The present invention relates to a document processing technique, and particularly to a document processing apparatus and a document processing method for processing a structured document having a hierarchical structure.
XML (eXtensible Markup Language) has been attracting attention as a data description format that allows the user to share data with other users via a network. This has promoted the development of applications for creating, displaying, and editing XML documents (see Patent document 1, for example). XML documents are created based upon a vocabulary (tag set) defined according to a document type definition.
XML permits the user to use multiple vocabularies in a single document. However, in some cases, such multiple vocabularies have the same element names or the same attribute names. In such a case, different elements having the same element name or different attributes having the same attribute name clash with one another. That is to say, this leads to a problem in that the vocabulary to which the element type or the attribute type belongs cannot be identified. In order to solve such a problem, XML uses the “namespace” concept, in which the namespace describes the vocabulary to which the element types and the attribute types used in a document belong.
Japanese Patent Application Laid-open No. 2001-290804
However, in some cases, a situation can occur in which the namespace has not been described properly. Accordingly, there is a demand for a technique for supporting the proper processing of a document even if such a situation has occurred.
The present invention has been made in view of the aforementioned problems. Accordingly, it is an object of the present invention to provide a technique for performing proper processing for a structured document even if information such as the namespace cannot be identified, thereby allowing such a document to be displayed and edited smoothly.
An aspect of the present invention relates to a document processing apparatus. The document processing apparatus comprises: a namespace detection unit which detects a namespace to which elements included in a document described in a markup language belong; a namespace identifying unit that provides a function whereby, in a case that the namespace detection unit has not detected any correct namespace, keywords are extracted from the document based upon a predetermined condition, and the namespace is identified based upon the keywords thus extracted; and a namespace information storage unit which stores information with respect to the relation between the keywords and the namespace. With such an arrangement, the namespace identifying unit identifies the namespace based upon the extracted keywords with reference to the namespace information storage unit. Furthermore, the document is displayed, in a manner that allows a user to edit the document, based upon the namespace thus identified by the namespace detection unit or the namespace identifying unit.
The “markup language” may be a kind of XML, e.g., XHTML (eXtensible HyperText Markup Language), SVG (Scalable Vector Graphics), MathML (Mathematical Markup Language), etc. Also, examples of the markup languages include SGML (Standard Generalized Markup Language), HTML (HyperText Markup Language), etc. Examples of the keywords include: a filename extension included in the file name of a document; an element name (tag name) or an attribute name described in the document; etc., which allow the namespace to be conjectured.
Also, the document processing apparatus may further comprise a namespace display unit which displays multiple namespaces detected by the data name identifying unit, and which allows the user to select one from among the multiple namespace candidates, thereby identifying the namespace. Also, the namespace information storage unit may store the information with respect to the relation between the namespace to which the elements included in the document belong and the keywords included in the document, for the past documents that have been processed every time the document is processed. With such an arrangement, the namespace may be identified based upon the information thus stored.
Another aspect of the present invention relates to a document processing method. The document processing method comprises: a step for detecting a namespace to which elements included in a document described in a markup language belong; a step whereby, in a case that no correct namespace has been detected in the detection step, keywords are extracted from the document based upon a predetermined condition, and the namespace is identified with reference to information with respect to the relation between the keywords and the namespace stored beforehand; and a step for displaying the document, in a manner that allows a user to edit the document, based upon the namespace thus detected or identified.
Note that any combination of the aforementioned components or any manifestation of the present invention realized by replacement of a system, a recording medium, and so forth, is effective as an embodiment of the present invention.
The present invention provides a technique for supporting appropriate processing for an structured document.
a) is a diagram which shows an example of a definition file used for mapping the XML document shown in
b) is a diagram which shows an example of a definition file used for mapping the XML document shown in
Description will be made below regarding the background technique for the present invention before detailed description of the present embodiment.
The main control unit 22 provides for the loading of a plug-in or a framework for executing a command. The editing unit 24 provides a framework for editing XML documents. Display and editing functions for a document in the document processing apparatus 20 are realized by plug-ins, and the necessary plug-ins are loaded by the main control unit 22 or the editing unit 24 according to the type of document under consideration. The main control unit 22 or the editing unit 24 determines which vocabulary or vocabularies describes the content of an XML document to be processed, by referring to a name space of the document to be processed, and loads a plug-in for display or editing corresponding to the thus determined vocabulary so as to execute the display or the editing. For instance, an HTML unit 50, which displays and edits HTML documents, and an SVG unit 60, which displays and edits SVG documents, are implemented in the document processing apparatus 20. That is, a display system and an editing system are implemented as plug-ins for each vocabulary (tag set), so that when an HTML document and an SVG document are edited, the HTML unit 50 and the SVG unit 60 are loaded, respectively. As will be described later, when compound documents, which contain both the HTML and SVG components, are to be processed, both the HTML unit 50 and the SVG unit 60 are loaded.
By implementing the above structure, a user can select so as to install only necessary functions, and can add or delete a function or functions at a later stage, as appropriate. Thus, the storage area of a recording medium, such as a hard disk, can be effectively utilized, and the wasteful use of memory can be prevented at the time of executing programs. Furthermore, since the capability of this structure is highly expandable, a developer can deal with new vocabularies in the form of plug-ins, and thus the development process can be readily facilitated. As a result, the user can also add a function or functions easily at low cost by adding a plug-in or plug-ins.
The editing unit 24 receives an event, which is an editing instruction, from the user via the user interface. Upon reception of such an event, the editing unit 24 notifies a suitable plug-in or the like of this event, and controls the processing such as redoing this event, canceling (undoing) this event, etc.
The DOM unit 30 includes a DOM provider 32, a DOM builder 34 and a DOM writer 36. The DOM unit 30 realizes functions in compliance with a document object model (DOM), which is defined to provide an access method used for handling data in the form of an XML document. The DOM provider 32 is an implementation of a DOM that satisfies an interface defined by the editing unit 24. The DOM builder 34 generates DOM trees from XML documents. As will be described later, when an XML document to be processed is mapped to another vocabulary by the VC unit 80, a source tree, which corresponds to the XML document in a mapping source, and a destination tree, which corresponds to the XML document in a mapping destination, are generated. At the end of editing, for example, the DOM writer 36 outputs a DOM tree as an XML document.
The CSS unit 40, which provides a display function conforming to CSS, includes a CSS parser 42, a CSS provider 44 and a rendering unit 46. The CSS parser 42 has a parsing function for analyzing the CSS syntax. The CSS provider 44 is an implementation of a CSS object and performs CSS cascade processing on the DOM tree. The rendering unit 46 is a CSS rendering engine and is used to display documents, described in a vocabulary such as HTML, which are laid out using CSS.
The HTML unit 50 displays or edits documents described in HTML. The SVG unit 60 displays or edits documents described in SVG. These display/editing systems are realized in the form of plug-ins, and each system is comprised of a display unit (also designated herein as a “canvas”) 56 and 66, which displays documents, a control unit (also designated herein as an “editlet”) 52 and 62, which transmits and receives events containing editing commands, and an edit unit (also designated herein as a “zone”) 54 and 64, which edits the DOM according to the editing commands. Upon the control unit 52 or 62 receiving a DOM tree editing command from an external source, the edit unit 54 or 64 modifies the DOM tree and the display unit 56 or 66 updates the display. These units have a structure similar to the framework of the so-called MVC (Model-View-Controller). With such a structure, in general, the display units 56 and 66 correspond to “View”. On the other hand, the control units 52 and 62 correspond to “Controller”, and the edit units 54 and 64 and DOM instance corresponds to “Model”. The document processing apparatus 20 according to the background technique allows an XML document to be edited according to each given vocabulary, as well as providing a function of editing the HTML document in the form of tree display. The HTML unit 50 provides a user interface for editing an HTML document in a manner similar to a word processor, for example. On the other hand, the SVG unit 60 provides a user interface for editing an SVG document in a manner similar to an image drawing tool.
The VC unit 80 includes a mapping unit 82, a definition file acquiring unit 84 and a definition file generator 86. The VC unit 80 performs mapping of a document, which has been described in a particular vocabulary, to another given vocabulary, thereby providing a framework that allows a document to be displayed and edited by a display/editing plug-in corresponding to the vocabulary to which the document is mapped. In the background technique, this function is called a vocabulary connection (VC). In the VC unit 80, the definition file acquiring unit 84 acquires a script file in which the mapping definition is described. Here, the definition file specifies the correspondence (connection) between the nodes for each node. Furthermore, the definition file may specify whether or not editing of the element values or attribute values is permitted. Furthermore, the definition file may include operation expressions using the element values or attribute values for the node. Detailed description will be made later regarding these functions. The mapping unit 82 instructs the DOM builder 34 to generate a destination tree with reference to the script file acquired by the definition file acquiring unit 84. This manages the correspondence between the source tree and the destination tree. The definition file generator 86 offers a graphical user interface which allows the user to generate a definition file.
The VC unit 80 monitors the connection between the source tree and the destination tree. Upon reception of an editing instruction from the user via a user interface provided by a plug-in that handles a display function, the VC unit 80 first modifies a relevant node of the source tree. As a result, the DOM unit 30 issues a mutation event indicating that the source tree has been modified. Upon reception of the mutation event thus issued, the VC unit 80 modifies a node of the destination tree corresponding to the modified node, thereby updating the destination tree in a manner that synchronizes with the modification of the source tree. Upon reception of a mutation event that indicates that the destination tree has been modified, a plug-in having functions of displaying/editing the destination tree, e.g., the HTML unit 50, updates a display with reference to the destination tree thus modified. Such a structure allows a document described in any vocabulary, even a minor vocabulary used in a minor user segment, to be converted into a document described in another major vocabulary. This enables such a document described in a minor vocabulary to be displayed, and provides an editing environment for such a document.
An operation in which the document processing apparatus 20 displays and/or edits documents will be described herein below. When the document processing apparatus 20 loads a document to be processed, the DOM builder 34 generates a DOM tree from the XML document. The main control unit 22 or the editing unit 24 determines which vocabulary describes the XML document by referring to a name space of the XML document to be processed. If the plug-in corresponding to the vocabulary is installed in the document processing apparatus 20, the plug-in is loaded so as to display/edit the document. If, on the other hand, the plug-in is not installed in the document processing apparatus 20, a check shall be made to see whether a mapping definition file exists or not. And if the definition file exits, the definition file acquiring unit 84 acquires the definition file and generates a destination tree according to the definition, so that the document is displayed/edited by the plug-in corresponding to the vocabulary which is to be used for mapping. If the document is a compound document containing a plurality of vocabularies, relevant portions of the document are displayed/edited by plug-ins corresponding to the respective vocabularies, as will be described later. If the definition file does not exist, a source or tree structure of a document is displayed and the editing is carried out on the display screen.
Here, the document processing apparatus 20 according to the background technique does not have a plug-in which conforms to or handles the display/editing of marks managing vocabularies. Accordingly, before displaying such a document in a manner other than the source display manner or the tree display manner, the above-described VC function is used. That is, there is a need to prepare a definition file for mapping the document, which has been described in the marks managing vocabulary, to another vocabulary, which is supported by a corresponding plug-in, e.g., HTML or SVG. Note that description will be made later regarding a user interface that allows the user to create the user's own definition file. Now, description will be made below regarding a case in which a definition file has already been prepared.
a) and
On the screen as shown in
Viewers or editors which can handle major vocabularies such as XHTML, MathML and SVG have already been developed. However, it does not serve any practical purpose to develop dedicated viewers or editors for such documents described in the original vocabularies as shown in
For example, when the source display and tree-view display are implemented by dedicated plug-ins, the source-display plug-in and the tree-display plug-in execute their respective displays by directly referring to the source tree without involving the destination tree. In this case, when the editing is done in any area of the screen, the source-display plug-in and the tree-display plug-in update the screen by referring to the modified source tree. Also, the HTML unit 50 in charge of displaying the area 96 updates the screen by referring to the destination tree, which has been modified following the modification of the source tree.
The source display and the tree-view display can also be realized by utilizing the VC function. That is to say, an arrangement may be made in which the source and the tree structure are laid out in HTML, an XML document is mapped to the HTML structure thus laid out, and the HTML unit 50 displays the XML document thus mapped. In such an arrangement, three destination trees in the source format, the tree format and the table format are generated. If the editing is carried out in any of the three areas on the screen, the VC unit 80 modifies the source tree and, thereafter, modifies the three destination trees in the source format, the tree format and the table format. Then, the HTML unit 50 updates the three areas of the screen by referring to the three destination trees.
In this manner, a document is displayed on a single screen in a plurality of display formats, thus improving a user's convenience. For example, the user can display and edit a document in a visually easy-to-understand format using the table 90 or the like while understanding the hierarchical structure of the document by the source display or the tree display. In the above example, a single screen is partitioned into a plurality of display formats, and they are displayed simultaneously. Also, a single display format may be displayed on a single screen so that the display format can be switched according to the user's instructions. In this case, the main control unit 22 receives from the user a request for switching the display format and then instructs the respective plug-ins to switch the display.
The displayed menu may be switched corresponding to the position of the cursor (carriage) during the editing of a document. That is, when the cursor lies in an area where an SVG document is displayed, the menu provided by the SVG unit 60, or a command set which is defined in the definition file for mapping the SVG document, is displayed. On the other hand, when the cursor lies in an area where the XHTML document is displayed, the menu provided by the HTML unit 50, or a command set which is defined in the definition file for mapping the HTML document, is displayed. Thus, an appropriate user interface can be presented according to the editing position.
In a case that there is neither a plug-in nor a mapping definition file suitable for any one of the vocabularies according to which the compound document has been described, a portion described in this vocabulary may be displayed in source or in tree format. In the conventional practice, when a compound document is to be opened where another document is embedded in a particular document, their contents cannot be displayed without the installation of an application to display the embedded document. According to the background technique, however, the XML documents, which are composed of text data, may be displayed in source or in tree format so that the contents of the documents can be ascertained. This is a characteristic of the text-based XML documents or the like.
Another advantageous aspect of the data being described in a text-based language, for example, is that, in a single compound document, a part of the compound document described in a given vocabulary can be used as reference data for another part of the same compound document described in a different vocabulary. Furthermore, when a search is made within the document, a string of characters embedded in a drawing, such as SVG, may also be search candidates.
In a document described in a particular vocabulary, tags belonging to other vocabularies may be used. Though such an XML document is generally not valid, it can be processed as a valid XML document as long as it is well-formed. In such a case, the tags thus inserted that belong to other vocabularies may be mapped using a definition file. For instance, tags such as “Important” and “Most Important” may be used so as to display a portion surrounding these tags in an emphasized manner, or may be sorted out in the order of importance.
When the user edits a document on an edit screen as shown in
Depending on the contents of the editing, modification of the display by the HTML unit 50 may change the overall layout. In such a case, the layout is updated by a screen layout management mechanism, e.g., the plug-in that handles the display of the highest node, in increments of display regions which are displayed according to the respective plug-ins. For example, in a case of expanding a display region managed by the HTML unit 50, first, the HTML unit 50 renders a part managed by the HTML unit 50 itself, and determines the size of the display region. Then, the size of the display area is notified to the component that manages the screen layout so as to request the updating of the layout. Upon receipt of this notice, the component that manages the screen layout rebuilds the layout of the display area for each plug-in. Accordingly, the display of the edited portion is appropriately updated and the overall screen layout is updated.
Let us say that an information processing apparatus according to an embodiment of the present invention is configured on the basis of the aforementioned background technique, and the information processing apparatus according to the present embodiment has a basic configuration including the document processing apparatus according to the aforementioned background technique in the form of a part of the present embodiment. Now, description will be made in the present embodiment mainly regarding an arrangement that provides a function of processing a document file structured with XML, which is an example of a structured document.
The namespace display unit 314 may display only one namespace identified by the namespace identifying unit 312, thereby allowing the user to confirm the namespace in the final stage. With such an arrangement, with reference to the namespace thus identified, the main control unit 22 or the editing unit 24 identifies the vocabulary of the XML document, and implements display and editing using a plug-in for displaying or editing that corresponds to the vocabulary. Here, let us consider a case in which the plug-in for processing the vocabulary of the namespace thus identified has not been installed on the aforementioned document processing apparatus. In this case, a message may be displayed for the user so as to prompt the user to load the plug-in. Also, the plug-in may be automatically downloaded.
Such an arrangement has a function of automatically identifying the namespace based upon the filename extension of the file or the element names, the attribute names, or the like, described in the document before the XML document is processed, even if there is no description of the namespace, or even if the description of the namespace is missing or erroneous. This enables the user to perform the document processing smoothly without involving time-consuming operations, e.g., a manual search for the portion where the namespace is missing, a manual search for the namespace, etc. Thus, such an arrangement reduces the time cost involved in the document processing. Furthermore, the present embodiment can be employed as a technique for avoiding the system crashes which often occur in reading out such an XML document. Such an arrangement shows the user the potential sources of such problems, and suggests ways to solve these problems, thereby offering a user-friendly document processing apparatus which is easy to understand.
Let us consider a case in which there are multiple namespace candidates. In this case, the namespace identifying unit 312 may instruct the namespace display unit 314 to display the multiple namespace candidates, instead of automatically performing a final identification of the namespace. For example, description has been made with reference to step S16 shown in
With such an arrangement, in a case that the namespace is identified based upon multiple keywords such as the filename extension, tag names, etc., the probabilities of the namespaces thus detected are calculated. This improves the precision of the namespace identification. On the other hand, let us consider a case in which the namespace has not been automatically identified. Even in this case, with such an arrangement, the namespace candidates are narrowed down, and the namespace candidates thus narrowed down are displayed for the user. Such an arrangement reduces the number of time-consuming operations required as compared with an arrangement that requires the user to perform manual operations throughout the procedure.
Any combination of searching based upon the filename extension and searching based upon the tag names may be made. Examples of such combinations include: processing in which a first-stage search is performed based upon the tag names; processing in which a search is performed based upon only the filename extension or the tag names; etc. Also, the search key is not restricted to the filename extension, the element name, the attribute name, or the like.
In the processing step, in a case of reception of the correspondence error detection signal from the namespace identifying unit 312, the namespace display unit 314 may display the XML document as-is, i.e., in the form of a source file, and a message advising the user that the namespace has not been detected may be displayed for the user. Also, a description may be inserted so as to allow the user to identify the namespace, for example, thereby allowing the user to directly correct the XML document.
Now, specific description will be made regarding a method used by the namespace identifying unit 312 for identifying the namespace, or for narrowing down the namespace candidates.
Let us consider a case in which multiple namespaces have been detected as a result of the search using the filename extension, or a case in which no namespace has been detected. In such cases, a rule-based technique may be applied with reference to a table obtained from the table structure 400 shown in
Let us consider a case in which multiple namespaces have been detected based upon the tag name extracted from the root node. In this case, identification of the namespace may be performed in the same way based upon an element or elements extracted from lower-level classes. With such an arrangement, the namespace information storage unit 316 may store a table including first-layer tag name fields, second-layer tag name fields, and the namespace information fields. In the example of an XHTML document shown in
The namespace information storage unit 316 may further include a probability classifier (not shown) which performs learning processing based upon training sets. Upon instructing the document processing apparatus to learn an XML document or the like including a description of the namespace, the probability classifier stores the information with respect to the relation between the tag names described in the document or the filename extension of the file and the namespace which has been used as a reference for processing the document. The probability classifier having a function of learning processing based upon the training sets may employ an existing method such as Bayes's theorem, the SVM (Support Vector Machine), or the like.
Description will be made below regarding a simple example of calculation for narrowing down the namespace candidates using Bayes's theorem. In the first stage, let us consider a case in which the classifier is instructed to learn a file as a training set for determining that a file is an SVG file, which includes one “svg” element, three “desc” elements, and three “rect” elements. Let us say that this training set will be referred to as “classification A”. Then, let us say that the classifier is instructed to learn another file as a training set for determining that a file is a MathML file, which includes one “math” element, four “mi” elements, and two “mfrac” elements. Now, let us say that this training set will be referred to as “classification B”. Furthermore, let us say that a document C is processed, the namespace of which cannot be identified, and which includes three “desc” elements, one “rect” element, and one “mi” element. In this case, the probability that the namespace of the document C is the classification A, and the probability that the namespace of the document C is the classification B, are calculated using Bayes's theorem as follows.
Accordingly, it can be assumed that it is highly probable that the namespace of the document C is that of an SVG document, which is the classification A. For example, with such an arrangement, let us consider a case in which an XHTML document including a description of the namespace URI as shown in
Such a stochastic method as described above provides namespace candidates with probabilities calculated based upon the number of occurrences of the tag names described in the document. Accordingly, an arrangement may be made in which only the top three most probable namespace candidates are displayed on the namespace display unit 314 as the namespace candidates in the final stage. Also, an arrangement may be made in which only the namespace candidates with probabilities higher than 50% are displayed on the namespace display unit 314 as the namespace candidates in the final stage. Also, an arrangement may be made which allows the user to select one from among the namespace candidates, thereby identifying the namespace in the final stage.
The identifying methods as described above may be applied, without any modification, to documents described in other languages such as XSL (eXtensible Stylesheet Language) etc. in addition to XML documents. Description will be made with reference to an example. First,
On the other hand,
Description has been made regarding the present invention with reference to the embodiments. The above-described embodiments have been described for exemplary purposes only, and are by no means intended to be interpreted restrictively. Rather, it can be readily conceived by those skilled in this art that various modifications may be made by making various combinations of the aforementioned components or processes, which are also encompassed in the technical scope of the present invention. Some of such possible modifications will be listed below.
Description has been made in the embodiment regarding an arrangement for processing an XML document. Also, the document processing apparatus 300 has a function of processing other markup languages, e.g., SGML, HTML, etc.
Description has been made in the aforementioned embodiment regarding an arrangement in which, after the namespace has been identified, a plug-in that corresponds to the vocabulary thereof is loaded, thereby allowing a document to be processed normally, i.e., to be displayed and edited. Also, an arrangement may be made in which plug-in candidates are displayed for the user according to a similar procedure instead of displaying the namespace candidates, and the document is displayed and edited using a plug-in selected by the user.
The present invention can be applied to an apparatus for processing a structured document.
Number | Date | Country | Kind |
---|---|---|---|
2004-312835 | Oct 2004 | JP | national |
Filing Document | Filing Date | Country | Kind | 371c Date |
---|---|---|---|---|
PCT/JP05/19824 | 10/27/2005 | WO | 00 | 12/7/2007 |