JPS62229364A - Document processor - Google Patents

Abstract

PURPOSE:To attain the automatic evolution and output of the document data in the form of a specified layout and according to a layout rule corresponding to the logic structure, by obtaining the logic structures of the chapters, paragraphs, itemization, the ends of paragraphs, etc., from the data of inputted sentences. CONSTITUTION:The document data supplied from an input part is stored in an original memory part 3 via a document control part 1. At the same time, a segmenting code, e.g., a renewing row code is detected out of the document data and a group of sentences which are segmented by the segmenting codes are successively extracted. Here the length of a single sentence is also measured. A header extracting part 6 decides whether or not an extracted sentence is possibly used as a header based on the length information of the single sentence and with reference to a headword dictionary 6a. Thus a header candidate is obtained if said sentence complies with a standard. This header candidate is compared with the header rule stored in a header rule dictionary 7a through a header deciding part 7. Then a header sentence is decided if the header candidate is coincident with the header rule. While a document text, like the end of a paragraph, etc., is decided if no header sentence is confirmed.

Description

[Detailed Description of the Invention] [Objective of the Invention] (Field of Industrial Application) The present invention effectively changes the output layout format of document data according to a hierarchical logical structure obtained by analyzing the document structure of document data. The present invention relates to a document processing device capable of supporting the document processing apparatus.

(Prior Art) In document processing devices such as word processors, document data is manually input as a series of code information such as character codes and punctuation mark codes. The document data indicated by the code information sequence is then registered in a document file or output to a printer or display.

However, if the character string indicated by the code information is simply output as is, the document will be extremely difficult to read. Therefore, the document format is generally formatted by inserting a line feed code at the break point of a coherent sentence, and by inserting a space code at the beginning of the sentence following the line break point to create a paragraph. things are done.

Furthermore, the entire document can be divided into multiple ranges, such as chapters and sections, and headings can be added to each range, and control codes such as tabs and indents can be inserted to make the document easier to read. It will be done.

However, creating a document while inserting control codes that are not directly related to the original document data, such as double line feed codes, interferes with the thinking process and causes a decrease in document creation efficiency. There is.

Further, when re-editing a document created in this manner, it is necessary to delete the above-mentioned control code or insert it in another location.

At this time, the document structure changes depending on the control code inserted into the document data, so when re-editing a document that spans several pages, it requires a great deal of effort to unify the document structure as a whole. There were problems such as. For example, in order to unify the document format, it is necessary to refer to the document format of several pages earlier. Therefore, it is not possible to proceed with the document editing process easily, and it is not possible to improve the processing efficiency.

(Problems to be Solved by the Invention) The present invention has been made in consideration of these circumstances, and its purpose is to actively utilize the hierarchical logical structure of documents to process documents. It is an object of the present invention to provide a document processing device that eliminates the troublesomeness of document processing described above and enables effective document processing.

[Structure of the Invention] (Means for Solving the Problem) The present invention refers to a headword dictionary and a headline rule dictionary for document data given as a series of code information, and extracts a header sentence from the document data. For each sentence extracted and divided into heading sentences and non-heading sentences, the document structure is determined by referring to the document structure rule dictionary, and the document data is divided into chapters, sections, bullet points, etc. according to this document structure.
A hierarchical logical structure represented by a document structure such as a paragraph is obtained, and the document data of each sentence is developed in the output section according to a layout rule corresponding to the document structure of each sentence represented by this hierarchical logical structure. It is something.

In other words, according to the hierarchical document structure determined for the input document data, each document data is expanded to the output section based on the layout rules corresponding to the logical structure of each sentence making up the document. .

(Operation) Thus, according to the present invention, the development structure of the document data of each sentence to the output section, that is, the output layout format, is automatically determined according to the hierarchical logical structure of the given document data.
The document will be displayed or printed in a layout format that corresponds to the document structure.

Therefore, when inputting document data or editing it, it is possible to process the document without considering its output layout format, that is, without considering the insertion position of line feed codes or setting indents and tabs. It becomes possible. Moreover, even if you arbitrarily process documents in this way,
The document data is expanded into a predetermined layout format and displayed on a display or printed out.

(Example) Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a schematic configuration diagram of an embodiment device. In FIG. 1, reference numeral 1 denotes a document management section forming the main body of the apparatus.

The document data inputted as a series of code information through the input section 2 consisting of a keyboard etc. has a document structure as shown in FIG. 2, for example, and is stored in the original text under the control of the document management section l. The document is stored in section 3 and is used for document processing. The document data processed by the document management section 1 is displayed on the display section 5 under the control of the display control section 4.

FIG. 3 shows a processing procedure for document structure analysis on input document data, and the functions of the present apparatus will be explained along this flow. The document management section 1 inputs document data from the input section 2, stores it in the original text storage section 3 in step a), detects a delimiter code, such as a line feed code, in the document data, and uses this delimiter code to A group of separated sentences is extracted in order. At the same time part 1
Measuring the length of sentences. Then, the execution of the following processes is managed and controlled for each extracted sentence.

The heading extractor 6 uses information on the length of the sentence measured as described above and the headword dictionary 6a to determine whether or not the sentence extracted by the document management unit 1 has a possibility of being a heading. A determination is made by referring to it (step b).

This headword dictionary 6a is a dictionary in which words and symbols that appear frequently as headers are registered in advance by classifying them into categories as shown in FIG. 4, for example.

Specifically, words and phrases that appear frequently as headings, such as "Introduction" and "Synopsis," are registered together in a category called "headline reserved words," and numbers and symbols that appear frequently as headings are registered in each category. Each item is registered together.

The heading extraction unit 6 determines whether the length of the extracted sentence is within a predetermined number of characters (for example, 40 characters), and if it is within the predetermined number of characters, determines that it may be a heading. ing. Then, for this sentence, the sentence (words, numbers, and symbols indicated by the code information series) is found in the headword dictionary 6.
In step b), if the corresponding word is registered in the headword dictionary Ba, it is set as a headline candidate (step C).

The heading determination unit 7 determines whether the sentence extracted as a heading candidate by the heading extracting unit 6 is the It is determined whether or not it matches the heading rule (step C). If the heading candidate matches the heading rule, it is determined that it is a heading sentence (step e). Note that sentences for which no headword is detected by the heading extraction unit 6, and sentences that are judged as heading candidates but do not fall under the heading rules are determined to be not heading sentences (step f). In other words, it is determined that it is the body of the document such as a paragraph.

For each sentence determined as a heading sentence or other sentence as described above, the document structure determination unit 8 performs the following, for example, as shown in FIGS. 6(a) to 6(d) According to the document structure rules as shown, the document logical structure, such as whether the sentence is a chapter heading, a section heading, or a paragraph, is determined (step g). The logical structure of each sentence determined by the document structure determining unit 8 is determined (step h), and information on the logical structure is stored in the logical structure storage unit 10 in association with each sentence (step i).
).

Note that for sentences for which the logical structure analysis based on the document structure rules has failed, it is determined that there is an error in the logical structure of the sentence, and error processing is performed, such as prompting the user to correct the error in the input document data (step j).

The document development section 9 develops each human-written sentence to the output section in a layout format determined according to the logical structure, according to the hierarchical logical structure of the input document data obtained by the above-described document structure analysis process. That is, the document development section 9
Layout rules! 9a stores layout rules for determining an output layout format determined according to the type of logical structure. The document expansion section 9 obtains a layout rule for displaying the document data of each sentence on the display section 5 from the layout rule dictionary 9a according to the logical structure obtained for each sentence as described above,
The document data of each sentence is developed in the display control section 4 according to the layout rules.

The display control unit 4 controls the display of the document data on the display unit 5 according to the expanded structure of the document data expanded in this manner.

According to the apparatus thus configured, the hierarchical logical structure of input document data is analyzed in the following manner.

When document data is input from the input section 2, the document data is sequentially stored in the original text storage section 3 and is delimited by the document management section 1. This delimiter process is performed when the manual code information is a line feed code, space code, or
This is performed by determining whether the input code is a delimiter such as J r; J r; At this time, the length of the sentence (number of characters) is measured by, for example, counting the length of the series of code information separated by the delimiter code.

However, now, to explain the case where the document data shown in FIG.
When the sentence "Okawa Abe" in the row is given, the heading extraction unit 6 determines that each of these sentences is not a heading, since the corresponding words are not registered in the headword dictionary 6a. The document structure determination unit 8 determines that the attribute of the sentence in the first line is a title because it conforms to the rule that it is a noun phrase that appears at the beginning of the document. The sentence in the second line is determined to be a proper noun, especially a proper noun indicating a person's name, and an author's name according to rules such as a sentence appearing after a title.

After that, when the sentence "1. Introduction" in the third line is given, the words rlJr and Jr. Introduction 1 forming this sentence are respectively found from the headword dictionary 6a. As a result, this sentence is determined as a heading candidate A, and at the same time, the category that constitutes the heading candidate is determined as "(number part) (suffix part) (headline reserved word)".

Then, the heading determination unit 7 refers to the heading rule dictionary 7a to check whether the structure of the sentence determined as the heading candidate A conforms to the heading rules.

In this heading determination process, first, the arrangement of categories constituting the heading candidate A is analyzed, and it is determined whether the analyzed structure satisfies the conditions for a heading shown in FIG. 5. In this case, the category "(number part) (suffix part) (headline reserved word)" is analyzed as shown in Figure 7 according to the rules shown in Figures 5(a) to (d), and forms a headline pattern. Since it is confirmed that the heading is heading B, it is determined that it is heading B. If it is determined through this determination process that the arrangement of categories does not match any of the conditions shown in FIG. 5, it is determined that the headline candidate is not a headline.

When the heading B is determined, the document structure determination section 8 determines which of the rules illustrated in FIGS. 6(a) to 6(d) the document structure corresponds to. In this case, the logical structure of the sentences analyzed so far is the "title" and "author name" as described above, and since no chapter headings appear, the heading rules shown in Figure 6 (a) are used. It is compared with As a result of this matching, the heading candidate for the sentence "1. Introduction" is found under the conditions (1) (1,
1) (1,1,1) (1,1,,1゜1) was found to match, and the heading "1゜Introduction" was found to be the chapter heading C
is uniquely determined to constitute the . Information on the logical structure is then stored in the logical structure storage section 10.

Thereafter, when a sentence spanning the fourth to fifth lines is input, the number of characters exceeds a predetermined number of characters that can be used as a heading, and therefore it is determined that the sentence is a sentence other than a heading. In this case, since the sentence corresponds to the document structure rules shown in FIG. 6(c), it is determined that the sentence constitutes a paragraph.

Thereafter, it is similarly determined whether the sentences separated by the delimiter codes are heading sentences, and the document structure rule dictionary 8
a and its document structure is sequentially cleared.

For example, when a sentence forming a heading is input again, it is detected as heading candidate A in the same way as in the previous example, and it is detected as heading candidate A in the same way as in the previous example.
It is determined that the heading rules shown in ) to (d) are met. Regarding this heading B, it is shown that a chapter heading has already been detected from the contents of the logical structure storage section 10, so first, referring to the document structure rules shown in FIG. 6(b), Find out which conditions the pattern applies to. Then, the pattern of the heading candidate is the condition (
1,1) (2,1) (3,1) (4,1), it is determined that the heading may be at the same level as the previously obtained chapter heading C. . After that, it is checked whether the rules shown in FIG. 6(a) are met, and the conditions (1) (1, 1) of the document structure rules shown in FIG. 6(a) are determined.
(1, 1, 2) (1, 1, 2, 2) (1, 1, 2, 2,
1), the heading is determined to be a chapter heading at the same level as the previously determined chapter heading.

In addition, when a heading with a circled number at the beginning is given, no heading with a similar pattern has been detected before, so it is checked against the rules (conditions) shown in Figure 6(b). Matching is unsuccessful.

As a result, it is determined that the heading is at a different level from the previously obtained heading. After that, by checking against the document structure rules shown in FIG. 6(a), for example, the condition (1
)(1,1)(1,1,2)(1,1,2,2)(1,
1, 2, 2, 2) (1, 1, 2, 2, 2, 1) is detected, and the heading sentence is determined to be a bulleted heading.

Note that when a sentence is determined to be a paragraph, it may not be possible to know what level of heading the paragraph has received. In such a case, the connection relationship between paragraphs and headings may be determined by referring to the rules shown in FIG. 6(d), and the level thereof may be determined.

Through such document structure analysis processing, the document structure of each sentence constituting the input document data is obtained, and information on the hierarchical logical structure is created corresponding to the input document data, for example, as shown in FIG. logical structure storage unit 10
is stored in

To briefly explain the logical structure data of the document shown in FIG. 8, information on the document structure is shown enclosed in parentheses in correspondence with the document data shown in parentheses.5. The numerical value at the beginning indicates the hierarchical level of the sentence, the next information indicates its character, and the data starting with 2 indicates the analysis result of the document data. The analysis results of this document data have meanings as shown in the following table, for example.

The analysis result is expressed by adding information about what the document data is or what character it is after the symbol having such a meaning.

Now, the document development section 9 develops the document data to the output section in the following manner according to the information on the hierarchical logical structure of the document data obtained as described above.

FIG. 9 shows the 71-ware configuration of the document development section 9. The logical structure data reading section 11 reads the document data as shown in FIG. 8 stored in the logical structure storage section 10 and its logical structure data. Each character is cut out and given to the document data allocation section 13. At this time, the civilian extractor 14 is the rttle Jr shown in FIG.
Civilian data such as outerJ is extracted and provided to the layout information detection section 15, and the layout information analysis section 16 reads out the logical structure information described with the above-mentioned symbol 2 attached.

Here, the layout rule dictionary 9a includes, for example, FIG.
As shown in a) and (b), layout rules are stored that consist of information about the frame (document frame) for outputting document data and information about how to layout and output the sentence according to the civilian character. . Specifically, as shown in FIG. 10(a), the frame information consists of a frame management number on an output device such as a display or printer, information on its output position i, and information on the size of the frame. As a result, the structure of a display screen on which document data is output or a frame on printing paper is defined as shown in FIG. 11, for example. This frame information is defined for each of a plurality of frames set to output document data.

The layout rule reading unit 17 reads out such a frame rule from the layout rule word #9a and stores it in the frame format information buffer 18, and also provides the information to the calculation unit 19, and calculates the maximum number of columns and the maximum number of columns of the frame. I'm looking for the number of lines. The data on the maximum number of columns and lines is set in the maximum column and line value buffer 20 and is used to control the expansion and output of character data, which will be described later.

Further, the layout rules related to civilian characteristics stored in the layout rule dictionary 9a are defined, for example, as shown in FIG. 10(b). This layout rule defines how document data for each type of civility is laid out on the frame defined as described above.

Here, for example, "I!r" indicates to force a line break at the end of the target character string, and rceJ indicates to center the target character string.

Also, after outputting the target document data with rds:IJ,
It is indicated that a line feed for one line is to be performed, and rus:IJ indicates that a line feed for one line is to be performed before outputting the target document data. Furthermore, "iS:1" indicates that a margin (blank space) of one character is provided from the left side of the frame.

In the case of a paragraph, the first character of the first line is blank.

Under the control of the layout information detecting section 15, such layout rules are read out from the layout rule dictionary 9a according to their civilian type and provided to the layout information analyzing section 16.

The layout information analysis section 18 controls the document data allocation section 13 based on the layout rules obtained in this manner, and sequentially develops the document data, which is cut out character by character as described above, into designated frames.

That is, the layout information analysis section 16 determines the frame in which the document data is to be expanded according to the information of the frame, and at the same time determines the size of the frame. Then, the document data is expanded and controlled in the designated frame according to the layout rules according to the civilian character.

At this time, each time the document data is cut out character by character by the character cutting section 12, the count value of the column/line counter 21 is counted up in synchronization with this. The count value by the column/line counter 21 is successively compared with the value set in the maximum column/line value buffer 20 in a comparison section 22, and the comparison result is determined in a determination section 23. The determination section 23 determines whether the layout position of the character cut out by the single character cutout section 12 is within the frame designated at that time.

When it is determined that the allocated position of the document data to be developed under layout control exceeds the specified frame, the frame switching section 24 is activated, and the document data allocated position by the document data allocation section 13 is is updated in the next frame.

In this way, the document development section 9 sets the frame in which the document data is to be output, according to the layout rule set in the layout rule dictionary 1f9a. Then, the document data is developed in the set frame according to the layout rule set according to the sentence attribute.

For example, in the document data with the logical structure obtained as shown in FIG. In 1G, the maximum column φ line number of the set frame is calculated from the maximum column/line value buffer 20, and the character position to be centered and displayed is calculated, for example, as (MAX line number of characters in the target sentence) + 2. With this data, the column/line counter 21
The count value is changed in a preset manner, and the document data of the target sentence is developed from the changed position, that is, the character position is assigned.

Thereafter, when the output assignment of the character data of the target sentence is completed, a line break is forcibly performed according to the information [ds;LJ, and a blank line of one line-sentence is set below the character string indicating the output title.

Such output (expansion) control of document data is performed sequentially for each extracted sentence as described above, according to the logical structure determined for that sentence, and the document data is sequentially expanded into predetermined frames. go. As a result, for example, the 12th
As illustrated in the figure, document data for which a hierarchical logical structure has been determined as shown in FIG. 8 is output and expanded according to the layout rules specified for the document data.

According to the device configured in this manner, the hierarchical logic (1) structure of the document is determined from the human document data.
According to the layout rules for document data according to the hierarchical logical structure, the document data is developed in a layout format determined for each set frame according to the sentence attributes.

Specifically, document data is sequentially allocated to predetermined frames, and controls such as forced line breaks and centering are performed according to the document structure information of each sentence that follows the hierarchical logical structure of the sentence, that is, the sentence attributes. The document data is expanded.

Therefore, document processing such as when creating a document or editing a document can be efficiently executed by focusing only on the content of the document without considering its layout format.

Moreover, even when a document is processed in this way, the output layout format of the document data is automatically determined by the hierarchical document structure analyzed as described above and the layout rules specified by that document structure (sentence attributes). Since the document is controlled, it is possible to create the document in an overall unified document format.

Therefore, the control of the output format of the document does not interfere with thinking about document creation, etc., and the document processing can be performed easily and effectively, which has great practical effects. .

Note that the present invention is not limited to the embodiments described above. For example, the data format that indicates the logical structure of a document, the rules for determining the logical structure, and the layout rules that determine the output format of document data may be determined according to the specifications. Further, the extraction of headings taking into account the deformation of the font (character font), the process of determining the document structure, and further the output control may be performed. In this case, it is of course possible to simultaneously control high-intensity display, reverse display, etc. of output characters, and it is sufficient to set such control information as a layout rule.

Furthermore, in the above example, the output layout format was controlled after determining the hierarchical logical structure of the document data.
The layout development process may be performed for each sentence. In this case, the logical structure may be used directly for layout development processing without being stored once. Furthermore, it is also possible to prepare a plurality of layout rules for one civilian type and selectively apply the layout rules from among them. This allows the same text to be converted into layout formats for different purposes, such as an academic paper format and an in-house report format, as appropriate, making it possible to easily perform various layout edits. Furthermore, although the explanation has been given using Japanese as an example, it goes without saying that it can be similarly applied to other languages as well. In addition, the present invention can be implemented with various modifications without departing from the gist thereof.

[Effects of the Invention] As explained below, according to the present invention, the logical structure of chapters, sections, bullet points, paragraphs, etc. of a document is obtained from input document data, and the document is created according to layout rules corresponding to the logical structure. Since data can be expanded and output in a predetermined output layout format, the thinking process is not hindered by controlling the document output format, and the document processing can proceed effectively. Great effects can be achieved.

[Brief explanation of drawings]

The figures show one embodiment of the present invention, in which Fig. 1 is a schematic configuration diagram of the embodiment device, Fig. 2 is a diagram showing an example of input document data, and Fig. 3 is document structure analysis processing in the embodiment device. Figure 4 is a diagram showing an example of a headword dictionary; Figures 5 (a) to (d) are diagrams showing an example of the structure of a header extraction rule dictionary; Figures 6 (a) to (d) is a diagram showing an example of the structure of a document structure rule dictionary; FIG. 7 is a diagram showing the analysis structure of a heading sentence; FIG. 8 is a diagram showing an example of document structure information stored in the logical structure storage unit; FIG. 9 is a diagram showing an example of the configuration of the document expansion section, FIG. 10 is a diagram showing an example of the configuration of the layout rule dictionary,
Figure 11 shows the frame (
FIG. 12 is a diagram showing an example of a document format output after layout development processing. ■...Document management section, 2...Input section, 3...Original text storage section, 4...Display control section, 5...Display section, 6...
Headline extraction unit, 6a... Headword dictionary, 7... Headline determination unit, 7a... Headline rule dictionary, 8... Document structure determination unit, 8a... Document structure rule unit, 9... Document development section, 9a... Layout rule dictionary, 10... Logical structure storage section. Applicant's agent Patent attorney Takehiko Suzue Figure 2 (a) Figure 6 (d) Figure 6 Figure 7 ■ 8 Figure 10 Figure 11 Figure 12

Claims (4)

[Claims] (1) A means for dividing input document data into sentences according to a predetermined delimiter code, determining whether each sentence is a heading sentence, and dividing each sentence according to document structure rules according to the result of this determination. means for determining the hierarchical logical structure of the document data by determining each document structure; means for storing information on the hierarchical logical structure in association with the document data; and layout rules for determining the output layout format of the document data. The present invention is characterized by comprising: a layout rule dictionary storing data corresponding to each of the document structures; and means for developing document data of the sentence in an output unit according to a layout rule corresponding to the document structure determined for each of the sentences. document processing device. (2) The document processing apparatus according to claim 1, wherein the layout rule dictionary stores rules for determining an output layout format for a hierarchical structure of a document for each document field to be processed. (3) A layout rule property stores a plurality of layout rules corresponding to one document structure, and a layout rule is selectively used from among these, as described in claim 1. document processing equipment. (4) The document processing device according to claim 1, wherein the hierarchical logical structure of the document is composed of document logical structures such as chapters, sections, bullet points, and paragraphs.