 |
Description  |
|
|
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a method and an apparatus for formatting a
document in accordance with the structure of the document.
2. Description of the Background Art
Recently, there has been remarkable progress in so-called desk-top
publishing as well as in Japanese word-processors, in addition to the
advances in the output devices such as displays and printers which make it
possible to produce documents in much more diverse and effective styles.
However, producing documents which are both neat-looking and
easily-readable requires thorough understanding of operations, commands
and formats of a document formatting apparatus to be used, which makes
such a document formatting apparatus almost inaccessible to those without
formal training.
As a solution to this situation, progress has been made in developing an
automatic document formatting system which utilizes an automatically
extracted logical structure of a document, along with a document
formatting system capable of arranging figures and articles according to
automatically derived referential relationships between figures and
articles.
Although such logical structures and referential relationships are
generally sufficient to provides structural information on the document
necessary for effective formatting, ambiguities involved in human
languages may lead to misapprehension of the logical structures and
referential relationships. Furthermore, apart from this problem, outputs
of such automatic document formatting apparatus may not satisfy personal
or temporal demands of a user.
On the other hand, there are document formatting systems which utilize
command codes indicating the manner of formatting to be embedded in the
document, such as `Roff`. More recently, there appeared document
formatting systems such as `Scribe` or `T.sub.E X` which use document data
and format data independently so a change in command codes can be made at
once on the format data, without looking for every embedded command code
as in `Roff`. Moreover, `Scribe` and `T.sub.E X` are capable of performing
more sophisticated formatting than others. But, in these document
formatting systems, a thorough understanding of command codes is
indispensable for skillful maneuvering. Furthermore, even in `Scribe` and
`T.sub.E X`, embedding of command codes is necessary which can easily be
tedious. Although automatic document formatting apparatuses mentioned
above are free of such problems concerning command codes, they are, as
described above, prone to misapprehension of the logical structures and
referential relationships due to the ambiguities in the human languages,
and the outputs of such automatic document formatting apparatus may not
satisfy personal or temporal demands of a user.
Thus, with conventional document formatting apparatuses, either a
possibility of misapprehension resulting from automatic extraction of
document structures or else difficulties in dealing with command codes
which need to be embedded in the document and thoroughly mastered by the
user, have to be tolerated.
SUMMARY OF THE INVENTION
It is therefore an object of the present invention to provide a method and
an apparatus for formatting a document which are simply manipulatable as
well as consistently correct.
According to one aspect of the present invention, there is provided an
apparatus for formatting a document which contains command codes
indicating prescribed commands concerning structural function, comprising:
a unit for entering the document in to the apparatus; a unit for
structurally analyzing the document and deriving structural information
from the document; a unit for detecting the command codes and adjusting
the analysis by the structurally analyzing a unit in accordance with the
indications of the command codes; and a unit for carrying out formatting
of the document in accordance with the structural information.
According to another aspect of the invention, there is provided a method of
formatting a document which contains command codes indicating prescribed
commands concerning structural functions, comprising the steps of:
structurally analyzing the document and deriving structural information
from the document; detecting the command codes and adjusting the analysis
at the structurally analyzing step in accordance with the indications of
the command codes; and formatting the document in accordance with the
structural information.
Other features and advantages of the present invention will become apparent
from the following description taken in conjunction with the accompanying
drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a block diagram of a document formatting apparatus according to
one embodiment of the present invention.
FIG. 2 is a tabulated illustration of examples of command codes to be
utilized in the document formatting apparatus shown in FIG. 1.
FIG. 3 is another tabulated illustration of logical structures to be
utilized in the document formatting apparatus shown in FIG. 1.
FIG. 4 is a flow chart for the operation of formatting by the document
formatting apparatus shown in FIG. 1.
FIG. 5 is an illustration of a document being formatted by the document
formatting apparatus shown in FIG. 1, showing the manner in which the
command codes are used in this embodiment.
FIGS. 6(A) and (B) are tabulated illustrations of logical structures for
the document shown in FIG. 5, obtained without and with the command codes.
FIG. 7 is another illustration of a document being formatted by the
document formatting apparatus shown in FIG. 1, showing the manner in which
the command codes are used in this embodiment.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
Referring now to FIG. 1, there is shown one embodiment of a document
formatting apparatus according to the present invention.
This document formatting apparatus comprises an input unit 10 from which a
document to be formatted with command codes embedded is entered, an
original document memory 20 for storing the document with the command
codes entered at the input unit 10, a format memory 30 for storing a
format into which the document is to be formatted, an analyzing unit 40
comprising a command code analyzing unit 50 for analyzing contents of the
command codes, and a document structure analyzing unit 60 for analyzing
logical and referential structures of the document in accordance with the
analyzed contents of the command codes, a document structure memory 70 for
storing the analyzed structures of the document, a formatting processing
unit 80 for carrying out formatting in accordance with the structures of
the document stored in the document structure memory 70 and the format
stored in the format memory 30, an output unit 90 for presenting the
document as formatted by the formatting processing unit 80, and an
administration unit 100 for administerring the operations by all these
parts of this document formatting apparatus mentioned above.
The input unit 10 may take a form of a keyboard, a mouse, or a
communication network. The output unit 90 may take a form of a CRT, a
display, or a printer.
In the analyzing unit 40, when the command codes are present in the
document, the command code analyzing unit 50 adjusts operation of the
document structure analyzing unit 60 such that the contents of the command
codes are reflected in the manner of analyzing the logical and referential
structures of the document. Thus, in this embodiment a user can
deliberately control the analysis of the logical and referential
structures. Moreover, the command code analyzing unit 50 also deciphers
those command codes which are directly concerned with the layout of the
document, so that the user can also have control over the layout of the
document.
An example of a set of command codes to be utilized in this embodiment are
shown in TABLE I. of FIG. 2. As given in the section (0) of FIG. 2, any
command code begins with a symbol in this embodiment, and what follows
this symbol designates a type of command code. Various different types
of command codes and their corresponding logical attributes are summarized
in the section (1) of FIG. 2. For example, a command code DATE indicates
that what follows is the date, a command code NODC indicates what follows
is the document number, and so on. In addition, there are analysis
prohibition codes shown in the section (2) of FIG. 2 which indicate a
portion of the document not to be structurally analyzed such as those
containing mathematical formulae and those requiring a special type of
formatting. Furthermore, there are graphic reference codes as shown in the
section (3) of FIG. 2 which indicate the presence of a reference in the
document to graphics, as well as the location of data on the graphics
being referred to and the layout of the graphics, i.e., how the graphics
are to be incorporated into the final output of the document. There is
also a compulsory return code shown in the section (4) of FIG. 2 which
indicates forcible changing to the next line, regardless of the logical
structure.
An example of the logical structure in the document structure memory 70 is
shown in TABLE II. of FIG. 3. The logical structure in the document
structure memory 70 comprises a sentence number labelling each sentence of
the document, an attribute for each sentence, a level given to each
attribute, and a header pattern for relevant sentences. For example, in
FIG. 3 a sentence labelled by the sentence number 4 is given an attribute
of Paragraph End in response to the command code HEAD present in that
sentence, whose level is given as 3 and which has no header pattern. Such
a logical structure will be utilized along with the format data in the
format memory 30 in carrying out formatting of the document.
Referring now to FIG. 4, the operation of document formatting by this
document formatting apparatus will be described.
At the step 100, the document with the command codes is entered from the
input unit 10 and stored in the original document memory 20.
Then, one sentence of the document is read out from the original document
memory 20 and provided to the analyzing unit 40 at the step 101.
Then at the step 102, whether there is any command code in this sentence is
determined by the command code analyzing unit 50.
When there is no command code in the sentence, an ordinary automatic
extraction of logical and referential structures of the document is
carried out by the document structure analyzing unit 60 at the step 103
and the process proceeds to the step 113 to be explained below.
On the other hand, when there is a command code in the sentence, the step
104 will be taken in which whether the command code present in the
sentence is one of the logical structure codes is determined by the
command code analyzing unit 50.
When the command code is one of the logical structure codes, the logical
structure of the sentence is extracted by the document structure analyzing
unit 60 at the step 105 in accordance with what the command code indicates
and the process proceeds to the step 113.
Otherwise, the step 106 will be taken in which whether the command code
present in the sentence is one of the analysis prohibition codes is
determined by the command code analyzing unit 50.
When the command code is one of the analysis prohibition codes, the
structural analysis by the document structure analyzing unit 60 is
controlled at the step 107 in accordance with what the command code
indicates and the process proceeds to the step 113.
Otherwise, the step 108 will be taken in which whether the command code
present in the sentence is one of the graphic reference codes is
determined by the command code analyzing unit 50.
When the command code is one of the graphic reference codes, the
referential structure of the sentence is extracted by the document
structure analyzing unit 60 at the step 109 in accordance with what the
command code indicates and the process proceeds to the step 113.
Otherwise, the step 110 will be taken in which whether the command code
present in the sentence is the compulsory return code is determined by the
command code analyzing unit 50.
When the command code is the compulsory return code, the information of the
compulsory return is deciphered and extracted by the command code
analyzing unit 50 at the step 111 and the process proceeds to the step
113.
Otherwise, the command code present in the sentence is in error since it is
not any one of command codes given in FIG. 2, so at the step 112 the error
in the command code is corrected, and the process proceeds to the step
113.
At the step 113, the result of the logical and referential structures
obtained up to this point is stored in the document structure memory 70.
Then at the step 114, whether all the sentences in the document have been
checked is determined. When all the sentences in the document have not
been checked, the process returns to the step 101 and the steps following
will be repeated.
Otherwise the process proceeds to the step 115, at which the formatting of
the document is carried out by the formatting processing unit 80 in
accordance with the logical and referential structures stored in the
document structure memory 70 as well as with the format stored in the
format memory 30, and as the resulting formatted document is outputted by
the output unit 90 the process terminates.
One practical example of the document and the use of the command codes is
shown in FIG. 5.
In this example of FIG. 5 which is a report 11 entitled `Intelligent
document processing system`, the first line "INTELLIGENT DOCUMENT
PROCESSING SYSTEM" can be identified as a title by the ordinary automatic
logical structure extraction so that no command code is necessary in this
first line. On the other hand, the second line "RACHI YOZAN" will most
likely not be identifiable as the name of the author as Rachi Yozan is a
very rare name so that this name cannot be found in a name data-base.
Thus, that "RACHI YOZAN" is a name of the author is indicated by placing
the command code AUTH at the top of this second line. Accordingly, the
document structure analyzing unit 60 can construe "RACHI YOZAN" as the
name of the author correctly. Likewise, "NICHIBEI SOFTWARE" in the third
line can be identified correctly as the name of the organization to which
the author belongs by placing the command code SECT at the top of this
third line. As for the fourth line, there is a compulsory return code at
the top of this fourth line, as this fourth line is to be left blank. The
fifth line "1. INTRODUCTION" can be identified as a section header with
`introduction` as a reserved word, by ordinary automatic logical structure
extraction so that no command code is necessary in this fifth line. Also,
the thirty-third line " S FIG. 10 ZF bunsho-1 ZP d E . . . " means that
"FIG. 10 is in the file `bunsho-1` and it is to be laid out in the lower
part of the current page . . . ".
How the logical structure obtained by the document structure analyzing unit
60 is affected by the presence of the command codes in the example of FIG.
5 is shown in FIGS. 6(A) and (B), where TABLE III. of FIG. 6(A) shows the
logical structure obtained from the document without the command codes and
TABLE IV. of FIG. 6(B) shows the logical structure obtained from the
document with the command codes. As can be seen from FIG. 6, without the
command codes, i.e., by a completely automatic logical structure
extraction, the second and the third lines are construed incorrectly as
sub-titles in FIG. 6(A), whereas with the use of the command codes this
misapprehension can be avoided in FIG. 6(B).
Another practical example of the document and the use of the command codes,
in particular the use of the analysis prohibition codes, is shown in FIG.
7.
In this example of FIG. 7 which is a portion of an article 12 containing
mathematical formulae, the analysis prohibition start line code X and the
analysis prohibition end line code Y are placed at the top and the
bottom, respectively, of the mathematical formulae so that this portion
will not be structurally analyzed, as the meaning of the mathematical
formulae is not analyzable by the document structure analyzing unit 60.
Apart from the mathematical formulae, the analysis prohibition codes can
similarly be used for those portions which are written in a special or
personal manner. In addition, the compulsory return code C can be placed
in a middle of a line as shown in FIG. 7.
As described, according to this embodiment it is possible to have a
document formatting apparatus which is both simply manipulatable and
consistently correct. It can be seen from the above description that this
is due to the particular use of both the automatic logical and referential
structures, and the command codes. Consequently, according to this
embodiment the misapprehension inevitably accompanying the completely
automatic structural analysis can be rectified by the use of the command
codes. At the same time, as this embodiment requires the use of the
command code in only those places which may causes misapprehension, and
not elsewhere, the encumbrances associated with the conventional usage of
the command codes can be lessened enormously.
It is to be noted that although in the above embodiment, the command codes
are placed at the top of lines, this can easily be modified to allow
placement of the command code anywhere in the lines.
Also, although in the above embodiment the only command code which is
directly concerned with the layout of the document is the compulsory
return code, it is possible to incorporate more complicated command codes
such as font command codes and style command codes used in `Scribe`,
thereby enhancing the scope of possible formatting.
Also, although in the above embodiment the structural analysis is performed
line by line, this can be performed in so-called Top-Down fashion, i.e.,
the entire document all at once.
Moreover, the analysis prohibition codes X and Y can be modified such
that within a portion indicated by them the analysis is to be performed by
a particular processing system such as T.sub.E X which is known to be
highly effective in dealing with mathematical formulae.
Furthermore, this embodiment can be effectively employed not only for a
document containing references to graphics as in the examples in the
foregoing description, but also for a document having references to a list
of references or bibliography, and for relational data-bases.
Besides these, many modifications and variations of the above embodiment
may be made without departing from the novel and advantageous features of
the present invention. Accordingly, all such modifications and variations
are intended to be included within the scope of the appended claims.
* * * * *
|
|
 |
|
 |
Description |
|
