JPH03212757A - Romaji/kana conversion system - Google Patents

Abstract

PURPOSE:To correctly recognize a left character string and to attain correct 'ROMAJI' (Roman letter)/'KANA' (Japanese syllabary) conversion by applying the 'ROMAJI'/'KANA' conversion to a character in a character string buffer corresponding to pointers belonging to a character string buffer and an attribute buffer. CONSTITUTION:If a cursor is moved and 'h' is inputted when 'S' which is not processed by 'ROMAJI'/'KANA' conversion exists on the way of the character string buffer, the conversion processing is executed and a 'ROMAJI' pointer (RP) points out the 'S'. Then the 'h' is inputted to the position of a cursor pointer CP and the 'ROMAJI'/'KANA' conversion is applied to a character string 'sh' esixting between both the pointers RP, CP. The 'ROMAJI'/'KANA' conversion is not applied to 'sh', but when 'U' is inputted to the position of the RP, 'shu' is used for the 'ROMAJI'/'KANA' conversion to obtain a 'KANA' character string 'shu'.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a Roman-ji-kana conversion method, and more particularly to a Roman-ji-kana conversion method for devices for inputting Japanese.

[Conventional technology]

Romaji-kana conversion is a technique used when entering characters on a computer keyboard or in a word processor. This technology is widely used as an effective means for inputting Japanese sounds. In romaji-kana conversion, data representing the pronunciation of the character you want to input is input as romaji using a combination of alphabet keys on the keyboard, and then the ``gana'' corresponding to the romaji is input.

In addition to this, Romaji-Kana conversion often includes the ability to move the cursor during input to edit the character string being input. That is, the user moves the cursor and deletes the character strings before and after the scroll, or performs further input at the cursor position.

[Problem to be solved by the invention]

In the conventional romaji-kana conversion described above, there are cases where the romaji are not yet converted to kana while inputting the romaji.
(For example, if you move the cursor while inputting rshj), the Roman alphabet you were inputting will remain as is, and even if you return to that part and continue typing, the Roman alphabet fragment you were inputting previously will be used. The disadvantage is that it cannot be used.

For example, if you move the cursor after inputting rshJ, and then move the cursor back immediately after that fit r sh J and input riJ, it will not be converted to rLl, but rsh.''
It had the disadvantage that it ended up looking like this.

[Means to solve the problem]

The Romaji-Kana conversion method of the present invention, in the Romaji-Kana conversion method for equipment that inputs Japanese using an alphabet display keyboard, has a character string buffer that holds a character string undergoing Romaji-Kana conversion, and a It has an attribute buffer that holds the attributes corresponding to the held characters, and performs Romaji-kana conversion for the characters in the string buffer that correspond to pointers attached to the string buffer and the attribute buffer, respectively. It is configured by doing the following.

〔Example〕

Next, the present invention will be explained with reference to the drawings.

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

When a user inputs Japanese using Romaji-Kana conversion, the user inputs using the keyboard 101, and the input is passed through the input character editing control unit 103 to the Romaji-Kana conversion control unit 104, where it is converted into Romaji-Kana. The progress and results of the process are displayed on the display device 102.

FIG. 2 is an explanatory diagram showing an example of the storage control that the Romaji-Kana conversion control unit 104 according to the present invention has internally for Romaji-Kana conversion.

The storage area is mainly composed of two similarly indexed buffers. The buffers include a string buffer 201 for holding a string that is being converted to Roman to kana, and an attribute buffer 20 for holding the attributes of each character stored in the string buffer 201.
Consisting of 2.

The string buffer stores the string that is currently being converted to Romaji-Kana, but the attribute buffer stores information about each character in the string buffer to determine whether that character was input after Romaji-Kana conversion. , a flag is set to indicate whether the alphabet is used for romaji-kana conversion.0For example, in Figure 2, among the characters in the string buffer, "ni", "yu", and "u" are in romaji or kana. "1" indicating that conversion is being performed is stored in the attribute buffer, and "s" among the characters in the character buffer is stored in the attribute buffer.
", "h", "0" indicating that Romaji-kana conversion has not yet been performed is stored in the attribute buffer.

Furthermore, for the buffer, two pointer memories are used to control the cursor position 1 < cursor pointer (C
P') 203> and control of the character string to be converted from Roman to Kana (Romaji pointer <RP) 204 in FIG. 2).

The character string in the character string buffer existing between the Romaji pointer (RP) 204 and the cursor pointer CP2O3 in FIG. 2 is the character string to be converted into Romaji-Kana.

The values of the cursor pointer and Romaji pointer are changed by the editing function when Romaji-Kana conversion input is performed. Changing the cursor, which is the place to insert input during input processing, means changing the cursor pointer (CP>203
will change. At this time, the romaji pointer (
RP) 204 also moves along with the cursor pointer CP.

FIG. 3 is a flowchart showing the flow of control when one Roman character is input and Roman character-kana conversion input is performed.

When Romaji is input according to this flow, the existence of Romaji that was previously input and is in the middle of conversion is checked, and if it exists, the process of inputting Romaji-Kana conversion including that Romaji is performed. It will be done.

In step 302, it is determined whether a check is to be made to see if a character that has not been converted to Romaji-Kana exists immediately before the cursor.

Cursor pointer (CP) and Roman pointer (RP)
Not being equal is a sufficient condition that the cursor has not been moved, in which case the check to see if there is an unconverted string in front of the cursor is omitted. .

Here, if the values of the cursor pointer (cp) and the Romaji pointer (RP) are equal, a search for a character string that has not been converted to Romaji-kana is performed in the following procedure.

First, the value of the Roman pointer (RP) is decremented. Next, the decremented Roman pointer (R
It is checked whether the character attribute indicated by P) is "1" indicating that the conversion has been completed. Decrementing the Roman character pointer (RP) and checking the attributes of a character string will cause the attributes of the character pointed to by RP to be converted, but
This is performed continuously until the part pointed to by RP is before the beginning of the character string buffer (steps 303 to 30).
5). By incrementing RP only once when the decrement of RP is completed (step 306),
RP points to the beginning of the character string that exists immediately before the cursor and has not been converted to Roman characters.

After this process, almost the same process as the Romaji-kana conversion seen in the prior art is performed. In other words, insert the entered Roman character at the cursor pointer,
Converts the Roman character pointer and the cursor pointer to Roman character and Kana, and replaces the character string between the Roman character pointer and the cursor pointer with the character string obtained as a result of the Roman character to Kana conversion.

However, in subsequent processing, when performing Romaji-Kana conversion, it is necessary to re-enter the value of the attribute buffer for the replaced part so that the unconverted character string can be recognized correctly. The result of romaji-kana conversion is divided into the part that has been converted to kana and the part that has not been converted to kana, so the attribute buffer for the part that has been converted to kana and romaji will indicate that it has been converted. "1" is set, and "0" indicating that the conversion has not been completed is set for the portion remaining in Roman letters (step 310).

FIGS. 4(a) to 4(d) represent the situation in which Romaji-kana conversion is performed in the flow shown in FIG. 40
1 to 404 are a character string buffer, an attribute buffer, a cursor pointer, and a Roman character pointer, respectively. FIG. 4(a) shows a state in which input has been made up to "ga n", "ka" has been converted, and "n" has not yet been converted.

The states shown in FIG. 4(b) to (d) correspond to the state of FIG. 4(a) in which input rke i J is input one character at a time.

To describe its operation, when "k" is input, rn
kJ is converted to romaji and kana, and "nk" is obtained as a result. Of "nk", "n" has been converted to Roman Kana, but rl (J has not been converted to Romaji Kana. As a result, the state shown in Figure 4 (b) is obtained. , "n" has an attribute indicating that it has been converted, and I"k4
has an attribute indicating that the conversion has not been completed. The romaji pointer is “
Move to the "k" part.

Furthermore, by inputting "e", "rke" is converted to romaji kana and "keJ" is obtained. The result is shown in Figure 4 (
This is the state shown in C). If you input "i" here, Romaji-kana conversion will be performed and you will get "i". The results are shown in FIG. 4(d).

FIGS. 5(a) to 5(d) are other examples showing how Romaji-kana conversion is performed according to the flow shown in FIG. 3.

This shows what kind of processing is to be performed when the character "S" that has not been converted to Romaji or Kana exists in the middle of the character string buffer. That is, if the cursor is moved and "h" is input when the state shown in FIG. 5(b) is reached, the processes shown in steps 303 to 305 are performed and the Roman character pointer points to "s". After that, "rh" is input at the cursor pointer (cp) position, and the Roman pointer (RP) and cursor pointer (
The character string rsh, , 1 between CP) is the target of Romaji-Kana conversion.

Romaji-kana conversion is not performed for "Sh", and both characters "S" and rhJ remain as unconverted characters,
The position of the Roman character pointer also remains pointing to "S".

If you enter "u" again in this state, "rshu" will be converted into Romaji-kana, "shiyu" will be obtained, and the fifth
The state shown in figure (d) is reached.

〔Effect of the invention〕

As explained in detail above, the present invention adds attributes to character strings that may be subject to Romaji-Kana conversion during Romaji-Kana conversion, thereby correctly recognizing character strings that remain without Romaji-Kana conversion. , it has the effect of being the correct target for romaji-kana conversion.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention, FIG. 2 is an explanatory diagram showing an example of the storage control that the Romaji-kana conversion control unit according to the present invention has internally for Romaji-kana conversion, and FIG. The figure is a flowchart showing the flow of control when one Roman character is input and Romaji-kana conversion input is performed, Figures 4 (a) to (d) and Figures 5 (a) to (d).
FIG. 4 is an explanatory diagram showing how the process proceeds using the flow shown in FIG. 3; 101...Keyboard, 102...Display device, 103...Input character editing control section, 104...Romaji-kana conversion control section.

Claims (1)

[Claims] In the Romaji-Kana conversion method of a device that inputs Japanese using an alphabet display keyboard, there is a string buffer that holds the character string that is undergoing Romaji-Kana conversion, and a character string that corresponds to the characters held in the string buffer. and an attribute buffer that holds the attributes, and performs Romaji-kana conversion on characters in the character string buffer that correspond to pointers attached to the character string buffer and the attribute buffer, respectively. Romaji-kana conversion method.