description example of the tag. In the tag shown in FIG. 15A, the attribute name “tsprint” and the attribute value “school” are assigned to the element with the element name “title”. An attribute having “CAI series No. 563 3 (3)” is given.
At this time, as shown in FIG. 15B, “a” is preset as a replacement character (shortened character string) of the element name “title” and registered in the tag dictionary, and also shown in FIG. 15C. As described above, “A” is preset as a replacement character (shortened character string) of the attribute name “tsprint” and registered in the tag dictionary.

  Then, by using the tag dictionary shown in FIG. 15 (B) or FIG. 15 (C), in the tag shown in FIG. 15 (A), the character strings of element names and attribute names are as shown in FIG. 15 (D). Replace with a 1-byte character string. At this time, in the verified XML document having the DTD (patterns (4) and (5)), the DTD is also converted corresponding to the character string replacement described above.

  Therefore, an SGML parser, an XML parser (processor) or the like analyzes the structured document converted as described above based on the DTD converted as described above. However, when searching for elements and attributes on the application software side, the element name and attribute name must be specified using a 1- or 2-byte shortened character string read from the converted DTD.

Hereinafter, the second embodiment of the present invention will be described in more detail and specifically with reference to FIGS. 14 and 16 to 24.
FIG. 14 is a block diagram showing the functional configuration of a structured document compression apparatus according to the second embodiment of the present invention. As shown in FIG. 14, the compression apparatus of the second embodiment is the same as that of the first embodiment. In addition to the same document storage unit 10, document realization value analysis unit 20, DTD analysis unit 30, new DTD file creation unit 60, and old and new DTD correspondence table output unit 70, tag dictionary creation unit 80, tag dictionary 90, document realization value A character string replacing unit 41 and a DTD character string replacing unit 51 are included.

Here, similarly to the first embodiment, the compression apparatus of the second embodiment is also realized by a computer system such as a personal computer configured by connecting a CPU, a RAM, a ROM, and the like through a bus line.
In other words, the RAM and ROM function as the document storage unit 10, and the RAM includes the document realization value analysis unit 20, the DTD analysis unit 30, the new DTD file creation unit 60, the old and new DTD correspondence table output unit 70, and the tag dictionary. An application program for realizing the creating unit 80, the document actual value character string replacement unit 41, and the DTD character string replacement unit 51 is stored. The tag dictionary 90 is recorded / stored on, for example, a RAM.

  When the CPU executes the application program, the document realization value analysis unit 20, the DTD analysis unit 30, the new DTD file creation unit 60, the old and new DTD correspondence table output unit 70, the tag dictionary creation unit 80, the document realization value. Functions as the character string replacement unit 41 and the DTD character string replacement unit 51 (details will be described later) are realized, and the structured document compression apparatus according to the second embodiment is realized.

  The program for realizing the compression apparatus of the second embodiment is also provided in a form recorded on a computer-readable recording medium such as a flexible disk or a CD-ROM, as in the first embodiment. Then, the computer reads the program from the recording medium, transfers it to the internal storage device or the external storage device, and uses it. The program may be recorded in a storage device (recording medium) such as a magnetic disk, an optical disk, or a magneto-optical disk, and provided from the storage device to the computer via a communication path.

  When the function as the compression device of the second embodiment is realized by a computer, the program stored in the internal storage device (for example, RAM) is executed by the microprocessor (for example, CPU) of the computer. At this time, the program recorded on the recording medium may be directly read and executed by the microprocessor.

In the compression apparatus of the second embodiment shown in FIG. 14, the document storage unit 10, the document realization value analysis unit 20, the DTD analysis unit 30, the new DTD file creation unit 60, and the old and new DTD correspondence table output unit 70 Since it performs substantially the same function as that described in the embodiment, its detailed description is omitted.
However, the document realization value analysis unit 20 of the first embodiment has a function as the tag dictionary creation unit 80 of the second embodiment, but corresponds to steps S41 and S42 of FIG. 6 in the second embodiment. The part that performs the processing will be described as a tag dictionary creation unit 80 functionally separated from the document realization value analysis unit 20.

  This tag dictionary creation unit 80, in accordance with the analysis results by the document realization value analysis unit 20 and the DTD analysis unit 30, from the character string (element name, attribute name) described in the tag of the document realization value and in the DTD and the character string. The tag dictionary 90 is made to correspond to a shortened character string (one or two-byte character string described above) that is shorter and can specify the character string.

  As described in the first embodiment with reference to FIG. 6, the tag dictionary creation unit 80 of the present embodiment creates the tag dictionary 90 using only the analysis result by the document realization value analysis unit 20. However, the tag dictionary 90 may be created using the analysis result by the DTD analysis unit 30 instead of the analysis result by the document actual value analysis unit 20, or the analysis result by the document realization value analysis unit 20 and the DTD analysis unit The tag dictionary 90 may be created by using both of the analysis results obtained by 30.

  The document actual value character string replacement unit 41 uses the tag dictionary 90 to replace a character string (element name, attribute name) described in the tag of the document actual value with a shortened character string corresponding to the character string. The replacement procedure will be described later with reference to the flowchart shown in FIG.

  When the XML document is a verified XML document (that is, when the XML document is the pattern (4) or (5)), the DTD character string replacement unit 51 converts the DTD description into the document realization value character string replacement unit 41. The character string (element name, attribute name) described in the DTD is replaced with a shortened character string corresponding to the character string using the tag dictionary 90 to match the description of the document realization value replaced by The replacement procedure will be described later with reference to the flowchart shown in FIG.

In the DTD character string replacement 51, when the XML document is the pattern (4), the DTD is read from the document storage unit, but when the XML document is the pattern (5), the DTD is read from the external file 100.
Also in the second embodiment, when the XML document is the pattern (4), the replacement result (compressed DTD) by the DTD character string replacement unit 51 is replaced with the replacement result (compressed by the document actual value replacement unit 41). Output / stored in the document storage unit 10 together with the document actual value), but may be output / stored together with the compressed document actual value in another recording medium or the like.

When the DTD exists in the external file 100 (when the XML document is the pattern (5)), the new DTD file creation unit 60 creates a file (new DTD file) for the DTD replaced by the DTD character string replacement unit 51 Is generated and output to the external file 100.
When the DTD exists in the external file 100 (when the XML document is the pattern (5)), the old and new DTD correspondence table output unit 70 clarifies the correspondence between the DTD before replacement processing and the new DTD after replacement processing. An old and new DTD correspondence table (see, for example, FIG. 24G) is created and output to the document storage unit 10.

As described above, the XML document compressed by the compression method (character string replacement) according to the second embodiment does not lose the characteristics of the XML document at all, and remains in the compressed state (without expansion) as an XML document. Can fulfill the functions of
At this time, if the tag dictionary 90 is held, by referring to the tag dictionary 90 and recognizing the correspondence between the character string before replacement and the shortened character string, replacement / compression is performed in the document realization value or DTD. The data in the XML document can be searched without expanding the character string.

In addition, in order to decompress / restore the description of the XML document replaced with the shortened character string as described above, the decompression configured to include the document realization value character string reverse replacement means and the DTD character string reverse replacement means. A device (not shown) is provided.
Here, the document actual value character string reverse replacement means performs a reverse process opposite to the document actual value character string replacement unit 41 described above, and is described in the tag of the document actual value using the tag dictionary 90. The abbreviated character string is replaced with the original character string (element name, attribute name), and the DTD character string reverse replacement means performs a reverse process opposite to the DTD character string replacement unit 51 described above. Using the tag dictionary 90, the abbreviated character string described in the DTD is replaced with the original character string (element name, attribute name).

Next, a second embodiment will be described with reference to FIGS.
First, a procedure for compressing a structured document (XML document) in the second embodiment will be described with reference to the flowchart shown in FIG. 16 (steps S111 to S129). Note that the circled numbers 1 to 5 in FIG. 16 correspond to the patterns (1) to (5) described above with reference to Table 2, respectively.
Although not shown in FIG. 14, the XML document stored in the document storage unit 10 is stored in any of the patterns (1) to (5) (see Table 2) in the compression apparatus of the second embodiment. It has a pattern recognition function for recognizing it. The processing by this pattern recognition function corresponds to the processing by steps S112 to S114 shown in FIG.

  When an XML document to be compressed is input and stored in the document storage unit 10 (step S111), it is determined whether or not “<! DOCTYPE” is described in the XML document (step S112) and is not described. If this is the case (NO route of step S112), the XML document is recognized as a well-formed XML document having no DTD, that is, an XML document of pattern (1), and steps S115, S116, and S129 are executed as described later. .

When “<! DOCTYPE” is described in the XML document (YES route in step S112), it is determined whether or not “[” is described thereafter (step S113).
When “<! DOCTYPE” is described but “[” is not described (NO route of step S113), the XML document is a verified XML document having the DTD as the external file 100, that is, the pattern (5). As described later, steps S121 to S129 are executed.

If “[” is described (YES route in step S113), it is determined whether “<! ELEMENT” (or “<! ATTLIST”) is described (step S114).
If "<! DOCTYPE" and "[" are described but "<! ELEMENT" is not described (NO route of step S114), a well-formed XML document having a DTD containing an internal or external entity declaration That is, it is recognized that the document is an XML document of the pattern (2) or (3), and steps S115, S116, and S129 are executed as in the case of the pattern (1).

When “<! DOCTYPE”, “[”, and “<! ELEMENT” are all described (YES route in step S114), a verified XML document having a DTD in the XML document, that is, the XML of the pattern (4) The document is recognized and steps S117 to S120 and S129 are executed.
Hereinafter, the compression process for each pattern (1) to (5) will be described with reference to specific examples (first to fifth examples) shown in FIGS.

FIGS. 20A to 20C are diagrams for explaining specific compression processing (first example) of an XML document according to the second embodiment.
The XML document before compression shown in FIG. 20A is an XML document of pattern (1), which is the same as that shown in FIG. Since “<! DOCTYPE” is not described in the XML document (pattern (1)) shown in FIG. 20A, the process proceeds from the NO route in step S112 to step S115, and the document realization value analysis unit 20 Thus, the description in the tag of the document actual value is analyzed, and the tag dictionary creating unit 80 registers and creates a tag dictionary 90 as shown in FIG.

  Here, in the example shown in FIG. 20A, only the element name is described in each tag of the document realization value, and no element has an attribute. Therefore, only the element name is detected, and each element name A tag dictionary 90 that associates a shortened character string with is registered and created. A tag dictionary 90 shown in FIG. 20C includes a short 1-byte shortened character string (hereinafter referred to as replacement) for the element names “book”, “title”, and “author” detected and recognized by the document realization value analysis unit 20. For example, “a”, “b”, and “c” are associated with each other.

If an attribute name is also described in the tag, a tag dictionary 90 for the attribute name is also registered / created as described later with reference to FIGS. 23 (D) and 24 (E).
Then, using the tag dictionary 90 obtained in step S115, the document actual value character string replacement unit 41 uses the element names “book”, “title”, and “author” described in the document real value tag to Each is replaced with 1-byte replacement characters “a”, “b”, and “c” (step S116).

Through steps S115 and S116, for example, the XML document shown in FIG. 20A is replaced and compressed with an XML document as shown in FIG. 20B, and then output and stored as a compressed document in the document storage unit 10 or the like. (Step S129).
In the XML document shown in FIG. 20B, the description of the XML declaration in the first line is the same as that before compression, but in the second to fifth lines, the element names “book”, “ “title” and “author” are replaced with 1-byte replacement characters “a”, “b”, and “c”, respectively.

FIGS. 21A to 21C are diagrams for explaining specific compression processing (second example) of an XML document according to the second embodiment.
The XML document before compression shown in FIG. 21A is an XML document of pattern (2), which is the same as that shown in FIG. In the XML document (pattern (2)) shown in FIG. 21A, both “<! DOCTYPE” and “[” are described, but “<! ELEMENT” and “<! ATTLIST” are described. Accordingly, the process proceeds from the NO route in step S114 to step S115, and the same process as the XML document having the pattern (1) described above is executed. At this time, the tag dictionary creation unit 80 registers and creates the same tag dictionary 90 as shown in FIG. 20C, as shown in FIG.

Thus, for example, the XML document shown in FIG. 21A is replaced and compressed with an XML document as shown in FIG. 21B, and then output and stored as a compressed document in the document storage unit 10 or the like (step) S129).
In the XML document shown in FIG. 21B, the descriptions of the first, third, and fourth lines are the same as before compression, but the element names “book”, “title”, “ “author” is replaced with 1-byte replacement characters “a”, “b”, and “c”, respectively.

22A to 22C are diagrams for explaining specific compression processing (third example) of an XML document according to the second embodiment.
The XML document before compression shown in FIG. 22A is an XML document of the pattern (3) described above, and an XML declaration indicating that this document is a version 1.0 XML document is described in the first line. The DTD including the entity declaration to the outside is described in the 2nd to 4th lines, and the document realization value is described in the 5th to 8th lines.

In the DTD on the second to fourth lines, the entity of the character string “para” used in the document realization value (XML instance) using the system identifier “SYSTEM” in the entity declaration (line 3) included in the document type declaration As described above, the use of an external file specified by the URL “http://www.xml.co.jp” is declared and defined.
The document realization value described in the 5th to 8th lines is almost the same as the description example in the 2nd to 5th lines of the XML document shown in FIG. 10 (A), but the example shown in FIG. 22 (A). Then, “Moto Sato &para;” is described as the content of the element “author” on the seventh line.

  Here, “&para;” is a description instructing to refer to the substance of the character string “para”. In a document that is actually output by display / printing, the URL “http: //www.xml The external file specified by “.co.jp” will be read and displayed.

  In the XML document (pattern (3)) shown in FIG. 22A, both “<! DOCTYPE” and “[” are described as in the XML document of pattern (2), but “< Since “ELEMENT” and “<! ATTLIST” are not described, the process proceeds from the NO route of step S114 to step S115, and the same process as the XML document of the pattern (1) or (2) described above is executed. The At this time, the tag dictionary creating unit 80 registers and creates the same tag dictionary 90 as shown in FIG. 20C, as shown in FIG.

Thus, for example, the XML document shown in FIG. 22A is replaced and compressed with the XML document as shown in FIG. 22B, and then output and stored as a compressed document in the document storage unit 10 or the like (step). S129).
In the XML document shown in FIG. 22B, the descriptions of the first, third, and fourth lines are the same as before compression, but the element names “book”, “title”, “ “author” is replaced with 1-byte replacement characters “a”, “b”, and “c”, respectively.

FIGS. 23A to 23D are diagrams for explaining specific compression processing (fourth example) of an XML document according to the second embodiment.
The XML document before compression shown in FIG. 23A is an XML document of pattern (4), which is almost the same as that shown in FIG. However, in the XML document shown in FIG. 23A, the description about the attribute “year” of the element “author” is omitted. That is, in the DTD, the attribute list declaration for the attribute “year” is omitted, and the attribute description in the start tag of the element “author” is omitted.

  In the XML document (pattern (4)) shown in FIG. 23A, “<! DOCTYPE” and “[” are described and “<! ELEMENT” or “<! ATTLIST” is also described. The process proceeds from the YES route in step S114 to step S117, the description in the tag of the document actual value is analyzed by the document actual value analysis unit 20, and the description of the DTD is analyzed by the DTD analysis unit 30 (step S118). ).

At this time, the tag dictionary creation unit 80 generates a tag dictionary 90 for element names as shown in FIG. 23C and a tag dictionary 90 for attribute names as shown in FIG. Registered / created.
Here, the tag dictionary 90 shown in FIG. 23C is the same as that shown in FIG. 20C, and the element names “book”, detected and recognized by the document realization value analysis unit 20 and the DTD analysis unit 30. “Title” and “author” correspond to 1-byte shortened character strings, for example, “a”, “b”, and “c”, respectively. In addition, the tag dictionary 90 shown in FIG. 23D adds a 1-byte shortened character string such as “A” to the attribute name “field” detected and recognized by the document realization value analysis unit 20 and the DTD analysis unit 30. It is something to be supported.

  Then, using the tag dictionary 90 shown in FIG. 23C and FIG. 23D, the document realization value character string replacement unit 41 uses the element names “book” and “title” described in the document realization value tag. , “Author” and attribute name “field” are replaced with 1-byte shortened character strings “a”, “b”, “c”, and “A”, respectively (step S119), and the document actual value character string conversion unit 41 The element name “book”, “title”, “author” and attribute name “field” described in the DTD are each shortened by 1 byte by the DTD character string replacement unit 51 in accordance with the character string replacement of the document realization value by The columns “a”, “b”, “c”, and “A” are replaced (step S120).

Thus, for example, the XML document shown in FIG. 23A is replaced and compressed with an XML document as shown in FIG. 23B, and then output and stored as a compressed document in the document storage unit 10 or the like (step) S129).
In the XML document shown in FIG. 23B, the descriptions in the first and seventh lines are the same as before compression, but the element names “book”, “title”, “author” and the like in the second to sixth and eighth to eleventh lines The attribute name “field” is replaced with 1-byte replacement characters “a”, “b”, “c”, and “A”, respectively.

FIGS. 24A to 24G are diagrams for explaining specific compression processing (fifth example) of an XML document according to the second embodiment.
The XML document before compression shown in FIG. 24A is an XML document of pattern (5), which is almost the same as that shown in FIG. However, in the XML document shown in FIG. 24A, the description about the attribute “year” of the element “author” is omitted. That is, in the DTD, the attribute list declaration for the attribute “year” is omitted, and the attribute description in the start tag of the element “author” is omitted.

In the DTD document type declaration on the second line, it is declared and defined that the DTD (file name “.. \ book.dtd”) held in the external file 100 is used by the system identifier “SYSTEM”.
The DTD of the file name “.. \ book.dtd” is described as shown in FIG. 24B corresponding to the configuration of the document realization value in FIG. Since the DTD (1st to 4th lines) shown in FIG. 24B is the same as the description example of the 2nd to 5th lines in the DTD shown in FIG.

  In the XML document (pattern (5)) shown in FIG. 24A, “<! DOCTYPE” is described, but after that “[” is not described, the DTD in the external file 100 is specified. Since the system identifier to be processed is described, the process proceeds from the NO route of step S113 to step S121, the description in the tag of the document realization value is analyzed by the document realization value analysis unit 20, and the DTD analysis unit 30 performs the analysis. The description of the DTD (file name “.. \ book.dtd”) read from the external file 100 according to the system identifier is analyzed (step S122).

  At this time, the tag dictionary creation unit 80 generates a tag dictionary 90 for element names as shown in FIG. 24E and a tag dictionary 90 for attribute names as shown in FIG. Registered / created. Here, the tag dictionary 90 shown in FIG. 24 (E) is the same as that shown in FIG. 20 (C), and the tag dictionary 90 shown in FIG. 24 (F) is the same as that shown in FIG. 23 (D). It is.

  After that, a new file name (for example, “.. \ book2.dtd”) different from the original file name is set for the new DTD obtained by changing / compressing the DTD shown in FIG. Then, by filling in the document realization value (step S123), the file name specified by the system identifier “SYSTEM” of the document type declaration in the document realization value is updated from the old file name “.. \ book.dtd”. Rewrite the file name “.. \ book2.dtd”.

  Then, using the tag dictionary 90 shown in FIGS. 24 (E) and 24 (F), the element names “book” and “title” described in the document actual value tag by the document actual value character string replacement unit 41 are used. , “Author” and attribute name “field” are replaced with 1-byte shortened character strings “a”, “b”, “c”, and “A”, respectively (step S124).

  As a result, the XML document shown in FIG. 24A is replaced and compressed with an XML document as shown in FIG. In the XML document shown in FIG. 13C, the description on the first line is the same as before compression, but the file name specified by the system identifier “SYSTEM” on the second line is the new file name “.. \ book2.dtd”. And the element names “book”, “title”, “author” and attribute name “field” in the 2nd to 6th lines are replaced by 1-byte replacement characters “a”, “b”, “c”, “ Is replaced by “A”.

  Next, the new DTD file creation unit 60 creates a new DTD file, copies the contents of the uncompressed DTD file read from the external file 100 to the DTD file (step S125), and then realizes the document actual value character. The element names “book”, “title”, “author”, and attribute name “field” described in the DTD are each set to 1 by the DTD character string replacement unit 51 in accordance with the character string replacement of the document realization value by the column conversion unit 41. The byte shortened character strings “a”, “b”, “c”, and “A” are replaced (step S126).

As a result, the DTD shown in FIG. 24B is changed and compressed into a DTD as shown in FIG. In the DTD shown in FIG. 24D, the element names “book”, “title”, “author” and the attribute name “field” in the first to fourth lines are 1-byte character strings “a”, “b”, Replaced by “c” and “A”.
The DTD file (new DTD file) replaced / compressed by the DTD character string replacement unit 51 is given a new file name “.. \ book2.dtd”, and the new DTD file creation unit 60 sends an external file. The data is output to and stored in 100 (step S127).

  Also, the old and new DTD correspondence table output unit 70 generates an old and new DTD correspondence table in which the correspondence between the old DTD and the new DTD (specifically, the correspondence between the old file name and the new file name) is clearly shown in FIG. ) And output / stored in the document storage unit 10 or the like (step S128), and the XML document replaced / compressed in step S124 is output / stored as a compressed document in the document storage unit 10 or the like. (Step S129). At that time, the tag dictionary 90 and the old and new DTD correspondence table may be added to the compressed document in the form of annotations instead of independent files.

Next, operations of the document actual value character string replacement unit 41 and the DTD character string replacement unit 51 that constitute the compression apparatus according to the second embodiment will be described with reference to FIGS. 17 and 18.
First, according to the flowchart shown in FIG. 17 (steps S151 to S158), the replacement procedure by the document actual value character string replacement unit 41 according to the second embodiment will be described. First, the document real value string replacement unit 41 creates a tag dictionary. After the tag dictionary 90 obtained by the unit 80 is input (step S151), the document realization value start is performed in the same manner as the document realization value analysis unit 20 of the first embodiment (see steps S31 to S34 in FIG. 6). Tags and end tags are discriminated (steps S152 to S155), and element names and attribute names in those tags are replaced using the tag dictionary 90 (steps S156 to S158).

  That is, while determining whether or not the document realization value to be compressed has been scanned to the end (step S152), the document realization value is scanned (step S153), and the description of the document realization value is sequentially recognized from the head, and “< "Is described or not (step S154). Note that “<” is not described in the content of the document realization value in the XML specification.

  When “<” is detected as the description of the document realization value (YES route of step S154), whether the tag starting with “<” is a start tag or an end tag is determined based on the description of 1 byte following “<”. Determination is made (step S155). The determination is made based on whether or not the description following “<” is “/”. That is, if the description following “<” is “/”, the tag is determined to be an end tag, and if the description following “<” is not “/”, the tag is determined to be a start tag. Is done.

  In the case of a start tag (YES route of step S155), the tag dictionary 90 is referenced to replace the element name described in the start tag with the corresponding shortened character string (step S156). If an attribute name is described in the start tag, the attribute name is also replaced with a corresponding shortened character string with reference to the tag dictionary 90 (step S157). If no attribute name is described in the start tag, the process of step S157 is omitted. After completing such replacement processing, the process returns to step S152.

On the other hand, in the case of an end tag (NO route of step S155), the tag dictionary 90 is referred to, and the element name described in the end tag is replaced with the corresponding shortened character string (step S158), and then step Return to S152.
In step S152, if it is determined that the document realization value to be compressed has been scanned to the end (YES route), the process ends.

  The replacement procedure performed by the DTD character string replacement unit 51 according to the second embodiment will be described with reference to the flowchart shown in FIG. 18 (steps S161 to S170). The DTD character string replacement unit 51 is first obtained by the tag dictionary creation unit 80. After inputting the tag dictionary 90 (step S161), while determining whether or not the DTD to be compressed has been scanned to the end (step S162), the DTD is scanned (step S163), and an element type declaration, that is, “< It is checked whether or not “ELEMENT” is described (step S164).

  When “<! ELEMENT” is detected in step S164 (YES route), the element name in the element type declaration is replaced with the corresponding short character string with reference to the tag dictionary 90 (step S165), The content model described in the element type declaration is detected (step S166), and when the child element name is described in the content model, the element name is also dealt with by referring to the tag dictionary 90. To replace the abbreviated character string (step S167). Note that if there is no description of the content model or if no child element name is described in the content model, the process of step S167 is omitted.

  Thereafter, it is checked whether or not an attribute list declaration, that is, “<! ATTLIST” is described (step S168). When “<! ATTLIST” is detected in step S168 (YES route), the tag dictionary 90 is referred to, and the element name in the attribute list declaration is replaced with the corresponding short character string (step S169). After replacing the attribute name in the attribute list declaration with the corresponding short character string (step S170), the process returns to step S162.

If it is determined in step S164 that no element type declaration is described (NO route), or if it is determined in step S168 that no attribute list declaration is described (NO route), the process proceeds to step S162. Return.
In step S152, if it is determined that the document realization value to be compressed has been scanned to the end (YES route), the process ends.

  By the way, as described above, in the second embodiment, there is provided a decompression device (not shown) for decompressing / restoring the description of the XML document replaced with the shortened character string. The character realization value character string reverse replacement means and the DTD character string reverse replacement means constituting the decompressing device are the same as the character realization value character string replacement unit 41 and the DTD character string replacement unit 51, respectively, as shown in FIGS. The reverse replacement process is performed according to the flowchart shown in FIG. However, in the reverse replacement process, the character string conversion directions in steps S156 to S158 in FIG. 17 and steps S165, S167, S169, and S170 in FIG. 18 are reversed.

Next, a structured document decompression procedure according to the second embodiment, that is, a reverse replacement procedure by the decompression apparatus described above will be described with reference to the flowchart shown in FIG. 19 (steps S131 to S142).
Note that the decompression apparatus according to the second embodiment is also provided with a pattern recognition function for recognizing that the compressed document to be decompressed is any one of the patterns (1) to (5). The processing by this pattern recognition function corresponds to the processing by steps S133 to S135 shown in FIG.

  First, when a compressed document to be decompressed is input (step S131) and the tag dictionary 90 created during the compression process is input (step S132), is “<! DOCTYPE” described in the compressed document? If it is not described (NO route of step S133), the compressed document is recognized as a well-formed XML document having no DTD, that is, an XML document of pattern (1). As will be described later, steps S136 and S142 are executed.

When “<! DOCTYPE” is described in the compressed document (YES route in step S133), it is determined whether or not “[” is described thereafter (step S134).
When “<! DOCTYPE” is described but “[” is not described (NO route of step S134), the compressed document is a verified XML document having the DTD as the external file 100, that is, the pattern (5). As described later, steps S139 to S142 are executed.

If “[” is described (YES route in step S134), it is determined whether “<! ELEMENT” (or “<! ATTLIST”) is described (step S135).
When “<! DOCTYPE” and “[” are described but “<! ELEMENT” is not described (NO route of step S135), the compressed document includes a DTD including an internal or external entity declaration. It is recognized that it is a well-formed XML document, that is, an XML document of pattern (2) or (3), and steps S136 and S142 are executed as in the case of pattern (1).

When “<! DOCTYPE”, “[”, and “<! ELEMENT” are all described (YES route in step S135), the compressed document is a verified XML document having a DTD in the XML document, that is, a pattern. The document is recognized as an XML document (4), and steps S137, S138, and S142 are executed.
Hereinafter, the decompression process for each of the patterns (1) to (5) will be described with reference to specific examples (first to fifth examples) shown in FIGS.

  When the decompressed compressed document is a compressed XML document having a pattern (1) as shown in FIG. 20B, since “<! DOCTYPE” is not described in the document, the process proceeds to the NO route in step S133. From step S136, the 1-byte shortened character string “a” described in the document realization value tag is written by the document realization value character string reverse replacement means using the tag dictionary 90 shown in FIG. ”,“ B ”, and“ c ”are reversely replaced with the original element names“ book ”,“ title ”, and“ author ”, respectively. As a result, for example, the compressed XML document shown in FIG. 20B is reversely replaced / expanded into an XML document as shown in FIG. 20A, and output / stored as an expanded document in the document storage unit 10 or the like (step) S142).

  When the decompressed compressed document is a compressed XML document having a pattern (2) as shown in FIG. 21B, both “<! DOCTYPE” and “[” are described in the document. Since “<! ELEMENT” and “<! ATTLIST” are not described, the process proceeds from the NO route in step S135 to step S136, and the document realization value is determined using the tag dictionary 90 shown in FIG. The 1-byte abbreviated character strings “a”, “b”, and “c” described in the document realization value tags are converted into the original element names “book”, “title”, “ Reversely replaced with “author”. Thus, for example, the compressed XML document shown in FIG. 21B is reversely replaced / expanded into an XML document as shown in FIG. 21A, and output / stored as an expanded document in the document storage unit 10 or the like (step) S142).

  When the compressed document to be decompressed is a compressed XML document having a pattern (3) as shown in FIG. 22B, any one of “<! DOCTYPE” and “[” is included in the document as in the pattern (2). However, since “<! ELEMENT” and “<! ATTLIST” are not described, the process proceeds from the NO route of step S135 to step S136, and the above-described patterns (1) and (2) Processing similar to that for the compressed XML document is executed.

  That is, by using the tag dictionary 90 shown in FIG. 22C, the 1-byte shortened character strings “a” and “b” described in the document actual value tag by the document actual value character string reverse replacement means. , “C” are reversely replaced with the original element names “book”, “title”, and “author”, respectively. Thus, for example, the compressed XML document shown in FIG. 22B is reversely replaced / expanded into an XML document as shown in FIG. 22A, and output / stored as an expanded document in the document storage unit 10 or the like (step) S142).

  When the decompressed compressed document is a compressed XML document having a pattern (4) as shown in FIG. 23B, “<! DOCTYPE” and “[” are described in the document and “<! ELEMENT "Or" <! ATTLIST "is also described, the process proceeds from the YES route in step S135 to step S137, and the document is searched using the tag dictionary 90 shown in FIGS. 23 (C) and 23 (D). The 1-byte shortened character strings “a”, “b”, “c”, and “A” described in the tag of the document actual value are converted into the original element name “book” by the actual value character string reverse replacement means. , “Title”, “author” and the attribute name “field” are reversely substituted.

Further, by using the tag dictionary 90 shown in FIGS. 23C and 23D, the DTD character string reverse replacement means uses the 1-byte shortened character strings “a”, “b”, “C” and “A” are reversely replaced with the original element names “book”, “title”, “author” and attribute name “field”, respectively (step S138).
Thus, for example, the compressed XML document shown in FIG. 23B is reversely replaced / expanded into an XML document as shown in FIG. 23A, and output / stored as an expanded document in the document storage unit 10 or the like (step) S142).

  When the decompressed compressed document is a compressed XML document having a pattern (5) as shown in FIG. 24C, “<! DOCTYPE” is described in the document, but “[” is included thereafter. Since the system identifier for designating the DTD in the external file 100 is described, the process proceeds from the NO route in step S134 to step S139, and the old and new DTD tables shown in FIG. Then, according to the old and new DTD tables, the file name specified by the system identifier “SYSTEM” of the document type declaration in the document realization value is changed from “.. \ book2.dtd” to the original file name “.. \ book. dtd "(step S140).

Then, using the tag dictionary 90 shown in FIGS. 24E and 24F, the document realization value character string reverse replacement means uses the 1-byte shortened character string “ “a”, “b”, “c”, and “A” are reversely replaced with the original element names “book”, “title”, “author”, and attribute name “field”, respectively (step S141).
Thus, for example, the compressed XML document shown in FIG. 24C is reversely replaced / expanded into an XML document as shown in FIG. 24A, and output / stored as an expanded document in the document storage unit 10 or the like (step) S142).

  At this time, since the DTD with the file name “.. \ book.dtd”, that is, the DTD shown in FIG. 24B is stored in the external file 100, the compressed DTD shown in FIG. It is not necessary to perform reverse replacement / extension by the reverse replacement means to obtain the DTD shown in FIG.

  As described above, according to the second embodiment of the present invention, a tag dictionary 90 is created according to the analysis result in the tag or DTD description of the document realization value, and the document realization is performed using the tag dictionary 90. By replacing a character string described in a value tag or DTD with a shortened character string (1- or 2-byte replacement character), the character string in the tag or the DTD is compressed without impairing the characteristics or structure of the XML document. Therefore, the compression rate of the XML document can be greatly increased, and consequently, the storage efficiency of the document data can be significantly increased in a system that handles a large-scale database.

  At this time, by replacing the element name and attribute name described in the tag or in the DTD with a shortened character string, the tag portion or DTD can be compressed while being kept in a searchable state. In other words, the element name and the attribute name are replaced, the tag dictionary 90 is held, and the attribute value is held in its original form, so that the search and document structure can be maintained while the document data is compressed without being decompressed. It is possible to grasp.

  Therefore, when performing a search by analyzing the structure of an XML document after compression, there is no need to decompress the compressed XML document, and even if the document data is compressed and stored in a large-scale database, the document data can be searched. Processing and the like can be performed in a short time.

  Since the attribute name and the attribute value can each be a search target, when both the attribute name and the attribute value are integrally compressed, the search cannot be performed unless decompression / restoration is performed. Therefore, in the second embodiment, compression is performed by replacing only the attribute name with a shortened character string and the attribute value is left in its original form, so that application software such as a browser performs decompression processing of the compressed document. It is possible to search for data in the compressed document while referring to the tag dictionary 90 without performing it.

[3] Description of Third Embodiment Next, a third embodiment of the present invention will be described.
In the third embodiment of the present invention, as in the second embodiment, as shown in FIG. 30A, in the XML document before compression, an area enclosed by “<” and “>” (DTD declaration statement) Alternatively, the character string described in the document realization value (start tag and end tag) is converted into a shortened character string (replacement character) and compressed using a tag dictionary as shown in FIG. When such compression is performed, the description format using “<” and “>” remains stored in the compressed XML document, and the characteristics and structure of the XML document are lost even after compression. There is nothing.

  In the third embodiment, after identifying the type of language used for describing plaintext, a dictionary (word dictionary) of the corresponding language, for example, a Japanese dictionary as shown in FIG. The plain text is converted to a fixed byte length shortened character string (word number) using the longest match method. Here, plaintext refers to the contents described between the start tag and the end tag.

In this way, for example, an uncompressed XML document as shown in FIG. 30A can be converted and compressed into an XML document as shown in FIG. 30B.
In the third and fourth embodiments, the language type is identified from, for example, three types of Japanese, English, and Chinese.

  In addition, in the case of a well-formed XML document (pattern (1)) having no DTD, the description in the tag in the document realization value is checked, and then a character string such as an element name in the tag is used using the tag dictionary. In association with the abbreviated character string (replacement character), the XML specification identifies the language of plaintext (contents) by looking at the language identification attribute (xml: lang), and a dictionary (corresponding to the language) Word dictionary) is selected, and the same compression as described above is performed. A specific example in which the compression method of the third embodiment is applied to a well-formed XML document having no DTD will be described later with reference to FIGS. 32 (A) to 32 (D).

Hereinafter, a third embodiment of the present invention will be described with reference to FIGS.
First, how the XML document before compression will be after compression will be described with reference to FIGS.

  30A to 30D are diagrams for explaining specific compression processing of an XML document (pattern (4): verified XML document in which DTD is described) according to the third embodiment. 30A shows a description example before compression of an XML document in which DTD is described, FIG. 30B shows a description example after compression, and FIG. 30C shows a tag used for the compression processing. An example of registered contents of a dictionary is shown, and FIG. 30D shows an example of registered contents of a Japanese dictionary used for the compression processing.

  FIGS. 31A to 31G all illustrate a specific compression process of an XML document (pattern (5): an XML document that specifies and uses a DTD of another file) according to the third embodiment. FIG. 31A shows an example of a pre-compression description of an XML document that refers to and uses a DTD stored in another file, and FIG. 31B shows a pre-compression description of the DTD stored in the other file. FIG. 31C shows an example of description after compression of the XML document, FIG. 31D shows an example of description after compression of the DTD (DTD of a new file), and FIG. An example of registered contents of the tag dictionary used for the compression process is shown, FIG. 31F shows an example of registered contents of the Japanese dictionary used for the compression process, and FIG. 31G holds the correspondence between the old and new DTDs. An example of registered contents of the correspondence table is shown.

  32A to 32D are diagrams for explaining specific compression processing of an XML document (pattern (1): a well-formed XML document having no DTD) according to the third embodiment. 32A shows a description example before compression of a well-formed XML document having no DTD, FIG. 32B shows a description example after compression, and FIG. 32C is used for the compression processing. An example of registered contents of the tag dictionary is shown, and FIG. 32D shows an example of registered contents of the Japanese dictionary used for the compression processing.

  FIG. 33 is a diagram for explaining the compression method of the XML document in the third embodiment. This FIG. 33 shows the character string described in the area surrounded by “<” and “>” in the XML document. This is for explaining a conversion method to a replacement character when there is a blank space.

First, with reference to FIGS. 30A to 30D, compression processing for an XML document in which DTD is described will be described.
As shown in FIG. 30A, the XML document before compression includes an XML declaration shown on the first line, a DTD shown on the second to seventh lines, a plain text content shown on the eighth to twelfth lines, and various symbols. It is filled in.

  In FIG. 30A, the first line describes an XML declaration indicating that this document is a version 1.0 XML document, and the second line describes the document type name (DOCTYPE name) of this document. ) Is “book”, and the description between “[” and “]” on the seventh line defines the structure of this document.

  In addition, the name “book” of the top-level element name on the third line needs to match the document type name “book”. In the third line, it is described as a content model that the element “book” has a child element “chapter”, and in the fourth to sixth lines, the element “chapter” It is described that it is composed of two child elements “title” and “paragraph”, and “]>” on the seventh line describes the internal subset description of the document type declaration (DOCTYPE declaration, DTD description) starting from the second line. Shown to finish.

  In the subsequent 8th line, “<book>” is a start tag of the element “book”, and “</ book>” of the 12th line is an end tag of the element “book”. The description between these tags (lines 9 to 12) is the content of the element “book”, and the description of the ninth line is that the content of the child element “title” of the element “chapter” is “XML overview” The description on the 10th line indicates that the content of the child element “paragraph” of the element “chapter” is “What is XML?”. In the ninth line, “<title>” is a start tag indicating the start of the element “title”, and “</ title>” is an end tag indicating the end of the element “title”. “</ Chapter>” on the 11th line is an end tag indicating the end of the element “chapter”.

A character string in the XML document before compression shown in FIG. 30A is converted into a replacement character or a word number by using the tag dictionary shown in FIG. 30C and the Japanese dictionary shown in FIG. By converting, the XML document is compressed as shown in FIG.
Here, the tag dictionary shown in FIG. 30C is created by the method described later with reference to FIG. 29, and the Japanese dictionary shown in FIG. 30D is a static dictionary created in advance. is there.

  In the example of FIG. 30A, “chapter”, “title”, and “paragraph” are replaced and compressed with fixed-length replacement characters “b”, “c”, and “d”, respectively, by the tag dictionary described above. In addition, according to the above Japanese dictionary, “XML”, “no”, “outline”, and “to” of the plain text part are respectively fixed-length word numbers “α”, “β”, “γ”, Replacement / compression with “δ” or the like is performed to obtain a compressed XML document as shown in FIG.

Next, a compression method when referring to a DTD stored in a file different from the XML document will be described with reference to FIGS. 31 (A) to 31 (G).
In the XML document before compression shown in FIG. 31 (A), another file having the file name “book.dtd” by the underlined part of the second line, that is, the system identifier “SYSTEM“ ../book.dtd ””. DTD is specified. The DTD of this separate file is described as shown in FIG.

  Using the tag dictionary shown in FIG. 31E and the Japanese dictionary shown in FIG. 31F, the character string in the XML document before compression shown in FIG. By converting, the XML document is compressed as shown in FIG.

  Further, by converting the character string in the DTD shown in FIG. 31 (B) into a replacement character using the tag dictionary shown in FIG. 31 (E), the DTD is converted into a new one as shown in FIG. 31 (D). Compress and create as file DTD. For example, “book2.dtd” is given as the file name of the new file.

As shown in FIG. 31C, in the second line of the compressed XML document, the file name “book2.dtd” is changed to the file name “book” of the old file in order to specify the DTD of the new file. instead of “.dtd”. Then, the new DTD file name and the old DTD file name before compression are entered in the new and old DTD correspondence table as shown in FIG.
Here, the tag dictionary shown in FIG. 31 (E) is created by the method described later with reference to FIG. 29 as well as the tag dictionary shown in FIG. 30 (C), and the Japanese dictionary shown in FIG. 31 (F). The dictionary is a static dictionary created in advance.

  Also in the example of FIG. 31A, “chapter”, “title”, and “paragraph” are replaced and compressed with fixed-length replacement characters “b”, “c”, and “d” by the tag dictionary described above. In addition, according to the above Japanese dictionary, “XML”, “no”, “outline”, and “to” of the plain text part are respectively fixed-length word numbers “α”, “β”, “γ”, Substitution / compression with “δ” or the like is performed, and a compressed XML document as shown in FIG. 31C is obtained.

  Further, in the DTD shown in FIG. 31C, “chapter”, “title”, and “paragraph” are replaced with fixed-length replacement characters “b”, “c”, and “d”, respectively, by the tag dictionary described above. Compressed and stored as a new DTD in a new file with the file name “book2.dtd”.

Next, a compression process for a well-formed XML document having no DTD will be described with reference to FIGS. 32 (A) to 32 (D).
In the XML document before compression shown in FIG. 32A, the DTD document type definition starting with the DOCTYPE declaration following the XML declaration in the first line is not described.
Even in such a case, the character string in the XML document before compression shown in FIG. 32A is replaced by using the tag dictionary shown in FIG. 32C and the Japanese dictionary shown in FIG. It is converted into characters and word numbers, and the XML document is compressed as shown in FIG.

Here, the tag dictionary shown in FIG. 32C is also created by the method described later with reference to FIG. 29 in the same manner as the tag dictionary shown in FIGS. 30C and 31E, and FIG. The Japanese dictionary shown in D) is a static dictionary created in advance.
Also in the example of FIG. 32A, “chapter”, “title”, and “paragraph” are replaced and compressed with fixed-length replacement characters “b”, “c”, and “d”, respectively, by the tag dictionary described above. In addition, according to the above Japanese dictionary, “XML”, “no”, “outline”, and “to” of the plain text part are respectively fixed-length word numbers “α”, “β”, “γ”, Replacement / compression with “δ” or the like is performed to obtain a compressed XML document as shown in FIG.

  In the present embodiment, as shown in FIG. 33, when a space (space) exists in the character string in the region surrounded by “<” and “>” in the XML document, the character string is blanked. Separate by parts. Each character string portion obtained by dividing in this way is associated with a replacement character (shortened character string).

  That is, in the example shown in FIG. 33, first, there is a start tag “table orient =” PORT ”tocentry =“ 1 ”” surrounded by “<” and “>”. By dividing the string with the blank part that exists in this string without registering as, the string is divided into three parts: "table", "orient =" PORT "", and "tocentry =" 1 "" To do.

  Then, as shown in FIG. 33, replacement characters (shortened character strings) “e”, “f”, and “g” are added to the divided portions. By shortening the registration contents in this way, the size of the tag dictionary can be reduced, the search target can be reduced, and the hardware amount of the search means can also be reduced.

  In the example shown in FIG. 33, the parts replaced with the replacement characters “e”, “f”, “g” and the portions replaced with the replacement characters “h” and “j” are tags. The tag replaced with the replacement character “h” indicates the start of the title, and the tag replaced with the replacement character “j” indicates the end of the title. The portion replaced with the replacement character “i” is a plaintext “function list” indicating the contents of the title. The parts replaced by the replacement characters “e”, “f”, and “g” are the element name of the tag <table orient = "PORT" tocentry = "1">, the first attribute (attribute name = Attribute value) and second attribute (attribute name = attribute value).

Next, the configuration and operation of the compression apparatus and decompression apparatus according to the third embodiment related to the above-described compression processing will be described with reference to FIGS.
FIG. 25 is a block diagram showing the functional configuration of the structured document compression apparatus according to the third embodiment of the present invention, and FIG. 26 shows the functional configuration of the structured document decompression apparatus as the third embodiment of the present invention. FIG. 27 is a flowchart for explaining a tag dictionary creation procedure (tag dictionary registration procedure) in the compression device shown in FIG. 25. FIG. 28 is a flowchart for explaining the compression procedure by the compression device shown in FIG. FIG. 29 is a flowchart for explaining a decompression procedure by the decompression apparatus shown in FIG.

  First, the compression apparatus according to the third embodiment will be described with reference to FIG. 25. In FIG. 25, 101 is a document storage unit, 102 is a DTD condition checking unit, 103 is a tag plaintext identification unit, and 104 is a tag character string. Registration unit, 105 is a tag dictionary, 106 is a character string comparison unit, 107 is a language identification unit, 108 is a Japanese dictionary, 109 is a Chinese dictionary, 110 is an English dictionary, 111 is a tag character string conversion unit, and 112 is a word number The conversion unit, 113 is a word number file, and 114 is a DTD entry unit.

The document storage unit 101 inputs and holds an XML document to be compressed, and is a memory that holds an XML document before compression as shown in FIG. 30A, for example.
Whether the XML document to be compressed has a built-in DTD as shown in FIG. 30A, or refers to the DTD of another file as shown in FIG. As shown in A), the pattern recognition function of the first and second embodiments is realized by identifying one of the three types without DTD.

  More specifically, if “<! DOCTYPE” is described in the second line of the XML document to be compressed, the DTD condition examining unit 102 identifies that the document has a DTD, and the document type name (here, “book”). ) That is, if “[” is described after the document type declaration, it is identified as an XML document with a built-in DTD, and if “[” is not described, it is identified as an XML document that refers to the DTD of the external file. To do. If “<! DOCTYPE” is not described in the second line of the XML document, it is identified as an XML document without DTD.

  The tag plaintext identifying unit 103 identifies whether the target character string of the XML document is a tag or plaintext. If the target character string is surrounded by “<” and “>” from the front and back, the target character string Is identified as a description in a tag, and if not enclosed, the character string of interest is identified as a description in plain text. For example, “XML overview” on line 9 in FIG. 30A and “What is XML” on line 10 or “αβγ” on line 9 and “αβγ” on line 10 in FIG. Since “αδ...” is not surrounded by “<” and “>”, it is identified as plain text.

The tag character string registration unit 104 registers / creates a tag dictionary 105 as shown in FIG. 30C, for example. The tag dictionary 105 is a character string divided as shown in FIG. (That is, “element name”, “attribute name =“ attribute value ””, etc.) are associated with abbreviated character strings (replacement characters), respectively.
The tag plaintext identification unit 103 and the tag character string registration unit 104 described above perform functions corresponding to the document realization value analysis unit 20, the DTD analysis unit 30, and the tag dictionary creation unit 80 in the second embodiment.

  Here, the tag character string registration unit 104 creates the tag dictionary 105 as described above according to the flowchart (steps S241 to S251) shown in FIG. 27. A tag dictionary creation procedure (tag dictionary registration procedure) by the registration unit 104 will be described.

First, characters are sequentially input to the tag character string registration unit 104 (step S241), it is determined whether or not the input character is EOF (End Of File) (step S242), and if it is EOF (YES route), The tag dictionary creation operation is terminated.
If it is determined in step S242 that the input character is not EOF (NO route), it is determined whether or not the input character is “<” (step S243). If it is not “<” (NO route), step S241 is performed. Return to and enter the next character.

  If it is determined in step S243 that the input character is “<” (YES route), the character string in the memory is emptied (#) (step S244), and the characters after “<” (characters in the tag) ) Are sequentially input (step S245). The character string input from step S241 and erased from the memory in step S244 is a plain text described before “<”.

  It is determined whether or not the character input in step S245 is a blank (step S246). If the character is blank (YES route), an appropriate replacement is performed on the character string stored in the memory until the blank is recognized. A character (for example, “b”) is determined, and the replacement character is associated with the character string and registered in the tag dictionary 105 (step S247). Thereafter, the process returns to step S244, and the same processing is repeatedly executed.

  If it is determined in step S246 that it is not blank (NO route), it is determined whether or not the character input in step S245 is "!" (Step S248). Since the character string following “!” Is not a tag but a comment, the process returns to step S241.

  If it is determined in step S248 that it is not “!” (NO route), it is determined whether or not the character input in step S245 is “>” (step S249), and if “>” (YES route). ), An appropriate replacement character is determined for the character string stored in the memory until the “>” is recognized, and the replacement character and the character string are associated with each other and registered in the tag dictionary 105 (step S250). .

Recognizing “>” in step S249 means that the description of the tag is finished, so the process returns to step S241 and the same processing is repeatedly executed.
If it is determined in step S249 that it is not “>” (NO route), the character input in step S245 this time is added to the character string stored in the memory as a new character string (step S251), the process returns to step S245.

Therefore, when a character string as shown in FIG. 33 is input, the character string is separated for each blank, and the character strings “table”, “orient =“ PORT ””, “tocentry =“ 1 ”” are separated. The replacement characters e, f, and g are respectively determined, and the correspondence between these character strings and the replacement characters is sequentially registered in the tag dictionary 105.
In addition, when the character string described in the ninth line of FIG. 30A is input, each time “<” and “>” are recognized, the character string is delimited, and the delimited character string “chapter ”And“ title ”are determined as replacement characters b and c, respectively, and the correspondence between these character strings and replacement characters is sequentially registered in the tag dictionary 105.

  In the compression apparatus shown in FIG. 25, the tag dictionary 105 is registered and created as described above by the tag character string registration unit 104. For example, FIG. 30 (C), FIG. 31 (E) and FIG. And a comparison table of the replacement character and the character string in the tag, as shown in FIG.

  The character string comparison unit 106 compares the character string held in the document storage unit 101 with the registered character string or word character string in the dictionaries 105 and 108 to 110, and the character string is compared with the registered character string or word character string. If they match, a replacement character is output for the character string. At this time, the character string in the tag is compared with a registered character string in the tag dictionary 105, and the character string in the plaintext part is compared with a word character string in any of the word dictionaries 108 to 110 described later. The XML document is compressed by converting the character string into a replacement character (shortened character string) output by the character string comparison unit 106.

  The language identification unit 107 determines the number of languages described in the content (plain text) of the XML document to be compressed, the character code name described in the encoding declaration (not shown) in the XML declaration, and an arbitrary The tag is identified by deciphering the value of the language identification attribute (xml: lang). The language identifying unit 107 of this embodiment identifies whether the language is, for example, Japanese, Chinese, or English, and in accordance with the identification result, the Japanese dictionary 108, China One of the word dictionary 109 and the English dictionary 110 is selected as a word dictionary for plain text.

When the language described in the plain text of the XML document is Japanese, the Japanese dictionary 108, for example, as shown in FIG. 30D, the Japanese word character string that forms the plain text and the corresponding word number ( This is a comparison table for correspondence with abbreviated character strings and replacement characters, and is a known table constructed in advance. Here, the word number is a shortened character string (replacement character) having a fixed byte length, which is shorter than the actual word character string in the plain text and can identify the word character string.
Similarly, each of the Chinese dictionary 109 and the English dictionary 110 is a comparison table for associating a word character string of each country constituting a plain text with a word number (abbreviated character string, replacement character).

  The tag character string conversion unit 111 converts the character string of the tag into a replacement character (shortened character string) given by the character string comparison unit 106 in response to a match signal from the character string comparison unit 106. Is to do. The tag character string conversion unit 111 and the character string comparison unit 106 described above perform functions corresponding to the document realization value character string replacement unit 41 and the DTD character string replacement unit 51 of the second embodiment.

  Similarly, the word number conversion unit 112 converts the plain text character string to the word number (replacement character, shortened character string) given by the character string comparison unit 106 in response to a match signal from the character string comparison unit 106. The word is converted into a fixed byte word number by the longest match method.

The word number file 113 is an XML document such as that shown in FIG. 30B obtained by the replacement character conversion output from the tag character string conversion unit 111 and the word number conversion output from the word number conversion unit 112, for example. Holds compressed data.
As shown in FIG. 31A, the DTD entry unit 114 enters a new file name for an XML document that refers to a DTD of another file.

Next, an extension device according to a third embodiment will be described with reference to FIG. In the figure, the same reference numerals as those already described indicate the same or substantially the same parts, and the description thereof will be omitted.
In FIG. 26, 120 is a word number file storage unit, 121 is a replacement character comparison unit, 122 is a tag character string reverse conversion unit, 123 is a word number reverse conversion unit, 124 is a document storage unit, and 125 is an old DTD entry unit. .

The word number file storage unit 120 holds an XML document after compression, for example, as shown in FIG.
The replacement character comparison unit 121 compares the replacement characters (word numbers) held in the word number file storage unit 120 with the replacement characters (word numbers) in the dictionaries 105 and 108 to 110, and these replacement characters match. In this case, a character string (word character string) is output for the replacement character. At this time, the replacement character in the tag is compared with the replacement character registered in the tag dictionary 105, and the word number (replacement character) of the plaintext part is compared with the word character string in any of the word dictionaries 108 to 110.

The tag character string reverse conversion unit 122 performs reverse conversion of the replacement character in the compressed XML document into a character string corresponding to the replacement character based on a signal from the replacement character comparison unit 121.
Similarly, the word number reverse conversion unit 123 reversely converts the word number (replacement character) in the compressed XML document into a word character string corresponding to the word number based on the signal from the number comparison unit 121. is there.

The document storage unit 124 describes the tag character string corresponding to the replacement character output from the tag character string reverse conversion unit 122 and the word character string corresponding to the word number output from the word number reverse conversion unit 123. The XML document that has been restored (restored document) is held.
The old DTD entry unit 125 restores the new DTD entered during the compression to the old DTD in the XML document.

Next, the operation of the compression apparatus according to the third embodiment described with reference to FIG. 25 will be described according to the flowchart (steps S201 to S217) illustrated in FIG. 28 with reference to FIGS.
First, for example, an XML document as shown in FIG. 30A created by an operator is held in the document storage unit 101. The DTD condition checking unit 102 refers to the XML document to be compressed stored in the document storage unit 101 and checks whether “<! DOCTYPE” is described in the XML document (step S201). , “<! DOCTYPE” is described (YES route), it is determined that the XML document has DTD. When it is determined that there is no DTD (NO route), the process proceeds to step S203.

  Subsequently, the DTD condition checking unit 102 determines whether or not “[” is described in the XML document having the DTD (step S202). If “[” is described (YES route), it is determined that the DTD is built in the XML document, and the process proceeds to step S203. On the other hand, if “[” is not described (NO route). ), It is determined that the DTD for the XML document is stored in the external file, and the process proceeds to step S209.

The XML document with the built-in DTD is compressed by performing the following steps S203 to S208.
First, the tag character string registration unit 104 performs tag character string replacement character registration (tag dictionary creation) processing according to the flowchart shown in FIG. 27, and creates a tag dictionary 105 as shown in FIG. 30C (step S203). ).

  Subsequently, the tag plaintext identifying unit 103 identifies the plaintext from the XML document held in the document storage unit 101 and sends it to the language identifying unit 107. At this time, the language identification unit 107 includes the character code name (not shown) described in the encoding declaration of the XML document from the tag plaintext identification unit 103, and the language identification attribute (xml: lang) in an arbitrary tag. By deciphering the value, the number of plaintext languages is identified, and a word dictionary corresponding to the language, for example, Japanese dictionary 108 in the case of Japanese is selected (step S204).

  Then, the character string comparison unit 106 compares the registered character string in the tag dictionary 105 with the character string in the tag in the XML document to be compressed, and if these character strings match, the character string comparison unit 106 corresponds to the registered character string. The replacement character (abbreviated character string) is read from the tag dictionary 105, and the tag character string conversion unit 111 converts the character string in the tag into a replacement character (step S205).

  Next, the character string comparison unit 106 compares the registered word character string in the Japanese dictionary 108 as shown in FIG. 30D with the word character string in the plain text in the XML document to be compressed, and these word characters. If the strings match, the word number (replacement character, abbreviated character string) corresponding to the registered word character string is read from the Japanese dictionary 108 and the word character string in the plaintext is converted to the word number by the word number conversion unit 112. Conversion is performed (step S206).

Thereafter, the tag dictionary 105 is output to a file (not shown) because it is necessary for reverse conversion processing, that is, decompression / restoration described later (step S207).
Further, the tag converted into the replacement character in step S205 and the plaintext converted into the word number in step S206 are output to the word number file 113 as an XML document compressed as shown in FIG. The word number file 113 is held in a storage device (not shown) or the like (step S208).

On the other hand, an XML document using DTD stored in an external file is compressed by performing the following steps S209 to S217.
If it is determined in step S202 that “[” is not described, the DTD condition examining unit 102 determines that it is necessary to recognize the DTD for the XML document to be compressed by an external reference. For example, FIG. As shown in the second line, the DTD of another file (see, for example, FIG. 31B) is recognized and referenced based on the file name “book.dtd” designated by the system identifier. Then, the tag character string registration unit 104 performs tag character string registration processing on the DTD, and a tag dictionary 105 as shown in FIG. 31E is created (step S209).

  Subsequently, as in step S <b> 204, the tag plaintext identifying unit 103 identifies the plaintext from the XML document held in the document storage unit 101 and sends it to the language identifying unit 107. At this time, the language identifying unit 107 interprets the character code name (not shown) described in the encoding declaration of the XML document from the tag plaintext identifying unit 103 to determine how many languages are described in plaintext. And a word dictionary corresponding to the language, for example, the Japanese dictionary 108 in the case of Japanese is selected (step S210).

  Further, the DTD condition checking unit 102 controls the DTD entry unit 114 to enter the DTD name “book2.dtd” of the new file in the old and new DTD table as shown in FIG. In the process of 112, the old DTD name “book.dtd” before compression is rewritten and entered into the DTD name “book2.dtd” of the new file (step S211).

  Then, the character string comparison unit 106 compares the registered character string in the tag dictionary 105 with the character string in the tag in the DTD of another file as shown in FIG. 31B, and when these character strings match. The replacement character (abbreviated character string) corresponding to the registered character string is read from the tag dictionary 105, and the character string in the tag is converted into the replacement character by the tag character string conversion unit 111, so that FIG. A new file DTD as shown is compressed and created (step S212).

  Similarly to step S205, the character string comparison unit 106 compares the registered character string in the tag dictionary 105 with the character string in the tag in the compression target XML document stored in the document storage unit 101, and these characters. If the strings match, the replacement character (abbreviated character string) corresponding to the registered character string is read from the tag dictionary 105, and the tag character string conversion unit 111 converts the character string in the tag into a replacement character. Then, the tag character string conversion unit 111 sends the XML document obtained by converting and compressing the character string in the tag to the replacement character as described above to the word number conversion unit 112 (step S212).

  Next, the character string comparison unit 106 converts the registered word character string in the Japanese dictionary 108 as shown in FIG. 31 (F) and the XML document from the tag character string conversion unit 111 (the character string in the tag is converted into a replacement character). If the word character strings match, the word numbers (replacement characters, shortened character strings) corresponding to the registered word character strings are read from the Japanese dictionary 108. Then, the word number conversion unit 112 converts the word character string in the plain text into a word number. At this time, the DTD entry unit 114 rewrites the old DTD name “book.dtd” into the new DTD name “book2.dtd” in the XML document. In this way, the XML document before compression as shown in FIG. 31A is compressed into an XML document as shown in FIG. 31C (step S213).

After that, the tag dictionary 105 is output to a file (not shown) because it is necessary for reverse conversion processing, that is, decompression / restoration described later, as in step S207 (step S214).
Also, the compressed XML document as shown in FIG. 31C converted in step S213 is output from the word number conversion unit 112 to the word number file 113 (step S215).

Further, the new file DTD compressed and created in step S212 is output to a file not shown (step S216), and the old and new DTD correspondence table shown in FIG. 31G created in step S211 will also be described later. Since it is necessary for reverse conversion processing, that is, decompression / restoration, it is output to a file not shown (step S217).
Note that the XML document having no DTD is also subjected to the processing in steps S203 to S208 described above, and the XML document shown in FIG. 32A is compressed as shown in FIG. 32B, for example.

Next, the operation of the decompression apparatus according to the third embodiment described with reference to FIG. 26 will be described according to the flowchart (steps S221 to S232) illustrated in FIG. 29 with reference to FIGS.
First, the tag dictionary 105 is extracted from a file (not shown), and the tag dictionary 105 is stored in a memory (not shown) (step S221).

Then, the compressed XML document word number file 113 corresponding to the tag dictionary 105 is stored in the storage unit 120 (step S222).
Through the processing in steps S221 and S222, the XML document compressed as shown in FIG. 30B is stored in the storage unit 120 for the tag dictionary shown in FIG. 30C, and the tag shown in FIG. The XML document compressed as shown in FIG. 31C is stored in the storage unit 120 for the dictionary, and the tag dictionary shown in FIG. 32C is compressed as shown in FIG. 32B. An XML document is stored in the storage unit 120.

Thereafter, the DTD condition checking unit 102 reads the XML document to be decompressed and restored from the storage unit 120, and first checks whether or not “<! DOCTYPE” is described in the XML document (step S223).
When “<! DOCTYPE” is described (YES route), the DTD condition examining unit 102 further checks whether “[” is described (step S224), and “[” is described. In the case (YES route), it is determined that the XML document to be decompressed / restored contains DTD.

  As described above, when the XML document to be decompressed / restored includes DTD, the language identifying unit 107 determines the plaintext language of the XML document from the value of the language identifying attribute (xml: lang) in an arbitrary tag. A word dictionary corresponding to the identification result, for example, a Japanese dictionary 108 as shown in FIG. 31D is selected from a group of word dictionaries that have been identified (step S225).

  The tag plaintext identification unit 103 identifies the tag portion of the compressed XML document, and the replacement character comparison unit 121 compares and collates the replacement character of the tag dictionary 105 with the replacement character of the tag portion output from the storage unit 120. If these replacement characters match, the in-tag character string corresponding to the replacement character is read from the tag dictionary 105 and sent to the tag character string reverse conversion unit 122, and is compressed by the tag character string conversion unit 122. A tag character string reverse conversion process is performed to convert the replacement character of the XML document into the character string in the tag (step S226). Thereby, for example, the replacement characters b, c, and d in FIG. 30B are converted into “chapter”, “title”, and “paragraph”, respectively, as shown in FIG.

  Next, the tag plaintext identification unit 103 identifies the plaintext part of the compressed XML document, and the word number (replacement character) of the plaintext part is registered in the replacement character comparison unit 121 by the registered word number (registration). If these word numbers match, the word character string corresponding to the word number is read from the Japanese dictionary 108 and sent to the word number reverse conversion unit 123, and this word number reverse conversion unit 123 In step S227, a word number reverse conversion process for converting the word number of the compressed XML document into a word character string is performed. Thereby, for example, the word numbers α, β, γ, and δ in FIG. 30B are converted into “XML”, “No”, “Outline”, and “What” as shown in FIG. The

The tag character string and plaintext that have been inversely converted in steps S226 and S227, respectively, become an uncompressed XML document as shown in FIG. 30A, that is, a restored document, and is held in the document storage unit 124.
On the other hand, if it is determined in step S224 that “[” is not described in the XML document to be decompressed and restored (NO route), the DTD condition examining unit 102 stores the DTD for the XML document in the external file. For example, a new and old DTD correspondence table as shown in FIG. 32G created and stored during the compression processing of the XML document is input (step S228).

  Similarly to step S225, the language identification unit 107 identifies the plaintext language of the XML document from the value of the language identification attribute (xml: lang) in the tag, Then, a word dictionary corresponding to the identification result, for example, a Japanese dictionary 108 as shown in FIG. 32 (F) is selected (step S229).

  Further, the DTD condition investigation unit 102 controls the old DTD entry unit 125 to recognize that the original file name of the DTD is “book.dtd” from the old and new DTD correspondence table read out in S228. In the XML document to be decompressed and restored as shown in FIG. 31C, the DTD name described as “book2.dtd” is rewritten and entered into the old DTD name “book.dtd” (step S230).

  Then, similarly to step S226, the tag plaintext identifying unit 103 identifies the tag portion of the compressed XML document, and compares the replacement character of the tag dictionary 105 with the replacement character of the tag portion output from the storage unit 120. When these replacement characters are matched by the comparison / matching unit 121, the in-tag character string corresponding to the replacement character is read from the tag dictionary 105 and sent to the tag character string reverse conversion unit 122, and this tag character string conversion is performed. In part 122, a tag character string reverse conversion process is performed in which the replacement character of the compressed XML document is converted into an in-tag character string (step S231).

Thus, for example, the replacement characters b, c, and d in FIG. 31C are converted into “chapter”, “title”, and “paragraph”, respectively, as shown in FIG. The number is output to the number conversion unit 123.
At this time, the DTD name described in the XML document is converted from the file name “book2.dtd” to the old file name “book.dtd” by the process of step S230.

  Next, as in step S 227, the tag plaintext identifying unit 103 identifies the plaintext part of the compressed XML document, and the word number (replacement character) of the plaintext part is determined in the replacement character comparison unit 121 in the Japanese dictionary 108. If these word numbers match, the word character string corresponding to the word number is read out from the Japanese dictionary 108 and sent to the word number reverse conversion unit 123. The number reverse conversion unit 123 performs word number reverse conversion processing for converting the word number of the compressed XML document into a word character string (step S232).

As a result, for example, the word numbers α, β, γ, and δ in FIG. 31C are converted into “XML”, “no”, “outline”, and “what” as shown in FIG. The
The tag character string or plaintext reversely converted in this way becomes, for example, an XML document before compression as shown in FIG. 31A, that is, a restored document, and is held in the document storage unit 124.

  If it is determined in step S223 that “<! DOCTYPE” is not described in the decompression / restoration target XML document (NO route), it is determined that the XML document has no DTD, and the process proceeds to step S225. The XML document is also subjected to the processing of steps S223 to S227 described above, and the XML document shown in FIG. 32A is expanded and restored as shown in FIG. Retained.

  Thus, in the third embodiment of the present invention described above, the XML document is compressed by converting the character string into a replacement character (shortened character string), and the tag dictionary 105 used at that time is compressed. The word dictionaries 108 to 110 for each national language are all stored without being compressed, so that the XML document can be searched without being decompressed. It is suitable for use when the system should not exceed 1 second.

Next, configurations and operations of a compression apparatus and an expansion apparatus as modifications of the third embodiment of the present invention will be described with reference to FIGS. 34 to 37.
Here, FIG. 34 is a block diagram showing a functional configuration of a structured document compression apparatus as a modification of the third embodiment of the present invention, and FIG. 35 is a structured document as a modification of the third embodiment of the present invention. FIG. 36 is a flowchart for explaining the procedure for compressing a structured document in the modification of the third embodiment, and FIG. 37 is a block diagram of the structured document in the modification of the fourth embodiment. It is a flowchart for demonstrating an expansion | extension procedure.

  First, a compression apparatus as a modified example of the third embodiment will be described with reference to FIG. 34. In FIG. 34, 131 is a document storage unit, 132 is a tag plaintext identification unit, 133 is a tag character string registration unit, 134 is a tag dictionary, 135 is a character string comparison unit, 136 is a language identification unit, 137 is a Japanese dictionary, 138 is a Chinese dictionary, 139 is an English dictionary, 140 is a tag character string conversion unit, 141 is a word number conversion unit, Reference numeral 142 denotes a variable length encoding unit, and 143 denotes a compressed file storage unit.

The document storage unit 131 holds an XML document to be compressed. For example, the document storage unit 131 is a memory that stores an XML document before compression as shown in FIG. Corresponding.
The tag plaintext identifying unit 132 identifies whether the target character string of the XML document is a plaintext tag. The tag plaintext identifying unit 132 corresponds to the tag plaintext identifying unit 103 in FIG. 25 and operates in the same manner as the tag plaintext identifying unit 103. To do.

The tag plaintext string registration unit 133 creates a tag dictionary 134 for converting a character string in a tag into a replacement character (shortened character string), and corresponds to the tag character string registration unit 104 in FIG. The tag dictionary registration operation is performed according to the procedure described above with reference to FIG.
The tag dictionary 134 is created by the tag plaintext string registration unit 133. For example, as shown in FIG. 30 (C), FIG. 31 (E), and FIG. 32 (C), the replacement character, the character string in the tag, And is held in memory. This tag dictionary 134 corresponds to the tag dictionary 105 in FIG.

  The character string comparison unit 135 compares the character string held in the document storage unit 131 with the registered character string or word character string in the dictionaries 134 and 137 to 139, and the character string is compared with the registered character string or word character string. If they match, a replacement character is output for the character string. At this time, the character string in the tag is compared with a registered character string in the tag dictionary 134, and the character string in the plaintext part is compared with a word character string in any of word dictionaries 137 to 139 described later. The XML document is compressed by converting the character string into the replacement character (shortened character string) output from the character string comparison unit 135. The character string comparison unit 135 corresponds to the character string comparison unit 106 in FIG.

  The language identification unit 136 determines the number of languages described in the content (plain text) of the XML document to be compressed, the character code name described in the encoding declaration (not shown) in the XML declaration, or an arbitrary The tag is identified by deciphering the value of the language identification attribute (xml: lang). The language identification unit 136 shown in FIG. 34 also identifies which of the three languages, for example, Japanese, Chinese, or English, and the Japanese dictionary 137, Chinese dictionary corresponding to the identification result. One of 138 and English dictionary 139 is selected as a plain text word dictionary and corresponds to the language identification unit 107 in FIG.

  When the language described in the plain text of the XML document is Japanese, the Japanese dictionary 137, for example, as shown in FIG. 30D, the Japanese word character string constituting the plain text and the corresponding word number ( 25 is a comparison table for correspondence with abbreviated character strings and substitution characters), which is a known table constructed in advance and corresponding to the Japanese dictionary 108 in FIG.

Similarly, each of the Chinese dictionary 109 and the English dictionary 110 is a comparison table for taking correspondences between word character strings of each language constituting plaintext and word numbers (abbreviated character strings, replacement characters). These are known ones that correspond to the Chinese dictionary 109 and the English dictionary 110 in FIG.
The tag character string conversion unit 140 converts the character string of the tag into a replacement character (shortened character string) given by the character string comparison unit 106 in response to a match signal from the character string comparison unit 135. This corresponds to the tag character string conversion unit 111 in FIG.

  Similarly, the word number conversion unit 141 converts the plain text character string to the word number (replacement character, shortened character string) given by the character string comparison unit 106 in response to a match signal from the character string comparison unit 135. A word is converted into a fixed-byte word number by the longest match method, and corresponds to the word number conversion unit 112 in FIG.

The variable length encoding unit 142 converts the result of conversion by the tag character string conversion unit 140 or the word number conversion unit 141 (replaces the character string described in the tag or DTD with a replacement character, and converts the word character string in the plaintext to the word number. The character string obtained by replacement), that is, the compressed file is variable-length encoded by a known method, and the compressed file subjected to variable-length encoding is output together with the tag dictionary.
The compressed file storage unit 143 holds the tag dictionary and the variable length encoded compressed file output from the variable length encoding unit 142.

Next, an extension device as a modification of the third embodiment will be described with reference to FIG. In the figure, the same reference numerals as those already described indicate the same or substantially the same parts, and the description thereof will be omitted.
35, 151 is a compressed file storage unit, 152 is a variable length decoding unit, 153 is a replacement character comparison unit, 154 is a tag character string reverse conversion unit, 155 is a word number reverse conversion unit, and 156 is a document storage unit. .

The compressed file storage unit 151 receives a compressed file that has been variable-length encoded by the compression device shown in FIG. 34, and may be the compressed file storage unit 143 itself shown in FIG. The compressed file output from the file storage unit 143 may be held.
The variable length decoding unit 152 decodes the compressed file read from the compressed file storage unit 151 from the variable length encoded state into a compressed file as shown in FIG. 30B, for example. .

  The replacement character comparison unit 153 compares the replacement characters (word numbers) of the compressed file decoded by the variable length decoding unit 152 with the replacement characters (word numbers) of the dictionaries 134 and 137 to 139, and replaces these replacement characters. If the characters match, a character string (word character string) is output for the replacement character. At this time, the replacement character in the tag is compared with the replacement character registered in the tag dictionary 134, and the word number (replacement character) of the plaintext part is compared with the word character string in any of the word dictionaries 137 to 139. This replacement character comparison unit 153 corresponds to the replacement character comparison unit 121 in FIG.

The tag character string reverse conversion unit 154 corresponds to the replacement character in the compressed file (the XML document after compression) decoded by the variable length decoding unit 152 based on the signal from the replacement character comparison unit 153. This is a reverse conversion to the character string, and corresponds to the tag character string reverse conversion unit 122 in FIG.
Similarly, the word number reverse conversion unit 155 uses the word number in the compressed file (the XML document after compression) decoded by the variable length decoding unit 152 based on the signal from the replacement character comparison unit 153. Is reversely converted to a word character string corresponding to, and corresponds to the word number reverse conversion unit 123 in FIG.

  The document storage unit 156 describes the tag character string corresponding to the replacement character output from the tag character string reverse conversion unit 154 and the word character string corresponding to the word number output from the word number reverse conversion unit 155. The stored XML document (restored document) is stored, and corresponds to the document storage unit 124 in FIG.

Next, the operation of the compression apparatus described in FIG. 34 will be described according to the flowchart (steps S261 to S267) shown in FIG. 36 with reference to FIG.
First, for example, an XML document as shown in FIG. 30A created by an operator is held in the document storage unit 101. Then, the tag plaintext identification unit 132 extracts a tag portion from the XML document held in the document storage unit 131, and based on the tag portion, the tag character string registration unit 133 follows the flowchart shown in FIG. The replacement character registration (tag dictionary creation) is performed to create a tag dictionary 134 as shown in FIG. 30C (step S261).

  Subsequently, the tag plaintext identification unit 132 identifies the plaintext from the XML document held in the document storage unit 131 and sends it to the language identification unit 136. At this time, the language identifying unit 107 uses the character code name (not shown) described in the encoding declaration of the XML document from the tag plaintext identifying unit 132, or the language identifying attribute (xml: lang) in an arbitrary tag. The plaintext language is identified by deciphering the value of and the word dictionary corresponding to the language, for example, the Japanese dictionary 137 in the case of Japanese is selected (step S262).

  The character string comparison unit 135 compares the registered character string in the tag dictionary 134 with the character string in the tag in the XML document to be compressed, and if these character strings match, the character string comparing unit 135 corresponds to the registered character string. The replacement character (shortened character string) is read from the tag dictionary 134, and the tag character string conversion unit 140 converts the character string in the tag into a replacement character (step S263).

  Next, the character string comparison unit 135 compares the registered word character string in the Japanese dictionary 137 as shown in FIG. 30D with the word character string in the plain text in the XML document to be compressed, and these word characters. When the strings match, the word number (replacement character, abbreviated character string) corresponding to the registered word character string is read from the Japanese dictionary 137, and the word number conversion unit 141 converts the word character string in the plaintext into the word number. (Step S264).

Thereafter, the conversion result [compressed file: XML document as shown in FIG. 30B] by the tag character string conversion unit 140 and the word number conversion unit 141 is variable length by the variable length encoding unit 142 by a well-known method. It is encoded (step S265).
The tag dictionary 134 is output to a file (not shown) because it is necessary for reverse conversion processing, that is, decompression / restoration described later (step S266).
Furthermore, the variable length encoding unit 142 outputs the compressed file that has been subjected to variable length encoding in S265 to the compressed file storage unit 143 (step S267).

Next, the operation of the decompression apparatus described in FIG. 35 will be described according to the flowchart (steps S271 to S276) shown in FIG. 37 with reference to FIG.
First, the tag dictionary 134 is extracted from a file (not shown), and the tag dictionary 134 is stored in a memory (not shown) (step S271).
Then, the compressed file encoded by the variable length encoding unit 142 in FIG. 34 corresponding to the tag dictionary 134 is input and stored in the compressed file storage unit 51 (step S272).

Thereafter, the variable length encoding unit 152 decodes the compressed file input to the compressed file storage unit 151 and restores the XML document as shown in FIG. 30B, for example (step S273).
Further, the language identification unit 136 identifies the plain text language of the XML document from a language identification tag (not shown), and selects a word dictionary corresponding to the identification result from a group of word dictionaries created in advance, for example, A Japanese dictionary 137 as shown in FIG. 30D is selected (step S274).

  Then, the tag plaintext identifying unit 132 identifies the plaintext part of the compressed XML document, and the word number (replacement character) of the plaintext part is registered with the registered word number (registration) of the Japanese dictionary 137 in the replacement character comparison unit 153. If these word numbers match, the word character string corresponding to the word number is read from the Japanese dictionary 137 and sent to the word number reverse conversion unit 155. This word number reverse conversion unit 155 The word number of the compressed XML document is converted into a word character string (step S275).

  Next, the tag plaintext identification unit 132 identifies the tag portion of the compressed XML document, and the substitution character comparison unit 153 compares and collates the substitution character of the tag dictionary 134 with the substitution character of the tag portion. If the two characters match, the in-tag character string corresponding to the replacement character is read from the tag dictionary 134 and sent to the tag character string reverse conversion unit 154. The tag character string conversion unit 154 uses the replacement character of the compressed XML document. Is converted into an in-tag character string (step S276). In this way, the XML document is restored and held in the document storage unit 156 as shown in FIG.

According to the third embodiment of the present invention, the compression rate and the search speed are as follows.
(i) When searching without decompression, that is, when the search time should not exceed 0.1 seconds, the compression ratio of the tag and the compression ratio of the plaintext are as follows.
The compression ratio of the tag is as follows. If the length of the character string in the tag is n bytes,
The original length of the tag is (n + 2) bytes,
The length of the compressed tag is 3 (= 1 + 2) bytes,
The compression rate is 3 / (n + 2). When n = 3, it becomes about 0.6.
The compression ratio of plain text is about 0.3 because most words are 4 bytes or 6 bytes.
The search speed of the original search system (for example, average 0.08 seconds)
Is preserved.

(ii) The method of searching by expanding, that is, when the search time may exceed 0.1 seconds, the compression rate of tags and the compression rate of plain text are as follows.
The compression ratio of the tag is as follows. If the length of the character string in the tag is n bytes,
The original length of the tag is (n + 2) bytes,
The length of the compressed tag is 2 bytes (fixed byte code),
The compression rate is 2 / (n + 2). When n = 3, it becomes about 0.4. Even if compression is performed in the 0th-order context by variable length coding, the compression rate is the same.
The compression ratio of plain text is about 0.4 when the word number is converted by a static dictionary and compressed in the 0th-order context.
The search speed is increased by 0.02 seconds to 0.03 seconds with respect to the search speed of the search system (eg, average 0.08 seconds).

  By the way, in the technique disclosed in Japanese Patent Laid-Open No. 11-53349, the tag symbols “<” and “>” and the character string sandwiched between them are encoded together. Unlike what encodes a character string, “(chapter)”, “chapter”, etc. shown in the 3rd to 6th lines of FIG. 30A cannot be encoded. The amount of compression is small.

According to the third embodiment of the present invention, the following advantages can be obtained.
A. The tag document 105, 134 is created by the tag character string registration unit 114, 133, the XML document is compressed by compressing the tag character string based on the tag dictionary 105, 134, and compressing the tag. Not only can the rate be increased, but this tag can also be made searchable in a compressed state.

  B. As described with reference to FIG. 27 and FIG. 33, when a tag character string is registered in the tag dictionary, within the tag area surrounded by “<” and “>” (however, “<!” And “>”). (Except for the comment area surrounded by), each time a blank character appears, the character string is separated to correspond to the replacement character, so the compression rate can be increased while accurately compressing the tag part. Long tags can be separated by short characters, the size of the tag dictionary can be reduced, the tag dictionary can be configured to be easily searched, and the tags can be searched in a compressed state.

C. Not only can the long tag portion be separated by short characters to facilitate searching, but this can be further compressed to further increase the compression rate.
D. The plain text language is recognized by the character code name in the XML document or the value of the attribute for language identification with an arbitrary tag, and the word dictionary corresponding to the language is determined. For example, Japanese, Chinese, English XML documents described in various languages such as can be handled accurately.

E. The plaintext part is compared with the word character string registered in the word dictionary of each language, and the word number conversion units 112 and 141 use the fixed-length byte number (replacement character, shortened character string) for the part that matches the registered word character string. Therefore, the plaintext part can be greatly compressed.
F. The tag portion is compared with the registered character string of the tag dictionary 105, 134, and the portion that matches the registered character string is converted into a fixed-length byte replacement character (shortened character string) in the tag character string conversion units 111, 140. The tag portion can be greatly compressed.

G. Since the word dictionary corresponding to the type of the corresponding language is determined by the language identification symbol of the XML document, the XML document can be restored while supporting multiple languages.
H. At the time of decompression / restoration processing, the replacement character comparison unit 121, 153 compares the replacement character in the tag with the registered replacement character of the tag dictionary 105, 134, and the tag character string reverse conversion unit 122, 154 replaces the replacement character that matches the registered replacement character. Since the character is reversely converted into the character string in the tag, the tag can be accurately restored.

I. At the time of decompression / restoration processing, the word number (replacement character) in the plaintext is compared with the registered word number of each language in the replacement character comparison units 121 and 153, and the word number reverse conversion units 123 and 155 match the registered word number. Since the number is reversely converted into a word character string, the plaintext word number can be accurately restored to each language.
J. et al. By further variable-length encoding the character string (replacement character and word number) after the replacement process by the variable-length encoding unit 142, the compression rate of the XML document can be further increased.

[4] Others The present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the spirit of the present invention.
For example, in the embodiment described above, the case where the structured document is an XML document has been described. However, the present invention is not limited to this, and other structured documents such as an HTML document and an SGML document are also included. It is applied in the same way, and the same effect as the above-described embodiment can be obtained.

It is a block diagram which shows the function structure of the compression apparatus of the structured document as 1st Embodiment of this invention. (A)-(D) are the figures for demonstrating the compression principle of the structured document (document realization value) in 1st Embodiment. (A)-(C) are the figures for demonstrating the compression principle of the structured document (document realization value) in 1st Embodiment. (A) And (B) is a figure for demonstrating the compression principle of the structured document (DTD) in 1st Embodiment. It is a flowchart for demonstrating the compression procedure of the structured document in 1st Embodiment. It is a flowchart for demonstrating the document realization value analysis procedure in 1st Embodiment. It is a flowchart for demonstrating the document type definition analysis procedure in 1st Embodiment. It is a flowchart for demonstrating the document realization value structure change procedure in 1st Embodiment. It is a flowchart for demonstrating the document type definition structure change procedure in 1st Embodiment. (A) And (B) is a figure for demonstrating the specific compression process (1st example) of the structured document (XML document) by 1st Embodiment. (A) And (B) is a figure for demonstrating the specific compression process (2nd example) of the structured document (XML document) by 1st Embodiment. (A) and (B) are diagrams for explaining a specific compression process (third example) of a structured document (XML document) according to the first embodiment. (A)-(D) is a figure for demonstrating the specific compression process (4th example) of the structured document (XML document) by 1st Embodiment. It is a block diagram which shows the function structure of the compression apparatus of the structured document as 2nd Embodiment of this invention. (A)-(D) are the figures for demonstrating the compression principle of the structured document in 2nd Embodiment. It is a flowchart for demonstrating the compression procedure of the structured document in 2nd Embodiment. It is a flowchart for demonstrating the document realization value character string replacement procedure in 2nd Embodiment. It is a flowchart for demonstrating the document type definition character string replacement | exchange procedure in 2nd Embodiment. It is a flowchart for demonstrating the expansion | extension procedure of the structured document in 2nd Embodiment. (A)-(C) are the figures for demonstrating the concrete compression process (1st example) of the structured document (XML document) by 2nd Embodiment. (A)-(C) are the figures for demonstrating the concrete compression process (2nd example) of the structured document (XML document) by 2nd Embodiment. (A)-(C) are the figures for demonstrating the concrete compression process (3rd example) of the structured document (XML document) by 2nd Embodiment. (A)-(D) is a figure for demonstrating the concrete compression process (4th example) of the structured document (XML document) by 2nd Embodiment. (A)-(G) are figures for demonstrating the concrete compression process (5th example) of the structured document (XML document) by 2nd Embodiment. It is a block diagram which shows the function structure of the compression apparatus of the structured document as 3rd Embodiment of this invention. It is a block diagram which shows the function structure of the expansion | extension apparatus of the structured document as 3rd Embodiment of this invention. It is a flowchart for demonstrating the tag dictionary creation procedure (tag dictionary registration procedure) in 3rd Embodiment. It is a flowchart for demonstrating the compression procedure of the structured document in 3rd Embodiment. It is a flowchart for demonstrating the expansion | extension procedure of the structured document in 3rd Embodiment. (A)-(D) is a figure for demonstrating the concrete compression process (1st example) of the structured document (XML document) by 3rd Embodiment. (A)-(G) are the figures for demonstrating the concrete compression process (2nd example) of the structured document (XML document) by 3rd Embodiment. (A)-(D) is a figure for demonstrating the specific compression process (3rd example) of the structured document (XML document) by 3rd Embodiment. It is a figure for demonstrating the compression method of the structured document in 3rd Embodiment. It is a block diagram which shows the function structure of the compression apparatus of the structured document as a modification of 3rd Embodiment of this invention. It is a block diagram which shows the function structure of the expansion | extension apparatus of the structured document as a modification of 3rd Embodiment of this invention. It is a flowchart for demonstrating the compression procedure of the structured document in the modification of 3rd Embodiment. It is a flowchart for demonstrating the expansion | extension procedure of the structured document in the modification of 4th Embodiment. (A)-(C) are the figures for demonstrating how to write the general tag in a structured document (XML document). It is a figure for demonstrating how to write the general attribute in the tag of a structured document (XML document). It is a figure for demonstrating the process of a general XML processor.

Explanation of symbols

10 Document storage unit 20 Document realization value analysis unit 30 DTD analysis unit (document type definition analysis unit)
40 Document Realization Value Configuration Change Unit 41 Document Realization Value Character String Replacement Unit 50 DTD Configuration Change Unit (Document Type Definition Configuration Change Unit)
51 DTD character string replacement part (document type definition character string replacement part)
60 New DTD file creation unit 70 New and old DTD correspondence table output unit 80 Tag dictionary creation unit 90 Tag dictionary 100 External file 101, 124, 131, 156 Document storage unit 102 DTD condition checking unit 103, 132 Tag plaintext identification unit (document realization value) Analysis Department)
104, 133 Tag character string registration unit (document realization value analysis unit, document type definition analysis unit,
Tag dictionary creation part)
105, 134 Tag dictionary 106, 135 Character string comparison unit (document realization value character string replacement unit, document type definition character string replacement unit)
107,136 Language identification unit 108,137 Japanese dictionary (word dictionary)
109,138 Chinese dictionary (word dictionary)
110,139 English dictionary (word dictionary)
111, 140 Tag character string conversion unit (document realization value character string replacement unit, document type definition character string replacement unit)
112, 141 Word number conversion part (word character string replacement part)
113 Word number file 114 DTD entry unit 120 Word number file storage unit 121,153 Replacement character comparison unit 122,154 Tag character string reverse conversion unit 123,155 Word number reverse conversion unit 125 Old DTD entry unit 142 Variable length coding unit 143 , 151 Compressed file storage unit 152 Variable length decoding unit

Claims (11)

A method for compressing a structured document, comprising:
A document realization value analyzing step for analyzing the description in the tag of the document realization value forming the structured document;
A tag dictionary that associates a character string described in a tag of the document realization value with a shortened character string that is shorter than the character string and can identify the character string in accordance with the analysis result in the document realization value analysis step. A tag dictionary creation step to be created;
Using the tag dictionary created in the tag dictionary creation step, a document realization value character string replacement step for replacing a character string described in the tag of the document realization value with a shortened character string corresponding to the character string; A method for compressing a structured document, comprising: A method for compressing a structured document, comprising:
A document realization value analyzing step for analyzing the description in the tag of the document realization value forming the structured document;
A document type definition analyzing step for analyzing a description of the document type definition forming the structured document;
According to the analysis result in the document realization value analysis step and the document type definition analysis step, the character string described in the tag of the document realization value and in the document type definition and shorter than the character string and the character string are specified. A tag dictionary creation step for creating a tag dictionary that associates possible shortened character strings;
Using the tag dictionary created in the tag dictionary creation step, a document realization value character string replacement step for replacing a character string described in the tag of the document realization value with a shortened character string corresponding to the character string; ,
A document type definition character string replacement step of replacing the character string described in the document type definition with a shortened character string corresponding to the character string using the tag dictionary created in the tag dictionary creation step. A method for compressing a structured document, characterized by:   The element name and attribute name described in the tag or in the document type definition are treated as the character string, and the element name and the attribute name are replaced with the abbreviated character string. 3. A method of compressing a structured document according to 2.   Using a word dictionary that associates a word character string with a shortened character string that is shorter than the word character string and that can specify the word character string, the word character string included in the content of the document realization value is converted into the word character 4. The method of compressing a structured document according to claim 1, further comprising a word character string replacement step for replacing with a shortened character string corresponding to the column.   A variable length in which a character string described in the tag or in the document type definition is replaced with the abbreviated character string, and after the word character string is replaced with the abbreviated character string, these character strings are compressed by variable length encoding. 5. The structured document compression method according to claim 4, further comprising an encoding step. An apparatus for compressing a structured document,
A document realization value analysis unit for analyzing the description in the tag of the document realization value forming the structured document;
In accordance with the analysis result by the document realization value analysis unit, a tag dictionary is created that associates the character string described in the tag of the document realization value with a shortened character string that is shorter than the character string and can identify the character string. A tag dictionary creation unit to
A document realization value character string replacement unit that replaces a character string described in the tag of the document realization value with a shortened character string corresponding to the character string using the tag dictionary created by the tag dictionary creation unit; A structured document compression apparatus, characterized by comprising: An apparatus for compressing a structured document,
A document realization value analysis unit for analyzing the description in the tag of the document realization value forming the structured document;
A document type definition analyzing unit for analyzing a description of the document type definition forming the structured document;
According to the analysis result by the document realization value analysis unit and the document type definition analysis unit, the character string described in the document realization value tag and in the document type definition is shorter than the character string and the character string is specified. A tag dictionary creation unit that creates a tag dictionary that associates a shortened character string
A document realization value character string replacement unit that replaces a character string described in the tag of the document realization value with a shortened character string corresponding to the character string using the tag dictionary created by the tag dictionary creation unit; ,
A document type definition character string replacement unit that replaces the character string described in the document type definition with a shortened character string corresponding to the character string using the tag dictionary created by the tag dictionary creation unit. A structured document compression apparatus, characterized by being configured.   The element name and the attribute name described in the tag or in the document type definition are treated as the character string, and the element name and the attribute name are replaced with the abbreviated character string. The structured document compression apparatus according to claim 7. A computer-readable recording medium storing a structured document compression program for realizing a function of compressing a structured document by a computer,
The structured document compression program is
A document realization value analysis unit for analyzing a description in a tag of a document realization value forming the structured document;
In accordance with the analysis result by the document realization value analysis unit, a tag dictionary is created that associates the character string described in the tag of the document realization value with a shortened character string that is shorter than the character string and can identify the character string. A tag dictionary creation unit, and
Using the tag dictionary created by the tag dictionary creation unit, as a document realization value character string replacement unit that replaces a character string described in a tag of the document realization value with a shortened character string corresponding to the character string A computer-readable recording medium storing a structured document compression program, which causes the computer to function. A computer-readable recording medium storing a structured document compression program for realizing a function of compressing a structured document by a computer,
The structured document compression program is
A document realization value analysis unit for analyzing a description in a tag of a document realization value forming the structured document;
A document type definition analyzing unit for analyzing the description of the document type definition forming the structured document;
According to the analysis result by the document realization value analysis unit and the document type definition analysis unit, the character string described in the document realization value tag and in the document type definition is shorter than the character string and the character string is specified. A tag dictionary creation unit that creates a tag dictionary that correlates with a shortened character string
Using the tag dictionary created by the tag dictionary creation unit, a document realization value character string replacement unit that replaces a character string described in a tag of the document realization value with a shortened character string corresponding to the character string, and,
Using the tag dictionary created by the tag dictionary creation unit, the computer as a document type definition character string replacement unit that replaces a character string described in the document type definition with a shortened character string corresponding to the character string. A computer-readable recording medium storing a structured document compression program, wherein   The structured document compression program treats the element name and attribute name described in the tag or in the document type definition as the character string, and replaces the element name and attribute name with the short character string in the computer. A computer-readable recording medium storing the structured document compression program according to claim 9 or 10, wherein the structured document compression program according to claim 9 or 10 is stored.

Images