JPH0360440B2 - - Google Patents

Description

[Detailed Description of the Invention] (1) Field of Application of the Invention The present invention provides a Japanese word processor for
The present invention relates to an audio read-aloud function that enables the correct reading of kanji or idioms, even when reading kanji or idioms.

(2) Prior Art Conventionally, when a kanji character appears in a sentence, a unique reading is given to the kanji character beforehand and the kanji character is uttered aloud. According to this method, even though kanji have different readings depending on the case, the readings were always limited to one reading, and there was a deviation from the correct reading.

(3) Purpose of the Invention The purpose of the present invention is to realize a voice reading function that allows kanji parts to be uttered in a natural manner, for checking the input of Japanese sentences containing kanji and kana.

(4) General explanation of the invention In order to check whether the contents of the sentences that have already been input and created are correct, the sentences are read out one after another by voice. Hiragana, katakana, alphanumeric characters, symbols, etc. can only be read in one way, so if you give them the reading in advance, you can read them correctly. On the other hand, kanji do not necessarily have one reading; some have as many as 11 different readings, such as the ``lower'' reading. In the case of reading aloud, it is not automatically determined which reading should be given, but the following trends can be observed.

1. When a kanji exists in isolation and is preceded and followed by kana, that kanji is often read in kun-yomi.

2 Even if a kanji exists in isolation, when it is followed by an alphabetic character or symbol other than kana, it is often read on-yomi.

3. When two or more kanji appear in a row, they are often read on-yomi.

The present invention is characterized in that already created sentences are read together according to the above-described rules.

(5) Embodiment An embodiment of the present invention will be described above with reference to FIG.

In FIG. 1, a text buffer 2 stores previously created Japanese texts such as sentences containing kanji and kana characters as character code strings.
A CPU (central processing unit) 4 sequentially reads character codes from the text buffer 2 and writes the codes to a refresh memory within the CRT display 1. The CRT display 1 scans the refresh memory at regular intervals and displays the character pattern in the character generator 5 that corresponds to the character code. Also, the above
A cursor indicating the position of a character is displayed on the CRT display 1, and information on the cursor position is monitored by a cursor position register 3. The cursor can also be moved by operating cursor keys on the keyboard 22.

The above-mentioned operations have already been performed before implementing the contents of the present invention, and the start of implementing the contents of the present invention is realized by pressing, for example, the voice reading key on the keyboard 22.

When a command to start the embodiment of the present invention is sent from the keyboard 22 to the CPU 4, the CPU 4 first sets the contents of the flag register 6 to 0. next,
The CPU 4 sends the code of the character corresponding to the cursor position on the CRT display 1 to the text buffer 2.
, and stores it in the character buffer 7 at the cursor position. Further, the CPU 4 obtains the code of the character next to the cursor position on the CRT display 1 (shifted one character backward) from the text buffer 2, and stores it in the next position character buffer 8.

The CPU 4 checks the character code in the character buffer 7 at the cursor position, and performs the following operations depending on the type of character code. That is, (1) If the character code is an alphabet other than kanji or kana, a number, a symbol, etc., the flag register 6
The content of is set to 0, the number of characters is 1, the character length is 1, and the above character code are stored in the code buffer 13, and at the same time, the cursor position information in the cursor position register 3 is advanced by one.

(2) If the character code in the cursor position character buffer 7 is either Hiragana or Katakana, the CPU 4 refers to the character code stored in the next position character buffer 8, and if the character code is If it represents a lowercase kana character,
Set the contents of flag register 6 to 0, store the number of characters 1, character length 2, character code in cursor position character buffer 7, and character code in next position character buffer 8 to code buffer 13, and at the same time store the cursor position register Advances the information at the cursor position in 3 by two. Also, if the character code stored in the next position character buffer 8 does not represent a lowercase kana character, then
The CPU 4 sets the contents of the flag register 6 to 0, stores the number of characters 1, character length 1, and the character code in the cursor position character buffer 7 in the code buffer 13, and at the same time stores the cursor position in the cursor position register 3. Advance the information by one.

(3) When the character code in character buffer 7 at the cursor position is a kanji.

The CPU 4 stores the character buffer 7 at the cursor position.
The character code inside is sent to a reading generation circuit 9, which will be described later. The reading generation circuit 9 receives the contents of the flag register 6 and the character code in the position character buffer 8 from the CPU 4, determines the reading for the character (kanji) in the sent cursor position character buffer 7, and calculates the number of characters. , character length,
The reading code is sent to the reading buffer 12, and at the same time, the cursor position information in the cursor position register 3 is advanced by one, and a signal pulse indicating the end of the operation is sent to the CPU 4.

As described above, when the CPU 4 sends a signal to the code buffer 13 (operation (1) or (2)) or receives an operation end pulse from the reading generation circuit 9,
The CPU 4 sends a command to the voice address conversion circuit 14. The voice address conversion circuit 14 receives the information of the flag register 6 from the CPU 4, and if the content is 0, it calculates the number of characters from the code buffer 13, and if the content is 1, it calculates the number of characters from the reading buffer 12. , receives the character length and character code, and creates a voice address table 15 corresponding to the character code of the reading as many as the number of characters.
The number of characters and the address are stored in the address buffer 16, and a pulse indicating the end of the operation is sent to the CPU 4.

The CPU 4 sequentially transfers the address information in the address buffer 16 to the audio LSI interface 1.
7 to the speech synthesis LSI 18. speech synthesis
The LSI 18 extracts the audio parameters from the audio parameter ROM 19 according to the address information and sends them to the audio synthesis circuit in the audio LSI 18.
A digital signal of the synthesized voice is obtained. The voice synthesis LSI sends the digital signal to the DA converter 20 to convert it into an analog signal, and when the signal is passed through the speaker 21, a voice is produced. Above audio parameters
Each utterance unit in the ROM 19 (in this example, kana 1)
Data indicating the end of the utterance is added to the end of the utterance of a character, alphabet, number, signal, and when the speech synthesis LSI 18 detects this,
The speech synthesis LSI end bit is set.

Each time the CPU 4 senses the bit through the audio LSI interface 17, it sends an address for one character to the audio LSI interface 17. When the address for the predetermined number of characters has been sent, the CPU 4
The code of the character indicated by the cursor on the CRT display 1 is extracted from the text buffer 2 and stored in the cursor position character buffer 7, and the code of the character one character after the cursor position is stored in the next position character buffer 8. Then repeat the operations described above.

The CPU constantly monitors input from the keyboard 22, and when a signal to stop reading arrives from the keyboard 22, it stops executing the operations described above.

By the above-described operation, the Japanese text containing Kanji and Kana characters in the text buffer 2 can be read out in a more natural way.

FIG. 2 is a block diagram showing the contents of the reading generation circuit 9 shown in FIG. In FIG. 2, the CPU 4 sends a kanji code to a kanji code buffer 901 and then to a subtraction circuit 902. When the subtraction circuit 902 subtracts a predetermined value from the sent kanji code, the kanji code is ,
This becomes a number representing the order of the kanji, and this is sent to the buffer 903. The CPU 4 refers to the contents of the flag register 6, and if the contents are 1,
When a command to read aloud is sent to the memory search circuit 904, the memory search circuit 904
The number of characters in the pronunciation, the character length in the pronunciation, and the code in the pronunciation in the on-yomi memory 10 corresponding to the number in the above are taken out and stored in the pronunciation buffer 12. Also, if the contents of the flag register 6 above are 0, the CPU
4 further refers to the contents of the next position character buffer 8, and if the contents are kanji, sends an on-yomi instruction to the memory search circuit 904, and the memory search circuit 904 reads the number in the buffer 903. The number of characters in the pronunciation, the character length in the pronunciation, and the code in the pronunciation in the on-yomi memory 10 corresponding to the above are taken out and stored in the pronunciation buffer 12. Also, if the content of the next position character buffer 8 is not a kanji, the CPU 4
When the memory search circuit 904 sends a command for reading, the memory search circuit 904 sends a command to the buffer 903.
The number of characters in the pronunciation, the character length in the pronunciation, and the code in the pronunciation in the kun-yomi memory 11 corresponding to the number in the above are taken out and stored in the pronunciation buffer 12.

FIG. 3 is a flowchart representing the processing of the CPU 4 in FIG. 1.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing one embodiment of the present invention. FIG. 2 is a block diagram illustrating the contents of the reading generation circuit 9 in FIG. 1. FIG. 3 is a flowchart explaining the operation of the CPU in FIG. 1.

Claims (1)

[Claims] 1.Equipped with a device that converts a stored character string into a series of single sounds, and a device that utters the single sound as a sound, and the generating device generates a sound when at least two Kanji codes appear consecutively in the character string. For example, if the onyomi reading of the kanji is given, and the kanji code exists isolated in the character string, and the next character of the kanji is a kana, then the kunyomi stem of the kanji is given. A voice reading device characterized in that it utters a voice as a voice.