 |
Description  |
|
|
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a structure retrieval apparatus in which
tags (part of the tags are omissible) are inserted in data to discriminate
portions of the data to thereby express a structure, and the structure of
the data is searched at high speed. For example, the invention is
applicable to an apparatus for searching the structure of a structured
document in which tags are inserted in a text to divide it into document
elements.
2. Description of the Related Art
Conventionally, in document editing apparatuses for document processing,
such as document editing apparatuses (word processors) in workstations, in
order to efficiently prepare a document, attempts have been made to
structure and edit the document by preparing in advance a plurality of
document parts such as headers and paragraphs and by determining
relationships among the respective document parts.
As examples of structured documents which incorporate the concept of a
structure with respect to a document, structured documents conforming to
international standards of ODA (ISO 8613: Open Document Architecture) and
SGML (ISO 8879: Standard Generalized Markup Language) are known. As for an
example of a document processing method using a structured document
conforming to ODA standards, reference is made to Japanese Unexamined
Patent Publication No. Hei. 5-135054 entitled "Document Processing Method.
"
Structured documents conforming to SGML, which have high affinity with
conventional text processing systems, have found widespread use
principally in the United States, and have already entered a stage of
practical use. This is because a conventional text processing system is
sufficiently capable of realizing the structured document since the
technique of the structured document conforming to SGML is a technique
whereby the document text is partially classified (e.g., divided as
document parts) by inserting marks called tags into the document text, the
document is structured by defining relationships among the divisions, and
a tree-structured document structure is thereby represented.
Next, by citing a structured document conforming to SGML as an example, a
description will be given of an example of processing a structured
document provided with marks. In a structured document conforming to SGML,
a pattern of a document structure is provided in advance, and the
structure of the document is constrained within the range of the provided
pattern. Such a pattern of the document structure is called a document
type definition (DTD) in SGML.
In a structured document conforming to SGML, a document type definition is
first set forth to regulate the structure of the document. Next, to
represent the structure, marks called tags are inserted in the document
text, and the document text is partially classified by the tags. For
example, one paragraph in a document is represented as shown below by
using a tag <para> having a name "para."
"<para>This is one paragraph.</para>"
The tag <para> here means a start of the paragraph, and is called a start
tag. The tag </para> means an end of the paragraph, and is called an end
tag. That is, in this example, the paragraph is marked by using two tags,
the start tag <para> and the end tag </para> having a name "para," and a
part of the document text is thereby partially classified. In other words,
the portion of the text sandwiched by the two tags indicates the content
portion of the structure indicated by the tags.
The tags which are given their names are respectively distinguished, and
their structural functions are defined in the document type definition. In
this context, the tag represents a structure of the document. Accordingly,
a structure of the structured document (an SGML-conformable document)
referred to hereafter shall mean that it is synonymous with a tag insofar
as confusion does not occur.
In addition, some tags are omissible in the structured document conforming
to SGML (hereafter abbreviated as a SGML document). In that case, whether
the omission is possible or not is designated by the document type
definition (DTD). The designation as to whether the omission is possible
or not is given independently for each start tag and each end tag. For
example, the end tag </para> is omissible in a case where a designation to
that effect is given in the document type definition, in which case the
above-described example can be written as
"<para>This is one paragraph."
The document type definition of the SGML document is written as shown in
FIG. 13, for example, In the document structure constrained by a document
type definition 130 shown in FIG. 13, it is defined that the respective
tags, including the start tag whose name is "header," the end tag of
"header," the end tag of "paragraph," the end tag of "figure," and the end
tag of "fig.sub.-- body," are omissible.
Next, a specific description will be given of the contents of the document
type definition 130 shown in FIG. 13. The document type definition (DTD)
here is written in conformity with the representation method of SGML. The
initial symbol "<!" on each line of the content of the document type
definition is a markup declaration delimiter, and an ensuing "ELEMENT"
which continues without a space is an element declaration keyword. In
other words, the initial word "<!ELEMENT" in the line constitutes a
reserved word for designating the content of its structure (a lower
structure) by an ensuing description. Then, the names (doc, chap, header,
para, fig, and fig.sub.-- body) of items described next represent the
names of object tags.
Ensuing symbols ("- - ,""- O,""O O," and so on) are symbols which represent
whether the object tags of the items are respectively omissible in the
order of the start tag and the end tag. The symbol "-" means that the tag
is not omissible, while the symbol "O" means that the tag is omissible.
For instance, if the symbols in a given line are "- O," it means that the
start tag is not omissible, and the end tag is omissible.
Items that further follow provide for definitions representing the lower
structure following the tag. The symbol "," means that items (tags) appear
in order, the symbol ".vertline." means that it suffices if either item is
present, and the symbol "*" means that the item (tag) is repeated a 0 time
or more. In addition, the symbol "?" means that the item (tag) may or may
not be present.
According, in a case where the lower structure following the tag is defined
as "chap.sub.-- header, para*, chap*)," the lower structure according to
the definition means that "first, there is a chapter header, followed by a
repetition of a paragraph a 0 time or more, further followed by a
repetition of a chapter a 0 time or more." To cite a specific example, in
a case where the lower structure following the tag is defined as "header,
(para.vertline.fig)*, chap*)" as in the second line of the document type
definition 130 shown in FIG. 13, the lower structure according to the
definition means that "there is a header, followed by a repetition of a
paragraph or a figure a 0 time or more, further followed by a repetition
of a chapter a 0 time or more."
In addition, "#PCDATA" in the lower structure following the tag and written
in the third and fourth lines is one of the reserved words of SGML, and
means in the definition of the structure that its content is character
data. Accordingly, in the example of the document type definition 130,
"#PCDATA" means that character data follows the tags on the "header" and
the "para" which constitute the "chap."
That is, in the pattern of the document structure according to the document
type definition (DTD) shown in FIG. 13, the relevant document is one which
starts with a tag "<doc>" constituted by a repetition of a "chap"
(chapter), and that "chap" has a "para" (paragraph) or a "fig" (figure)
which is repeated a 0 time or more following the "header," further
followed by a repetition of a "chap" a 0 time or more. In addition, the
"header" and the "para" in this example are constituted by character data.
As detailed rules, it is defined that the content of the "fig" in the
document structure is comprised of the "header" and the "fig.sub.-- body"
(figure body) that follows, and the "fig.sub.-- body" is defined as having
not lower structure ("EMPTY") since, for instance, an external image file
is referred to. In addition, as for the designation concerning whether the
tags of the structure are omissible or not, it is defined that the tags of
the "doc" and the "chap" are not omissible, that, as for the tags of the
"para," the "fig," and the "fig.sub.-- body," only their end tags are
omissible, and that, as for the "header," its both tags are omissible.
An example of an actual document conforming to such a document type
definition (hereafter such a document will be referred to as an object
document) is, for example, an SGML document 140 shown in FIG. 14.
Incidentally, in the document shown in FIG. 14, the document is
represented by varying the indentation depending on the depth of the
structure of the document, but this representation is provided only for
the purpose of facilitating viewing in the light of explanation of an
example of the document of a structured document. Actual documents are not
indented in many cases.
Referring to FIG. 14, as can be appreciated from the SGML document 140 of
the structured document in this example, as for the tags representing the
"header" in the lower structure constituting the "chap," both their start
tag and end tag do not appear in the document. In actuality, however, the
start tag "<header>" between the tag "<chap>" and its content portion
"What is SGML?" in the second line is omitted. Incidentally, whether or
not such a tag is omitted cannot be determined unless reference is had to
the aforementioned document type definition 130. Accordingly, the accurate
structure of the object document cannot be understood unless it is
considered in combination with the document type definition to which the
object document constantly conforms.
Since tags are thus omitted in an SGML document, processing for analyzing
the document structure (syntactic analysis processing using an SGML
parser) is first required when processing the SGML document. In the
processing of analysis of a document structure, processing is mainly
carried out in which collation is made with a document type definition
while analyzing the object document, and the omitted tags in the object
document are restored. In the syntactic analysis processing which is
executed in actual document processing, other processing (processing such
as restoration of attributes and expansion of entities) is also carried
out. Since attention is focused herein on the structure restoration
processing alone, a description will be given hereafter under the
assumption that the syntactic analysis processing is simply equivalent to
structure restoration processing.
If restoration processing of tags (structures) is carried out with respect
to the SGML document 140 illustrated in FIG. 14 as the object document, an
SGML document 150 such as the one shown in FIG. 15 is obtained. In the
SGML document 150 shown in FIG. 15, the underlined portions indicate
restored tags (structures). In this object document, the omitted tags are
restored by collating with the document type definition 130 such as the
one shown in FIG. 13. That is, in view of the rule on the structure of
"chap," the tag "<header"> must always exist next to the tag "<chap>," so
that the tag "<header>" is first restored next to the tag "<chap>."
Similarly, since the tag "<header>" must always exist next to the tag
"<fig>," so that the tag "<header>" is restored next to the tag "<fig>."
In addition, since their end tags are omitted, their respective end tags
"</header>," "</para>" and the like are restored next to their content
portions (at positions behind their corresponding ensuing tags). Thus, the
respective tags (structures) are restored as underlined in the drawing.
Next, a description will be given of processing in a case where a structure
is retrieved in the SGML document 150 in which the tags are restored and
its structure is represented. In a structured document, when document
editing is carried out, not only the retrieval of mere character strings
of the text but also the retrieval of structures making use of the
document structure becomes important processing in document processing.
This is because in cases where processing of a structured document is
performed, edit processing which makes use of structures of the document
structure is actively undertaken.
In the retrieval of a structured document, not only the conventional
retrieval of character strings (text retrieval) but also retrieval making
active use of the structures is effectively utilized. For example, in a
case where an attempt is made to retrieve a figure related to SGML in a
document, according to conventional retrieval processing, full-text
retrieval (character string retrieval) is carried out with respect to the
document, and the "related figure" is located from the character string in
the text.
However, if the structure itself of the document structure is used for
retrieval, it becomes possible to effect retrieval by pointing to a
structure in the document structure as in "a figure in which SGML is
included in the title of the figure" or "a header in the lower structure
of a figure," thereby making it possible to effect retrieval by narrowing
down the object. In addition, in retrieval processing in that case, since
retrieval can be effected by narrowing down the object range in accordance
with the structure of the document, there is an advantage in that the
efficiency of retrieval processing improves.
As described before, since the SGML document has a document architecture of
a type in which tags for marking are embedded in the text, its affinity
with a conventional text processing system is high. That is, since the
structures are represented by tags for marking, it is unnecessary to use a
special apparatus or processing program when retrieving the structure, and
it is possible to retrieve the document structure by using
character-string retrieval for retrieving a character string representing
the symbols of the tags. In other words, the SGML document can be prepared
by using a conventional text processing apparatus (such as a document
editor), and structure retrieval can be basically carried out by
retrieving the start tags and their corresponding end tags by using the
conventional text retrieving technique for character-string retrieval in
which character strings of the tags are retrieved.
As described above, in the SGML document, the tags (structures) are omitted
according to the designation of a document type definition. Accordingly,
there are cases where omissible tags are designated as objects to be
retrieved. For example, if a description is given by citing the SGML
document 140 illustrated in FIG. 14, in a case where the user performs
retrieval by designating the tag "<header>" by desiring to retrieve the
content of the structure called "header" from the document structure, this
tag is omitted in the original SGML document, so that the conventional
text retrieval method cannot be used unless the restoration processing of
the tag is conducted beforehand.
Accordingly, in the structure retrieval of a representation format such as
an SGML document, structure restoration processing for restoring the
omitted structure (tag) generally becomes indispensable. That is, before
the structure subject to retrieval is searched for, the structure (tag) is
restored by conducting the structure restoration processing with respect
to the entire object document, and processing for searching for the tag
subject to retrieval is then carried out by text retrieval.
Since the structure restoration processing must be carried out while
referring to the document type definition of the object document,
complicated processing is required, and a long processing time is
required, with the result that it takes time in retrieval processing. The
longer the object document is, the more time it takes in the restoration
processing of the structure, and the more retrieval time increases. For
this reason, such processing presents a problem in practical applications
when handling an SGML document as a structured document.
In contrast, when, for instance, the object document (SGML document) is
stored, it is possible to use a technique whereby the results of
restoration processing of the tags omitted are stored in advance after
being converted into an internal data structure. If this technique is
used, the structure restoration processing during retrieval of the
structure becomes unnecessary, so that the retrieval processing speed can
be improved.
With the above-described technique, however, since the object document must
be stored after being converted into the internal data structure, there is
a drawback in that a large storage area of such as an external storage
device is required. In addition, when documents are frequently exchanged
with external documents, processing for converting the documents into
internal data structures (structure restoration processing) is, in fact,
required on each such occasion, with the result that there arises a
drawback that the overall throughput cannot be improved. Namely, although
the above-described technique is effective for a large-scale document
database for managing object documents in one place, it cannot be
generally said that the technique provides an effective method when
processing a group of small-scale structured documents.
SUMMARY OF THE INVENTION
To solve the above problems, it is an object of the present invention to
provide a structure retrieval apparatus in which tags (part of the tags
are omissible) are inserted in data to discriminate portions of the data
to thereby express a structure, and the structure of the data is searched
at high speed.
To this end, in accordance with a first aspect of the present invention,
there is provided a structure retrieval apparatus comprising: data storing
means (11) for storing data in which tags are partially omissible when the
tags are inserted in the data, and the data is partially discriminated by
the tags so as to represent a structure; type storing means (12) for
storing a pattern of the structure represented by the tags; restoration
processing means (13) for restoring an omitted portion of the tag in the
data stored in the data storing means on the basis of the pattern of the
structure; and structure retrieving means (14) for controlling the
restoration processing means when a designated structure is retrieved, for
effecting processing of partially restoring the structure with respect to
necessary and minimum partial data concerning the tag of the structure
subject to retrieval, and for retrieving the tag of the structure subject
to retrieval on the basis of the restored partial data.
In accordance with a second aspect of the present invention, there is
provided a structure retrieval apparatus comprising: data storing means
(111) for storing data in which tags are partially omissible when the tags
are inserted in the data, and the data is partially discriminated by the
tags so as to represent a structure; type storing means (112) for storing
a pattern of the structure represented by the tags; restoration processing
means (113) for restoring an omitted portion of the tag in the data stored
in the data storing means on the basis of the pattern of the structure;
structure retrieving means (114) for retrieving a tag of a designated
structure; essential-structure searching means (115) for searching the
structure of the pattern for a tag which is located at a higher level than
that of the structure subject to retrieval and is not omissible, in a case
where the tag concerning the structure subject to retrieval is omissible;
and control means (116) for controlling the restoration processing means
so as to effect partial structure restoration processing on the basis of
necessary and minimum partial data concerning the tag by using the tag
found by the essential-structure searching means, and for controlling the
structure retrieving means so as to retrieve the structure subject to
retrieval.
In accordance with a third aspect of the present invention, there is
provided a structure retrieval apparatus comprising: data storing means
(121) for storing data in which tags are partially omissible when the tags
are inserted in the data, and the data is partially discriminated by the
tags so as to represent a structure; type storing means (122) for storing
a pattern of the structure represented by the tags; restoration processing
means (123) for restoring an omitted portion of the tag in the data stored
in the data storing means on the basis of the pattern of the structure;
structure retrieving means (124) for retrieving a tag of a designated
structure on the basis of the data stored in the data storing means;
essential-structure searching means (125) for searching the structure of
the pattern for a tag which is located at a higher level than that of the
structure subject to retrieval and is not omissible, in a case where the
tag concerning the structure subject to retrieval is omissible; and
control means (126) for controlling the restoration processing means so as
to effect structure restoration processing with respect to necessary and
minimum partial data concerning the tag by using the tag found by the
essential-structure searching means, and for replacing corresponding data
stored in the data storing means by restored data.
In the structure retrieval apparatus in accordance with the first aspect of
the present invention, the data storing means (11) stores data in which
tags are partially omissible when the tags are inserted in the data, and
the data is partially discriminated by the tags so as to represent a
structure. In addition, the type storing means (12) stores a pattern of
the structure represented by the tags. The restoration processing means
(13) restores an omitted portion of the tag in the data stored in the data
storing means on the basis of the pattern of the structure.
When a structure designated by, for instance, a user is retrieved with
respect to the data stored in the data storing means, the structure
retrieving means (14) controls the restoration processing means, effects
processing of partially restoring the structure with respect to necessary
and minimum partial data concerning the tag of the structure subject to
retrieval, and retrieves the tag of the structure subject to retrieval on
the basis of the restored partial data.
As a result, since the structure retrieving means (14) retrieves the tag
subject to retrieval by performing the processing for partially restoring
the structure with respect to only necessary and minimum partial data
concerning the tag of the structure subject to retrieval, the substantial
retrieval time can be shortened. For this reason, it is possible to effect
structure retrieval at high speed.
In the structure retrieval apparatus in accordance with the second aspect
of the present invention, the data storing means (111) similarly stores
data in which tags are partially omissible when the tags are inserted in
the data, and the data is partially discriminated by the tags so as to
represent a structure. The type storing means (112) stores a pattern of
the structure represented by the tags. The restoration processing means
(113) for restores an omitted portion of the tag in the data stored in the
data storing means on the basis of the pattern of the structure
When a structure designated by, for instance, a user is retrieved from the
data stored in the data storing means, the structure retrieving means
(114) retrieves a tag of the designated structure, and at that time the
essential-structure searching means (115) searches the structure of the
pattern for a tag which is located at a higher level than that of the
structure subject to retrieval and is not omissible, in a case where the
tag concerning the structure subject to retrieval is omissible. Then, the
control means (116) controls the restoration processing means so as to
effect partial structure restoration processing on the basis of necessary
and minimum partial data concerning the tag by using the tag found by the
essential-structure searching means, and controls the structure retrieving
means so as to retrieve the structure subject to retrieval.
Since the tag subject to retrieval can be retrieved as the control means
(116) controls the restoration processing means (113) by using the tag
found by the essential-structure searching means (115) and by merely
effecting partial structure restoration processing on the basis of
necessary and minimum partial data concerning the tag, the substantial
retrieval time can be shortened. For this reason, it is possible to effect
structure retrieval at high speed.
In the structure retrieval apparatus in accordance with the third aspect of
the present invention, the data storing means (121) similarly stores data
in which tags are partially omissible when the tags are inserted in the
data, and the data is partially discriminated by the tags so as to
represent a structure. The type storing means (122) stores a pattern of
the structure represented by the tags. The restoration processing means
(123) for restores an omitted portion of the tag in the data stored in the
data storing means on the basis of the pattern of the structure.
The structure retrieving means (124) retrieves a tag of a designated
structure on the basis of the data stored in the data storing means (121).
When so doing, the essential-structure searching means (125) searches the
structure of the pattern for a tag which is located at a higher level than
that of the structure subject to retrieval and is not omissible, in a case
where the tag concerning the structure subject to retrieval is omissible.
Then, the control means (126) controls the restoration processing means
(123) so as to effect structure restoration processing with respect to
necessary and minimum partial data concerning the tag by using the tag
found by the essential-structure searching means, and replaces
corresponding data stored in the data storing means by restored data.
As a result, since the data in which the tags subject to retrieval are
partially restored are consecutively replaced and are stored in the data
storing means (121), the structure retrieving means (124) subsequently
retrieves the tag of the designated structure on the basis of the data
stored in the data storing means (121). In this case as well, since the
tag subject to retrieval can be retrieved as the control means (126)
controls the restoration processing means (123) by using the tag found by
the essential-structure searching means (125) and by merely effecting
partial structure restoration processing on the basis of necessary and
minimum partial data concerning the tag, the substantial retrieval time
can be shortened. For this reason, it is possible to effect structure
retrieval at high speed. In addition, since the data in which the tags
subject to retrieval are partially restored are consecutively replaced and
are stored in the data storing means (121), in structure retrieval which
is subsequently conducted, there are cases where it is unnecessary to
carry out the structure restoration processing. Consequently, there is an
additional advantage in which the substantial retrieval time can be
shortened.
Thus, in accordance with the structure retrieval apparatus in accordance
with the present invention, a structure retrieval apparatus is provided in
which tags are inserted in data, the data is partially discriminated by
the tags so as to represent a structure, and the structure can be
retrieved at high speed from the data in which the tags are partially
omissible. In addition, a structure retrieval apparatus is provided in
which, by converting the data into document text, tags are inserted into
the text, the text is divided into document elements, and the structure
can be retrieved at high speed from the structured document.
The above and other objects, features and advantages of the present
invention will become more apparent from the following detailed
description of the invention when read in conjunction with the
accompanying drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a block diagram illustrating a configuration of a structure
retrieval apparatus in accordance with a first embodiment of the present
invention;
FIG. 2 is a diagram illustrating an example of an operating screen in a
case where a structure retrieving instruction is given;
FIG. 3 is a flowchart explaining structure retrieval processing by a
structure retrieving section;
FIG. 4 is a flowchart explaining processing for retrieving an object
structure from a partially restored structure;
FIG. 5 is a flowchart illustrating a processing flow of an
essential-structure searching section;
FIG. 6 is a flowchart illustrating a processing flow of a structure
restoring section;
FIG. 7 is a diagram specifically explaining the manner in which a structure
of a structured document is partially restored in a case where a
corresponding end tag is omitted;
FIG. 8 is a diagram illustrating another form of document type definition
stored in a type storing section;
FIG. 9 is a diagram illustrating another example of the structure
retrieving instruction;
FIG. 10 is a diagram illustrating still another example of the structure
retrieving instruction;
FIG. 11 is a block diagram illustrating a configuration of a structure
retrieval apparatus in accordance with a second embodiment of the present
invention;
FIG. 12 is a block diagram illustrating a configuration of a structure
retrieval apparatus in accordance with a third embodiment of the present
invention;
FIG. 13 is a diagram illustrating an example of the document type
definition (DTD) of SGML;
FIG. 14 is a diagram explaining an example of an SGML document in which
tags are omitted; and
FIG. 15 is a diagram explaining an example of the SGML document in which
the omitted tags are restored.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
Referring now to the accompanying drawings, a description will be given of
the embodiments of the present invention. FIG. 1 is a block diagram
illustrating a configuration of a structure retrieval apparatus in
accordance with a first embodiment of the present invention. In FIG. 1,
reference numeral 11 denotes a document-data storing section; 12, a type
storing section; 13, a structure restoring section; 14, a structure
retrieving section; 15, and an essential-structure searching section.
Numeral 16 denotes a structure retrieving instruction for retrieving a
structure, and 17 denotes a retrieved result.
Document data for obtaining a structured document is stored in the
document-data storing section 11, wherein the structured document is
formed such that tags are inserted in the text data of a document, and the
document text is partially distinguished (as document parts) by the tags.
This document data is similar to, for instance, the SGML document 140
(FIG. 14) described before, and is the document data of a document
architecture in which the structure of the document is represented by
tags, and the tags are partially omissible. As a pattern of the structure
represented by tags, a document type definition 130, such as the one shown
in FIG. 13, is stored in the type storing section 12 in correspondence
with the document data of the structured document. In addition, the
structure restoring section 13 restores omitted portions of the tags in
the document data stored in the document-data storing section 11, on the
basis of the pattern (document type definition) stored in the type storing
section 12.
Upon receiving the structure retrieving instruction 16 from a user or
another apparatus, the structure retrieving section 14 effects retrieval
processing of the structure with respect to the document data stored in
the document-data storing section 11. At that time, by controlling the
structure restoring section 13, the structure retrieving section 14
effects processing for partially restoring the structure with respect to
necessary and minimum partial data concerning the tags of a structure
subject to retrieval, and retrieves the tags of the structure subject to
retrieval by means of the restored partial structure.
This structure retrieving section 14 includes the essential-structure
searching section 15 as a part of its processing function, and the
essential-structure searching section 15 searches for a necessary and
minimum portion concerning the tags of the structure subject to retrieval.
Subsequently, the structure restoring section 13 partially restores the
structure with respect to the found necessary and minimum document data,
and processing is then effected for searching for the object structure
(tags) in the document by means of the restored partial structure. The
pattern (document type definition) of the structure stored in the type
storing section 12 is referred to at the time of structure restoration
processing by the structure restoring section 13, and is also referred to
when the essential-structure searching section 15 searches for the
necessary and minimum portion concerning the tags of the structure subject
to retrieval.
In the above-described manner, the structure retrieving section 14 effects
processing for partially restoring the structure with respect to only the
necessary and minimum partial data concerning the tags of the structure
subject to retrieval, retrieves the tags (structure) subject to retrieval,
and outputs the retrieved result 17. For this reason, as for the retrieval
of the structure on the basis of the document data of the structured
document here, the substantial retrieval time can be shortened, and the
retrieval of the structure can be effected at high speed.
FIG. 2 is a diagram illustrating an example of an operating screen in the
case where an instruction for receiving the structure is given. A
structure retrieving request from the user or another apparatus or the
like opens a subwindow 20 of a structure retrieving property, and a
retrieving instruction is given there by designating constraints in
retrieving the structure. As for the constraints in the structure
retrieval, an object structure is designated by a field 21 for designating
an object structure, a constraint of the content is designated by a field
22 for designating a content constraint, and a prescribed condition for
constraining the structure is designated by a field 23 for designating a
structural constraint. All the constraints in retrieving the structure
designated by these three fields may not necessarily be designated, and
the retrieval of the structure may be executed by the designation of some
constraints. In that case, however, the desired structure is not
sufficiently narrowed down, and there is a possibility that many
structures satisfying the conditions may be retrieved.
As the structure retrieving instruction 16 having such constraints is
given, the structure retrieving section 14 starts the retrieval processing
of the structure. Next, a more detailed description will be given of the
contents of processing by the structure retrieving section 14. In the
basic processing of retrieving the structure by the structure retrieving
section 14, the structure (tags) subject to retrieval is accepted from the
user or another apparatus, the essential-structure searching section 15
and the structure restoring section 13 are accessed to request processing,
and processing is effected in which the partial structure of the restored
document is searched for the structure (tags) subject to retrieval, and
the retrieved result is returned to the user or the apparatus.
Incidentally, as for the form of the structure retrieving instruction 16
from the user or another apparatus which is delivered to the structure
retrieving section 14, an instruction for retrieving the structure is
given in the form of a structure retrieving property sheet (FIG. 2) or a
command in conformity with the form of the apparatus such as an apparatus
whereby the user directly edits a document, such as a word processor or a
document editor. In addition, in the case of an apparatus which is
connected to a network, e.g., an apparatus such as a retrieval server, its
structure retrieving instruction is given by a predetermined protocol
through the network from another apparatus. In either case, there is no
difference in the contents (constraints of the structure) of the given
information in their structure retrieving instruction.
Also, the presentation methods for outputting the retrieved result here are
similar in the respective cases. The retrieved result is outputted in
conformity with the form of the apparatus such as an apparatus whereby the
user directly edits a document, such as a word processor or a document
editor. In | | |
