JPS6243769A - Kana-to-kanji converting device - Google Patents

Abstract

PURPOSE:To select the desired KANJI (Chinese character) data at a high speed by using a means which extracts the KANJI data corresponding to the reading data coincident with the input reading of KANJI out of a phrase block and selecting the data as a candidate of conversion. CONSTITUTION:A reading character string extracting part 23 extracts a character string less than the maximum reading length 6 in a phrase dictionary 10 at the head of a reading input buffer 22. The start address of a reading table 12 corresponding to the reading head character and the reading length is obtained out of a reading index table 11. In the same way, the start address of the corresponding phrase block 14 is obtained out of a phrase index table 13 and stored. The reading data code is extracted out of the corresponding address of the table 12. Then a group of phrase candidates are read out of the block 14 when the coincidence is obtained between the extracted reading and the reading of the table 12. Thus the replacement is possible with the increment value fixed in relation to the reading length. As a result, a subject of conversion is retrieved at a high speed.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a kana-to-kanji conversion device, and particularly to a kana-to-kanji conversion device used in, for example, a Japanese word processor that performs bunsetsu conversion.

[Conventional technology]

In Japanese word processors and the like, when converting input pronunciations into kanji, an idiom dictionary or the like is used in which pronunciation data and each kanji data are registered in correspondence with each other in advance. The reading data registered in the dictionary is usually ``Aiueo・
...'', and in order to speed up dictionary searches, a reading index indicating a start address in the dictionary, such as the first character of two readings, is generally provided.

FIG. 7 shows an example of the structure of a conventional phrase dictionary.

The conventional pronunciation index is 1. For example, as shown in Figure 7 (a), the start address of the pronunciation starting with "a", the start address of the pronunciation starting with "i", etc. Pointers to reading and kanji data are registered using characters or character strings including the first character as items.

Further, the contents of the dictionary are as shown in FIG. 7 (b), for example. Following the reading, the number of kanji and compound words corresponding to the reading are arranged in the same number of homonyms, and the end code indicates the end. This is followed by the primary reading data.

[Problem that the invention seeks to solve]

According to the conventional method shown in Fig. 7, the reading length in the dictionary is variable, so when searching for reading data that is -t for the input reading, the end code is found. It is necessary to find the secondary reading. That is, the end
You need to check the code byte by byte and follow the readings.

There is a problem that searching for readings takes time.

In order to avoid having to check the end code, it is possible to provide a pointer to the primary reading for each reading data in the dictionary, but this would increase the storage capacity of the dictionary,
Another problem is that the process of registering new pronunciations in the dictionary becomes complicated.

[Means for solving problems]

If the reading length is fixed and readings of the same length are arranged consecutively, there is no need to search for the end code etc. when searching from one reading to the next reading. It is thought that it is possible to sequentially search for readings at high speed by adding addresses such as It should be kept separately and collectively.

Moreover, the reading data is grouped by reading length. Then, the reading index items are classified by character or character string including the first character and reading length, and for each classified item, the reading index table is used to
It points to the corresponding reading data.

When searching for a pronunciation that matches the input pronunciation and obtaining conversion candidate kanji data, refer to the pronunciation index table based on the character or string containing the first character of the input pronunciation and the pronunciation length. Obtain the start address of read data grouped by reading length. Then, while updating the address according to the reading length, the reading data is extracted and the readings are compared, so that desired kanji data can be selected at high speed. below.

An embodiment will be described with reference to the drawings.

〔Example〕

Figure 1 shows the configuration of an embodiment of the present invention, Figure 2 shows an example of the configuration of a reading index, Figure 3 shows an example of the configuration of an idiom index, Figure 4 shows an example of the configuration of a reading table, and Figure 5 shows the configuration of an idiom block. For example, FIG. 6 shows a processing explanatory diagram of an embodiment of the present invention.

In Fig. 1, 10 is an idiom dictionary, 11 is a reading index table, 12 is a yomi table in which reading data is registered, 13 is an idiom index table, 14 is an idiom block in which kanji data is registered, 20 is an input of readings, etc. 21 is an input control unit, 22 is a reading cover sofa, 23 is a reading character string extraction unit that extracts a reading character string from the data of the reading cover sofa 22, and 24 is a reading length control that instructs the reading length to be converted. 25 is a pronunciation index table search unit that searches the pronunciation index table 11; 26 is an idiom index table search unit that searches the idiom index table 13; 27 is a pronunciation index table search unit that searches the pronunciation index table 11; 27 is a pronunciation index table search unit that searches the pronunciation index table 11; Selling tape flare address management section, 28 is an idiom block address management section that manages pointers to the idiom block 14, 29 is a reading table search section that extracts reading data from the reading table 12, and 30 is an input L7 reading and 7 reading table. 12 is a pronunciation comparison section that compares and matches the pronunciations in idiom block 14.
g with evening as an idiom candidate! 1 word candidate selection section; 32, phrase determination section; 33, display control section for controlling the display; and 34, a display.

The detailed structure of the reading index table 11 is as follows.

It is as shown in Figure 2. In the reading index table 11, not only start addresses for readings whose first character is "A", for example, but also reading lengths are registered. That is, in this embodiment, the maximum value of the reading length is 6 (the number of characters in the reading is 6 digits), the first character is "A", and the start address of the reading data with the reading length of 6 is "A". So, the start address has a reading length of 5.

..., a start address whose first character is "A" and whose reading length is 1, and so on. Start addresses classified by reading length are set. Note that the maximum value of the reading length does not necessarily have to be 6, and the key information of the index may be two or more characters, such as 1, for example, 1 ah J, "Ai", etc., instead of the first character.

Similarly, the detailed structure of the phrase index table 13 is as follows:
As shown in FIG. 3, this also corresponds to the reading index table 11 and is classified in consideration of the reading length. Note that the reading index table 11 and the idiom index table 13 may be combined into one table.

The reading table 12 is configured, for example, as shown in FIG. In other words, pairs of readings with a reading length of 6 and their idiom block lengths are arranged as many times as there are reading data with a reading length of 6, and then pairs of readings with a reading length of 5 and their idiom block lengths are arranged as many times as there are reading data with a reading length of 6. are made to continue in sequence. Note that since the first character of the reading is determined by searching from the reading index table 11, it is omitted from the reading table 12 and registered. Idiom block length is Idiom block 1
This is the length of the kanji data group of homophones in 4.

Since the reading length is known in advance, searching for readings in the reading table 12 can be performed by adding a constant to the address.

The phrase block 14 is configured as shown in FIG.

That is, ? 'J, a couminator indicating the end of two kanji data are registered in the reading order in the reading table 12.

Next, the processing operation of the second embodiment will be explained with reference to FIG.

The following numbers - [phase] correspond to the process numbers shown in FIG.

(1) The illustrated reading character string extracting unit 23 extracts the character string with the maximum reading length (6) or less in this embodiment from the beginning of the reading cover sofa 22, and sets it as the first reading. By sequentially searching from the longest character string, the longest −
The losing candidate will be selected first.

(2) From the reading index table 11, find the start address of the reading table 12 corresponding to the first character of the reading and the 5 reading lengths.

■Similarly, from the phrase index table 13.

The start address of the idiom block 14 corresponding to the first character of the pronunciation and the five pronunciation lengths is determined and stored.

■ From the corresponding address in the reading table 12.

Extract the reading data code corresponding to the reading length.

(2) Determine whether or not the reading extracted in the above process (2) matches the reading in the reading table 12. If they do not match, control is transferred to processing [phase].

■ When the readings match, from Ikugo Pro, ri 14.

A group of phrase candidates corresponding to the corresponding phrase block length in the reading table 12 is read out.

■ Display the read phrases on the display 34 one after another.

■ Determine whether the idiom is correct based on the input from the keyboard 20. If the idiom is confirmed, 1
After the two conversion processes are completed, the data is passed to the document editing section (not shown). Or continue the primary conversion process or reading input.

■ If the phrase has not been determined, it is determined whether all phrases corresponding to the reading have been displayed. If not completed, control is returned to process ■.

Show next suggestion. When all the phrases have been displayed, it means that there are no phrases corresponding to the current reading length, so control is transferred to the process [phase].

[Phase] When the pronunciations do not match in the judgment of process ①, add the length of the idiom block in the reading table 12 to the pointer that saves the address of the idiom block 14 and add 2.
The block address of the next idiom block 14 is determined in advance.

o Add the increment value corresponding to the reading length to the pointer that saves the address of the reading table 12 to find the next data address in the 8-selling table 12.

■ Check whether the data for the relevant principal has been completed.

■ If the data has not been completed, control is returned to process (■), and reading data is similarly extracted and reading comparison processing continues. When the data is completed, the following is executed as the primary process.

(1) The reading length control unit 24 shown in FIG. 1 reduces the reading length by one character.

0 Determine whether the reading length has become “0”.

If the reading length is not “0”, control is returned to processing ■, ■,
Repeat the process in the same way.

[Phase] When the reading length is "0", control is transferred to a processing unit (not shown) for when there is no corresponding kanji, and the conversion process is ended.

〔Effect of the invention〕

As explained above, according to the present invention, the idiom block in which kanji data is stored and the reading table in which reading data is stored are separated, and the reading index items are classified by reading length.

The address to search for reading data can now be updated in fixed increments relative to the reading length.

You will be able to quickly search for conversion targets. Also.

The capacity of the dictionary is also relatively small.

[Brief explanation of the drawing]

Figure 1 shows the configuration of an embodiment of the present invention, Figure 2 shows an example of the configuration of a reading index, Figure 3 shows an example of the configuration of an idiom index, Figure 4 shows an example of the configuration of a reading table, and Figure 5 shows the configuration of an idiom block. For example, FIG. 6 is a processing explanatory diagram of one embodiment of the present invention, and FIG. 7 shows an example of a conventional phrase dictionary. In the figure, 10 is an idiom dictionary, 11 is a reading index table, 12 is a reading table, 13 is an idiom index table, 14 is an idiom block, 20 is a keyboard, 21 is a human control unit, 22 is a reading cover sofa, 23 is a reading character 24 is a reading length control unit, 25 is a reading index table search unit, and 26 is an idiom index table search unit. 27 is a reading table address management section, 28 is a phrase block address management section, 29 is a reading table search section, 30 is a reading comparison section, and 31 is a phrase candidate selection section. Reference numeral 32 represents a phrase determining unit, 33 represents a display control unit, and 34 represents a display. Patent applicant: Patent attorney representing Alps Electric Co., Ltd.
Mori 1) Hiroshi (3 others) Fig. 2 Fig. 3 Fig. 3 Il] IIL71-Jun Example E? Diagram 6

Claims (1)

[Scope of Claims] Kana that includes a dictionary having a reading index whose items are characters or character strings including at least the first character of the reading, inputs the reading, searches the dictionary, and converts the reading into the corresponding kanji. In the kanji conversion device, the reading index items are classified by character or character string including the first character and reading length, and a reading index table (11) points to the corresponding reading data for each classified item. A reading table (12
) and the idiom block (14) in which the kanji data is stored corresponding to the reading registered in the reading table (12).
), means (25) for searching the reading index table (11) based on a character or character string including the first character of the reading and the reading length; and a reading pointed from the reading index table (11). Regarding the data, a means to update the address based on the reading length and sequentially compare it with the input reading (2
7, 29, 30), and means (31) for extracting kanji data corresponding to reading data that matches the input reading from the compound word block (14) and selecting it as a conversion candidate. Kana-kanji conversion device.