JPH05128102A - Sentence compression device - Google Patents

Abstract

PURPOSE:To provide a sentence compression method which can enlarge a compression rate by encoding a long character string. CONSTITUTION:This device is provided with dictionaries 14 and 15 registering one of the similar character strings from a sentence including plural equal or similar character strings formed from plural words, a control part 11 which is connected with the dictionaries 14 and 15 and replaces the similar character strings in the sentence into identification data and difference data of one character string registered in the dictionary 14 and 15, a data base 16, buffers 17 and 18 and a memory 19.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a word processor and an electronic book sentence compression apparatus.

[0002]

2. Description of the Related Art A conventional text compressing apparatus registers a character string registered in a dictionary with a plurality of identical character strings in units of a word or a character string having a certain number of characters, and identifies the character string registered in the dictionary. A text compression method is used in which the same character string in a text is replaced by a code to compress the text.

FIG. 9 shows the structure of a dictionary for registering word-by-word character strings used in the above-described conventional sentence compression apparatus.

In the conventional sentence compression device using the dictionary shown in FIG. 9, for example, a sentence such as "I also use elemental magic that uses the power of the wind." And use [word 01411] [word 12959].

[Word 01411], [Word 1] in this compressed sentence
2959] and the like indicate coded words, and the original character string can be restored by searching the dictionary in FIG. 9 with the word number included in the code.

[0006]

However, in the above-described conventional sentence compression apparatus, the character string registered in the dictionary is 2 characters.
When the number of types exceeds 56, 2 bytes are required as an identification code. However, since many Chinese words are formed of two characters and the target word is only a few characters long, there is a problem that the compression rate of the entire sentence cannot be increased.

In view of the above-mentioned problems in the conventional text compression apparatus, the present invention provides a text compression apparatus which can code a character string to increase the text compression rate.

[0008]

According to the present invention, there is provided registration means for registering a specific character string among similar character strings from a sentence including a plurality of identical or similar character strings formed from a plurality of words, This is achieved by a sentence compression device that is connected to the registration means and that has conversion means for replacing similar character strings in a sentence with identification data and difference data of a specific character string registered in the registration means. ..

[0009]

In the sentence compressing apparatus of the present invention, the registering means registers a specific character string among similar character strings from a sentence including a plurality of identical or similar character strings formed from a plurality of words, and converting means. Is connected to the registration means and replaces a similar character string in the sentence with the identification data and the difference data of the specific character string registered in the registration means.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the sentence compression apparatus of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing the configuration of an embodiment of the sentence compression apparatus of the present invention.

The text compression apparatus shown in FIG. 1 is connected to the control unit 11 and the control unit 11, which is a part of conversion means for controlling the restoration of compressed text data and the display of character strings, and the text to be restored. Is connected to the input unit 12 and the control unit 11 for displaying the restored text, such as a cathode ray tube (CRT) and a liquid crystal display (LCD), and the control unit 11. One of the registration means for storing the character string to be restored from the data obtained by compressing (see FIG. 2) and coding the same character string in a long character string (hereinafter referred to as sentence code). Dictionary 14 which is a part of the dictionary, which is connected to the control unit 11 and which is a part of the registration means for storing the character string restored from the word code, and which is connected to the control unit 11 and which is a sentence code. A part of conversion means that organically combines coded and compressed data with difference data A certain database 16, a buffer 17 that is connected to the control unit 11 and is a part of the conversion means that stores the finally restored result, and a character string that is connected to the control unit 11 and that is retrieved from the dictionary 14. Is composed of a buffer 18 which is a part of a converting means for temporarily storing the data, a buffer 19 connected to the control section 11 and a memory 19 which is a part of the converting means for storing the character position of the compressed data. ing.

Next, each of the above components will be described.

The control unit 11 controls the decompression of compressed text data and the display of character strings, and also controls the operation of other parts.

The input unit 12 is constituted by a keyboard so as to instruct a sentence to be restored.

The display unit 13 is composed of a cathode ray tube (CRT), a liquid crystal display (LCD) or the like so as to display the restored text.

As shown in FIG. 3, the dictionary 14 is composed of an index 141 corresponding to sentence numbers in a one-to-one correspondence and a main body 142 for storing variable length character strings.

The index 141 indicates the beginning of the corresponding character string of the main body 142, and the length of the character string can be known from the difference from the character position pointed to by the next element of the index 141.

The dictionary 15 has word numbers and 1 as shown in FIG.
The structure is one-to-one correspondence, and the database 16 is
It is configured such that the sentences coded and compressed by the sentence code and the difference data are organically combined and collected, and a necessary portion can be taken out by an instruction of the input unit 12.

The buffer 17 stores the finally restored result, and the buffer 18 is configured to temporarily store the character string retrieved from the dictionary 14.

The memory 19 includes a pointer 191, which indicates the processing position of the text database 16, a pointer 192, which indicates the end position of the restoration buffer 17, a pointer 193, which indicates the end position of the restoration buffer 18, and an insertion position of the restoration buffer 18. Pointer to
It is composed by 194.

FIG. 2 shows a part of an original sentence compressed by the sentence compressing apparatus of this embodiment. Further, FIG. 3 shows an example of the structure of a dictionary in which long character strings are registered in the text compression device of this embodiment, and FIG. 4 is a diagram showing a long character string coded by the text compression device of this embodiment. An example is shown below.

The sentence compressing apparatus of this embodiment registers long character strings such as sentences, clauses or phrases in the dictionary 14 to code the same character strings (sentence code), and When the character string registered in the dictionary 14 does not completely match, the difference data is added. That is, the text compression apparatus of the present embodiment uses the text compression method for compressing the original text as shown in FIG. 2 into the shape shown in FIG.

Next, each part of the sentence of FIG. 4 compressed by the sentence compressing device of FIG. 1 will be described in detail.

[Sentence] in FIG. 4 is a separator that separates the sentence code and the difference data from the normal character code. Also, [00023] of FIG. 4 shows a sentence code,
The original character string is restored by searching the dictionary of FIG. 3 with the sentence number included in the code.

[01-04] in FIG. 4 is deletion data from the character string restored from the sentence code. The number placed after the hyphen indicates the beginning position of the deletion character, and the hyphen. The number placed in front of indicates the number of deleted characters.
The deletion data and the insertion character string following the deletion character data constitute difference data.

FIG. 5 is a schematic diagram showing an example in which both a long character string and a word-unit character string are coded in the sentence compressing apparatus of this embodiment, and FIG. 6 is of this embodiment. It is a figure for demonstrating the code data structure in a text compression device.

The content shown in FIG. 5 is the latter half of FIG. 4 further compressed using the word dictionary of FIG. As an actual code, characters are usually expressed in 2 bytes (most significant byte (MSB) “0”) using JIS Kanji code, and other compression codes are shown in FIG. The configuration is as shown. To accommodate this structure, if the difference is 256 characters or more from the beginning, the sentence to be registered is divided and adjusted within 255 characters, and if the difference is 8 characters or more, adjusted within 7 characters. To do.

Next, referring to the flow charts of FIGS. 7 and 8, the operation of the text compression apparatus of FIG. 1, particularly the procedure for decompressing the compressed document designated by the input unit 12 (see FIG. 1). Will be explained.

Taking the first line of the sentence in FIG. 5, first,
Pointer 191 points to the beginning of the indicated compressed data, and pointer 192 points to the beginning of buffer 17 and pointer 193.
Are initialized so that they respectively point to the beginning of the buffer 18 (step S1), the type of data pointed to by the pointer 191 is determined (step S2), and in the case of FIG. 5, the separator [sentence] comes. Proceed to step S3. From here, enter the sub-routine of FIG. 8 for restoring the sentence code data.

Since there is a sentence code immediately after the sentence separator, the sentence number [00023] is taken out (step S3).
1), dictionary 42 with index 41 with this sentence number
To find the character string “against the wind ...” (step S32), copy the found character string to the beginning of the buffer 18 so that the end position is pointed to by the pointer 193 (step S33). The pointer 191 is advanced to the next (step S34), and the type of data pointed to by the pointer 191 is determined (step S35). In this case, since the next data is the deleted character data [01-04], the process proceeds to step S36.

The character position and the character length to be deleted are extracted from the buffer 18, and the character position is stored in the pointer 194 (step S36). In this case, only the first character of the fourth character (the character to be deleted) (in this case,
"Wind")) is deleted from the buffer 18, and the pointer 193 is operated so as to point to the end of the character string (step S37). Then, the process returns to step S34 and the next data is determined in step S35.
Since the normal character is "he", the process proceeds to step S39.

This character is inserted into the position indicated by the pointer 194 of the buffer 18, and the pointers 193 and 194 are advanced by the number of inserted characters to point to the end position and the next insertion position, respectively (step S39), step S34, step Loop S35 and step S39,
After inserting "ha" in the same manner, since the next data is the word code [word 08394] in the determination of step S35, the process proceeds to step S38.

The dictionary 15 is searched from the word number of the word code and the corresponding character string, in this case, "earth" is taken out (step S38). The "ground" taken out in step S39 is inserted into the buffer 18 and the process returns to step S34.

At step S35, since it is found that the next data is the sentence separator, the process returns to the main routine of FIG.

Character string (character string prepared for buffer 18)
Is added to the position pointed to by the pointer 192 of buffer 17.

The pointer 193 is made to point to the beginning of the buffer 18 (step S5), the pointer 191 is advanced (step S6), and the end of data is determined (step S5).
7). In this case, since the data is finished here, the process proceeds to step S8, the character string stored in the buffer 17 is displayed on the display unit 13 and the process is finished (step S8).

Next, the third line of the sentence shown in FIG. 5 will be described.

First, after the initialization of step S1 is performed as in the case of the first line, the first character is the normal character "ta" in the determination of step S2, so the process proceeds to step S5 and this character is stored in the buffer 17. Input (step S5), pointer 191
(Step S6), the end of data is determined (Step S7). In this case, since there is still data, the process returns to step S2. Then, step S2, step S5, step S6 and step S7 are looped,
After "da", "shi", and "," are added to the buffer 17 in the same manner, the next data is the word code [word 00808] according to the determination in step S2, and therefore the process proceeds to step S4.

The dictionary 15 is searched from the word number of the word code and the corresponding character string, in this case "stone crushed", is taken out (step S4), and this character string is buffered in step S5.
Add to. After this, add a normal character string with a similar loop, and finally add the sentence code [000 as in the case of the first line sentence.
25] and restore it to buffer 17.

In step S7, the end of data is determined,
The result restored in step S8 is displayed and the process ends.

As described above, sentences, clauses registered in the dictionary,
The original character string is restored from the sentence data obtained by coding a character string that is the same as or similar to a long character string such as a phrase and compressed.

[0043]

EFFECT OF THE INVENTION The sentence compression apparatus of the present invention includes a registration means for registering a specific character string among similar character strings from a sentence including a plurality of identical or similar character strings formed from a plurality of words, Since it is connected to the registration means and has a conversion means for replacing the similar character string in the sentence with the identification data and the difference data of the specific character string registered in the registration means, there are many similar expressions. The compression ratio of the appearing sentence can be improved.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of an embodiment of a text compression device of the present invention.

FIG. 2 is an explanatory diagram showing a part of an original sentence for explaining an example of compression by the sentence compression device in FIG.

FIG. 3 is an explanatory diagram showing a configuration example of a dictionary for registering a long character string of the sentence compression device in FIG.

FIG. 4 is an explanatory diagram showing an example sentence in which a long character string is coded by the sentence compression device of FIG.

5 is a schematic diagram showing a structure of an example in which both a long character string and a character string of a word unit are coded in the sentence compression apparatus of FIG.

FIG. 6 is an explanatory diagram showing the structure of code data in the sentence compression apparatus of FIG.

FIG. 7 is a flowchart of a main routine for explaining the operation of the sentence compression apparatus of FIG.

FIG. 8 is a sub-routine flowchart for explaining the operation of the sentence compression apparatus of FIG.

FIG. 9 is an explanatory diagram showing a structure of a dictionary for registering character strings in word units of a conventional sentence compression device.

[Explanation of symbols]

 11 Control unit 12 Input unit 13 Display unit 14 Character string dictionary 15 Word dictionary 16 Compressed sentence database 17 Restored sentence buffer 18 Character string dictionary Buffer 19 Memory set 141 Character string dictionary index 142 Character string dictionary body 191 to 194 Memory Pointer

Claims (1)

[Claims] 1. A registration unit for registering a specific character string of a similar character string from a sentence including a plurality of identical or similar character strings formed of a plurality of words, and a registration unit connected to the registration unit. A sentence compression apparatus, comprising: a conversion unit that replaces the similar character string in the sentence with the identification data and difference data of the specific character string registered in the registration unit.