JPS6323591B2 - - Google Patents

Description

[Detailed description of the invention]

The present invention relates to a character input device capable of inputting Japanese sentences mixed with kanji and kana using the touch type method. (Description of Prior Art and Object of the Present Invention) When entering Japanese text mixed with kanji and kana (hereinafter simply referred to as Japanese text) using the touch type method, the problem is that it is difficult to use a Roman typewriter type keyboard with a small number of keys. The problem is how to input kanji, which have over several thousand character types. Two methods have been proposed as effective input methods to deal with this problem, and some of them have even been put into practical use.
One is a two-stroke code memorization input method, and the other is a kana-kanji conversion input method. However, neither of these two methods has so far been evaluated as a Japanese text input method that has general and versatile practicality. First, there is a two-stroke code memorization input method, which uses kana, Roman, alphabet, etc. for the hiragana, katakana, kanji, numbers, punctuation marks, symbols, etc. used to write Japanese sentences. In this method, fixed length codes of characters, that is, two-stroke codes are assigned, these two-stroke codes are stored in advance, and the two-stroke code of the desired character, symbol, etc. is input by pressing two keys. In order for this method to be effectively established as a touch-type input method for Japanese text, the operator must input more than 2,000 kinds of two-stroke codes (of which 1,800 to 1,900 are kanji codes) in advance using reflexive finger movements. In order to be able to enter the code as , you have to memorize it as a finger movement rather than a head movement. For this reason, it initially takes about 400 hours to memorize the two-stroke code with the fingers (training the fingers to hit the chord reflexively), but this is a major drawback of the two-stroke code memorization input method. It is as follows. Since this method normally uses a 48-key keyboard, it is possible to input 2304 types of two-stroke codes (48 x 48 = 2304). If you also use the space key, you can enter 2401 different two-stroke codes (49 x 49 = 2401). However, if you look at the breakdown of the types of two-stroke codes that can be entered, it is approximately
60% of the chords are based on a stroking pattern (fingering pattern) that is not easy to strike, and half of these chords are based on a stroking pattern that is difficult to strike. It is clear that this coding system is the main cause of the drawbacks of this method. This is because the main premise of this method is the memorization of chords using fingers, and the biggest obstacle to memorizing chords using fingers is chords that are difficult to strike or have a stroking pattern that is not easy to strike. In view of this, the two-stroke code storage input method must be improved as described below. (1) All characters, symbols, etc. should be input using only two-stroke codes that correspond to easy-to-type stroke patterns. For this reason, (for example, when using a 43-key keyboard), there are two types of strokes that correspond to the stroking pattern that makes it easy to type 800 kanji, which have a total usage rate of 90%. Enter the code. Kanji other than the 800 character types listed above are entered using a 4-stroke code, and the 4-stroke code can be easily created by combining two types of 2-stroke codes for the 800 character types mentioned above. In this way, all kanji can be input using two easy-to-type two-stroke codes. In other words,
A 4-stroke chord that is easy to strike is given priority over a 2-stroke chord that is difficult to strike, thereby reducing the number of 2-stroke chords to be memorized. The remaining two-stroke codes that are not assigned to the above 800 kanji are used for hiragana, numbers, punctuation marks, symbols, etc. (Of course, two-stroke codes may be used for Katakana as well.) Codes that are compatible with stroking patterns that are not easy to type or are difficult to type are not used for inputting codes for characters, symbols, etc. These codes may be used mainly to input function control characters (function characters) that are used less frequently. (2) In order to make it possible to freely mix the above two types of fixed length codes (2-stroke code and 4-stroke code), there is no need for a key dedicated to switching the input mode between the two fixed-length code input modes, To enable smooth normal key operation with low rate. (3) It also eliminates the need for shift key operations between inputting a two-stroke code in hiragana and a two-stroke code in katakana. (The first method that can be considered is to include a Katakana code in the normal 2-stroke code, but in that case, punctuation marks, symbols, and 2-stroke
This will result in reducing the number of kanji characters in the typing code, so
Here, let's input Katakana using a different input method. However, this point is simply a matter of code assignment, so the Katakana input method may be changed as necessary. ) By making the improvements as described above, touch type input can be achieved that allows both 2-stroke and 4-stroke chords to be input. One immediate thought would be that this would make high-speed input (high-speed keystrokes), which is the biggest advantage of the conventional two-stroke code memorization input method, impossible, but that is not necessarily the case. . In general Japanese writing, kanji account for about 40% to 50%, but according to this method of combining 2-stroke and 4-stroke codes, the number of strokes required per character is (0.4-0.5) × 0.9 × 2 + (0.4 to 0.5) x 0.1 x 4 + (0.6 to 0.5) x 2 = 2.08 to 2.10 (stroke/
), which is only 4% compared to the ideal case where everything is entered using two-stroke codes.
The stroke will increase by ~5%. When comparing the input speed of this 2-stroke code/4-stroke chord combination method with that of the conventional method, firstly, the conventional method uses many chords that are difficult to type, whereas this method uses many chords that are difficult to type. Since this method inputs all chords using a stroking pattern that is easy to type, the input speed of the compatible method is correspondingly faster. If we consider the increase in speed to be a modest 4% to 5%, the time consumption due to the 4% to 5% increase in the number of required strokes in the combination method can be covered. Second, the number of strokes required per character in the dual-strike method (2.08 to 2.10) includes the number of strokes required to input kanji (external characters) that are input with a 4-stroke code, but in the conventional method Therefore, the number of strokes required to code a custom character with four strokes must also include the number of keystrokes required to switch modes, and taking these factors into account, the input speeds of the conventional method and the compatible method will be the same. . From the above, it can be said that the input speed achieved by the conventional method can be sufficiently maintained by the compatible method. In this way, according to the combination method, all chords are input using easy-to-strike chords, which greatly shortens the training period (memorizing chords with fingers and mastering how to create 4-stroke chords by combining those chords), reduces fatigue, and reduces errors. The keystroke rate will be reduced, resulting in a significant improvement over the conventional method.
(Please add a note to the numbers in the calculation formula above. "0.4~0.5" means kanji content of 40%~50%,
"0.9" is the total usage rate of 90% of the 800 kanji that are entered in the 2-stroke code, "2" is the number of keystrokes required to input the 2-stroke code, and "0.1" is the 4-key usage rate for the 800 kanji that are input in the 2-stroke code. "4" is the number of keystrokes required to enter a 4-stroke code of 10%, and "0.6 to 0.5" is the total usage rate or inclusion of kana, numbers, punctuation, etc. The rates are 60% to 50%, respectively. ) In order to realize the above-mentioned improvements, a key dedicated to mode switching between 2-stroke code input mode and 4-stroke code input mode and a key dedicated to shift switching between 2-stroke hiragana input and 2-stroke katakana input are required. It is necessary to automatically result in switching between the two modes or between shifts without having to operate the system. The present invention allows you to freely input 2-stroke fixed-length codes and 4-stroke fixed-length codes so that you can input Japanese sentences using only easy-to-type stroke patterns without using any dedicated mode switching keys or dedicated shift keys. It is an object of this invention to provide a character input device that can be used in combination, and for this purpose, it is an object of the present invention to make it possible to give a double meaning to each key on a keyboard consisting of a limited number of keys. A dual meaning keyboard such as that described above can also be constructed using keys with special key switch functions. For example, a keyboard consisting of keys that has the function of generating an input signal A when a normal key press operation (or a weak and shallow key press/keystroke) and generating an input signal A' when a key press operation is performed strongly and deeply. You can also consider using it. With this kind of keyboard, when there are N keys, you can also create ^N4 types of two-stroke codes, which shifts a much larger amount of information compared to a normal keyboard that can only input ^N2 types of two-stroke codes. This is because input can be made without using a switching key or a mode switching key. However, a system that can be configured based on the strength of key press operations (shallow and deep) makes it impossible to use widely used keyboards. It is not necessarily a good idea to use a keyboard that can switch the input signal depending on the depth (shallow and deep). Furthermore, there is currently no data to show that key press force (shallow and deep) operations are practical from an ergonomic point of view, and furthermore, such special keyboards are expensive, etc. Considering the conditions, I would like to refrain from using a special keyboard. After all, it is desirable to obtain the above-mentioned effects by using an ordinary keyboard which is widely used. Note that a normal keyboard mainly refers to a keyboard consisting of keys called so-called tactile keys. Therefore, the present invention allows you to freely input two-stroke chords and four-stroke chords (both fixed length) so that you can input Japanese sentences smoothly using only easy-to-type stroke patterns using an ordinary keyboard. The purpose is to provide a character input device (especially a signal processing circuit for a keyboard) that can be used interchangeably. Another object of the present invention is to propose a new kana-kanji conversion input method. The biggest problem with conventional kana-kanji conversion input methods is that the conversion rate cannot be significantly improved unless a text analysis/semantic analysis program is included in the automatic conversion program based on kana input to kanji. It's there. At present, this method has only been put to practical use in a limited work area, and it has not yet become a fully automatic conversion method. The present invention aims to obtain a Japanese text input method that utilizes the current technology of this method and has general and versatile practicality.
We propose the following modified improvement measures. There are many kanji that correspond to the kana character strings ``sei'' and ``kou.'' Kana string "Seiko"
Although the number of kanji idioms corresponding to ``is smaller'', it must be said that the number of candidates is still large. However, the kanji idioms that correspond to the character strings ``Seikou'', ``Seikou'', ``Seikou'', and ``Seikou'', which are written in a mixture of kanji and kana, are ``success'' and ``exquisite or exquisite work'', respectively. , "Straight attack", "Raw hard"
Is it uniquely specified?
The number of candidates is also extremely limited. By the way, in order to input "Seikou" and "Seikou" separately, it would have been better to input them together as "Seikou" and "Seikou", but in this case, after "Seikou" was converted, "Seikou" would be entered. If you want to input ``, then you have no choice but to correct ``technique'' to ``technique.'' However, when making the corrections, do not use the kana-kanji conversion method, but directly
In many cases, the method of inputting is better. Thus, it is possible to consider a touch-type input method for Japanese text that focuses on these two points. That is, the present invention provides an input method that can achieve a high conversion rate without adding a text analysis/semantic analysis program to the conversion program by inputting kana characters to be converted into kanji together with kanji, and To provide a character input device that enables combination with the Japanese input method using 2-stroke codes and 4-stroke codes, and allows these two different input methods to be freely used together as necessary using the same keyboard. That is. This composite input method uses both 2-stroke and 4-stroke codes to clearly distinguish between kana to be converted into kanji (this is called converted kana) and other characters and symbols. In other words, it can be constructed by making it possible to freely input mixed inputs without any input. Therefore, in this composite input method, it is necessary to be able to input characters belonging to different categories using the following four types of input modes (shifts). (1) Two-stroke code input for hiragana, numbers, punctuation, symbols, and kanji (2) Two-stroke code input for Katakana (or shift. When not inputting in the input mode of (1) above) ( 3) Inputting a 4-stroke code (kanji other than those that are input with a 2-stroke code) (4) Inputting a 2-stroke code for converted kana At this time, input these 4 types of characters by performing key operations dedicated to mode (shift) switching. It needs to be done smoothly and smoothly. The present invention provides a keyboard that can freely mix (combine) the above four types of character input without operating a dedicated input mode (shift) switch key, and input characters in each of the above categories using a stroke pattern that is easy to type. This paper also attempts to present technical measures to achieve this by using a non-specialized general keyboard. (Summary of the Present Invention) One of the objects of the present invention described above is to eliminate one specific key by simultaneously pressing one specific key on the keyboard and one desired key from the remaining keys. This is accomplished by assigning different meanings (codes) to other keys. That is, all keys except one specific key are assigned two meanings (codes) depending on whether or not they are pressed at the same time as the specific key. By assigning two meanings (codings) to each key on this keyboard (excluding one specific key),
That is, by using them properly, it is possible to double the number of input codes, and also eliminates the need for a key (operation) dedicated to mode switching between the 2-stroke code input mode and the 4-stroke code input mode, and
A dedicated key (operation) for shifting between two-stroke hiragana input and two-stroke katakana input can be made unnecessary. What should be noted here is that the above-mentioned specific key is neither a mode switching key nor a shift switching key. In other words, one specific key used for simultaneous keystrokes is a key that is also used as a code input key when inputting a two-stroke code or a four-stroke code, but due to physical conditions (one key cannot be pressed). Since simultaneous keystrokes cannot occur, it is simply impossible to assign two meanings (codes) to simultaneous keystrokes. The first keystroke (or second keystroke) of a 4-stroke code input
By introducing the above simultaneous keystrokes in the keystrokes), 2
It becomes possible to automatically identify that the input code is not a typed code. Also, 2 in kana input
By introducing the above-mentioned simultaneous keystroke at the first keystroke (or second keystroke) of the typing code input, two-stroke hiragana input and two-stroke katakana input can be automatically distinguished. The simultaneous keystroke processing circuit detects whether or not the specific key and one of the other keys are pressed simultaneously, and if they are pressed simultaneously, converts the key code (assigns a different meaning or code). thus,
A two-stroke code input or a four-stroke code input is determined from a converted key code and an unconverted key code, or a two-stroke hiragana input and a two-stroke katakana input are determined. The above-mentioned one specific key is desirably arranged at a position where it can be easily operated with the thumb. This is to facilitate simultaneous keystrokes. In the example,
Bar keys arranged horizontally at the bottom of the keyboard are used as the above-mentioned specific keys. In addition, in the embodiment, in addition to the 2-stroke code input and 4-stroke code input, a 3-stroke key input method is shown as a kanji input method, but this uses an automatic shift method as a modification of the 2-stroke code input. I just showed you what I would get. Another object of the present invention is to code the kana input to be converted into kanji differently from normal hiragana or katakana (as a converted kana code), and to code the character strings before and after the converted kana (especially the converted kana). This is achieved by automatically converting the converted kana into a predetermined kanji with high accuracy, judging from the relationship between the kanji and the kanji input before and after the kana. In other words, by identifying the kana to be converted into kanji as a conversion kana from other kana, and inputting the kanji to be combined with this conversion kana, it is possible to perform highly accurate kana-to-kanji conversion that is close to fully automatic. ing. Converted kana is entered as a two-stroke code, but by inputting the second keystroke (or the first keystroke, i.e., the opposite one from when inputting katakana) using the above simultaneous keystrokes, the meaning (code) is automatically entered. It is becoming more and more common to (Description of an Embodiment) An embodiment of the present invention will be described in detail below with reference to the accompanying drawings. FIG. 1 is a schematic block diagram of the overall configuration of an embodiment of a character input device according to the present invention, and an example of the key arrangement of the keyboard section 10 in this embodiment is shown in FIG. The keyboard section 10 consists of a key group 11 consisting of a plurality of push-type key switches and one horizontally elongated bar key BK disposed below the key group 11. On the key top of each key in the key group 11, characters, symbols, etc. are displayed as illustrated in FIG. For the key group 11, the touch fingering method is applied. "I", "O",
The keys "A" and "E" are the home positions of the four fingers of the left hand excluding the thumb; "R", "N", "T",
The "S" key is the home position of the four fingers of the right hand excluding the thumb. Each key in the key group 11 is arranged so as to be most suitable for operation with a predetermined finger, as shown in FIG. Berkey BK
shall be operated by pressing with the thumb of the free hand. Within the key group 11, simultaneous multiple keystrokes are prohibited, but rollover input is possible. Of course, it is possible to press the bar key BK and key group 11 at the same time. For convenience, each key in the keyboard unit 10 is divided into five key sets G1 to G5 and processed. In FIG. 2, keys belonging to each key set G1 to G5 are shown surrounded by broken lines. In this embodiment, the keyboard unit 10 shown in FIG. 2 is used to input hiragana, katakana, numbers, punctuation/symbols, kanji, and kana for kana-kanji conversion (hereinafter referred to as converted kana). It's becoming possible. In order to understand the outline of character input according to the mode of the present invention, an outline of the key operation method will first be explained. Outline of key operation method As a general rule, each character or symbol and frequently used kanji are entered by pressing the key twice (two strokes). Exceptionally, kanji (for example, many kanji that are used infrequently) require 3 or 4 key strokes.
Some items are entered by typing. Furthermore, in principle, pressing one key is one keystroke, but when one key in the key group 11 and the bar key BK are pressed at the same time, it is also regarded as one keystroke. key group 1
By simultaneously pressing the 1 key and the bar key BK, the same effect as doubling the number of keys in the key group 11 can be obtained. Hiragana is input using two keystrokes using pseudo-Roman characters. A desired Roman consonant key among the keys of the key set G2 in the key group 11 is pressed with the first key stroke, and a desired Roman alphabet vowel key among the keys G4 of the key set G4 is pressed with the second key stroke. For example, if you press the "K" key and the "A" key in sequence, the hiragana "ka" will be displayed.
is input. For the hiragana in the "A" line, the "," key of key set G2 is pressed as the first key press, and the desired vowel key is pressed as the second key press. Numbers are entered in two keystrokes by pressing a desired number key in the key set G1 or G3 as the first keystroke, and pressing the bar key BK as the second keystroke. Punctuation marks and symbols are input with two keystrokes by pressing a predetermined key in the key group 11 and the bar key BK back and forth in a predetermined order. There are two types of kanji: those that are entered with two keystrokes (hereinafter referred to as two-stroke kanji) and those that are entered with three keystrokes (hereinafter referred to as "two-stroke kanji").
There are 3-stroke kanji) and 4-stroke kanji (hereinafter referred to as 4-stroke kanji). "Two-stroke kanji" is a method to input frequently used kanji (purpose As mentioned above, approximately 800 kanji) are assigned to this. "3 stroke kanji"
is input by a shift operation and a predetermined two-stroke operation in the key group 11, but there is no special shift key, and the first key to input a "two-stroke kanji" is pressed twice. By doing so, the shift state is automatically entered, and in the end, a ``3-stroke kanji'' is input with a total of 3 keystrokes. "4 stroke kanji"
The first keystroke is performed by simultaneously pressing the bar key BK and a desired key in the key group 11, and the subsequent second, third, and fourth keystrokes are performed by pressing a desired key in the keyboard section 10. I'll do it. Like hiragana, katakana is entered using two keystrokes using pseudo-roman characters, but in order to distinguish it from hiragana, the first
Keystrokes are performed using the desired key in key set G2 and the bar key.
This is done by pressing keys simultaneously with BK, and the second keystroke is a normal vowel key input. Conversion kana (kana for kanji-to-kanji conversion) is input using two keystrokes using pseudo-roman characters, just like hiragana, but in order to distinguish it from hiragana or katakana, the first keystroke is a normal consonant key input, and the second keystroke is is the desired vowel key of key set G4 and bar key BK
The keys must be pressed at the same time. The converted kana input in this way is eventually automatically converted to kanji. In addition, stroke patterns that are easy to type will be assigned to hiragana, numbers, punctuation marks, and symbols that are input in two strokes, and katakana, which consists of a stroke pattern similar to hiragana (the only difference is that keys can be pressed simultaneously with bar keys) It can be said that inputting converted kana is also an easy-to-type stroking pattern. Outline of Overall Configuration of Embodiment In FIG. 1, a key scanning detection encoding circuit 12 scans each key in the key group 11 to detect a pressed key and generates a code signal for the pressed key. This key scanning detection encoding circuit 12 is configured to enable key input by rollover. In other words, if multiple keys are pressed at exactly the same time, it will be considered an error input and no key code will be generated, but if the next key is pressed before the previously pressed key is released, multiple keys will temporarily be pressed. Even if a key is pressed, key codes for each pressed key are generated sequentially in the order in which they are pressed. Such a rollover key detection circuit is already known, for example,
The key scanning detection encoding circuit 1 can be easily constructed according to the configuration shown in No. 17017 or Japanese Patent Application Laid-Open No. 52-58423.
2 can be configured. The key code of the pressed key in the key group 11 generated by the key scan detection encoding circuit 12 is added to the (A) input of the simultaneous key press processing circuit 13. On the other hand, when the bar key BK is pressed, a key code indicating the bar key BK is generated from the encoder 14 and is applied to the (B) input of the simultaneous key press processing circuit 13. The simultaneous keystroke processing circuit 13 determines whether or not the (A) input and (B) input, that is, the key of the key group 11 and the bar key BK, are pressed at the same time, and if they are pressed at the same time, the (A) input and the Converts the value of the key code being pressed (to a value that does not overlap with the key code indicating each key on the keyboard section 10) and outputs it to the key code bus
The key code added to (A) input or (B) input is output as is to key code bus X. (A) The key code added to the input is shown as "A", (B) the key code added to the input is shown as "B", (A) the key code converted from the input key code A is shown as "a". The input/output relationship in the simultaneous keystroke processing circuit 13 is shown in Table 1. “〇” in Table 1
The mark indicates that the key code has been entered,
A "-" mark indicates that no key code has been input.

[Table] Note that the simultaneous keystroke processing circuit 13 can tolerate not only the case where the key of the key group 11 and the bar key BK are pressed at exactly the same time, but also even if there is some time lag between the timings of the two keystrokes. If there is a predetermined time difference, the processes are treated as simultaneous. FIG. 3 shows a specific example of the simultaneous keystroke processing circuit 13 that recognizes a certain time error when determining simultaneous keystrokes. In FIG. 3, the rising edge detection circuits 15 and 16 are
It detects that the content of the key code given to (A) input or (B) input changes to a new key code (the rising edge of the key code), and generates one short pulse when the rising edge is detected. . The rising edge detection pulse outputted from the rising edge detection circuit 15 is delayed by one cycle of the system clock pulse φ ₁₀ (for example, about 10 to several tens of microseconds) by a delay flip-flop 17, and becomes a load command for the register 18. The rising edge detection pulse output from the rising edge detection circuit 16 serves as a load command for the register 19. Therefore, when a new key is pressed in the key group 11, a pulse is generated from the rising edge detection circuit 15,
The key code of the new pressed key is stored in the register 18. Further, when the bar key BK is pressed, a pulse is generated from the rise detection circuit 16 at the beginning of the press, and the key code of the bar key is stored in the register 19. When some key code is stored in the register 18 or 19, "1" is output from the OR circuit 20 or 21 into which the output is input, and the timer 22 or 23 is set. timer 2
The operation time (T) of 2 and 23 corresponds to the maximum allowable time difference that is considered to be simultaneous keystrokes. For example, the operating time (T) is about 50 milliseconds. Timer 22, 23
The output is "1" from the time the timer is set until the operating time (T) ends or until the timer is cleared, and is "0" otherwise. The key code output from the register 18 is input to the gate 24 and also to the gate 26 via the key code conversion circuit 25. The key code of Barkey BK output from the register 19 is input to the gate 27. Each gate 24,2
As will be described later, 6 and 27 are electrically connected in accordance with the result of determining whether the key of the key group 11 and the bar key BK are pressed simultaneously. One gate 24, 26, 2 conductive
The key code output from 7 is sent to multiplexer 2.
8 to the key code bus X. Multiplexer 28 is comprised of a plurality of OR circuits that connect the output key codes of each gate 24, 26, 27 to key code bus X. When the gate 24 is conductive, the key code of the pressed key of the key group 11 given to the input (A) is given as is to the key code bus X, and when the gate 26 is conductive, the pressed key code of the key group 11 is converted. The key code a is applied to the key code bus X, and when the gate 27 is conductive, the key code of the bar key BK applied to the input (B) is applied to the key code bus The key code value converted by the key code conversion circuit 25 is not only a value that does not overlap with the key code of each key of the keyboard section 10, but also a predetermined value that differs depending on the input key code value. converted to a value. There are seven possible time relationships between the key codes applied to the (A) input and (B) input of the simultaneous keystroke processing circuit 13, as shown in FIGS. 4a to 4g. Here, (A) input and (B)
The inputs are considered simultaneous in FIG. 4d or e. Note that when the (A) input and (B) input are completely simultaneous, they belong to either d or e in Figure 4.
The operation of the circuit of FIG. 3 will be explained below with respect to each of FIGS. 4a to 4f. It is assumed that the registers 18 and 19 are cleared in the initial state. As shown in Figure 4a, if key codes are input consecutively to (A) input with a time difference longer than the timer time T, when the first key code is added to (A) input,
A short pulse is generated from the rising edge detection circuit 15 and sent to the gate 24 via the OR circuit 29 and the AND circuit 30.
An enable pulse is applied to the key code bus X, but since the contents of the register 18 are cleared at this time, no key code is applied to the key code bus X even if the gate 24 is temporarily turned on. It is assumed that the width of the output short pulse of the rising edge detection circuit 15 is equal to or shorter than one cycle of the system clock pulse _φ10 . The first key code is stored in register 18 when the pulse delayed by delay flip-flop 17 is applied to the load input of register 18. When the key code is stored in the register 18, the output of the OR circuit 20 rises to "1", and the timer 22
is set. When the time T of the timer 22 has elapsed, the inverter 31 inverts the output of the timer 22.
The output rises to "1", and "1" is applied to the AND circuit 30 via the OR circuit 29. Since "1" obtained by inverting the output "0" of the timer 23 by the inverter 32 is added to the other input of the AND circuit 30, the condition of the AND circuit 30 is satisfied, and the output of the AND circuit 30 is "1". The gate 24 is made conductive. Therefore, the key code stored in the register 18 (the first key code given to the input (A)) is transferred to the key code bus X via the gate 24.
given to. At this time, the gate 26 is closed. The output “1” of the inverter 31 is the differentiating circuit 3
3 is input. Differentiating circuit 33 outputs a short pulse at a time delayed by one period of clock pulse _φ10 from when the input rises to "1". This differential short pulse clears the register 18 via the OR circuit 34. When the second key code is applied to the (A) input, the gate 24 is turned on when the timer 22 expires, and the register 18 is then cleared, as described above. As shown in Figure 4b, if key codes are given consecutively to (A) input at a time interval shorter than the timer time T (this can happen since key group 11 can be rolled over), the first key code is First, it is stored in the register 18 as described above. Next, when the second key code is applied to the (A) input, a rising edge detection pulse is generated from the rising edge detection circuit 15, and the gate 24
becomes conductive, and the first key code stored in register 18 is supplied to bus X. The rising edge detection pulse delayed by the delay flip-flop 17 is applied to the load input of the register 18 and clears the timer 22 via the OR circuit 35. Therefore, the timer 22 that was in the middle of the operating time T is once cleared, but the output of the OR circuit 20 becomes "1" immediately.
and resume tire operation. The second key code is applied to the bus X via the gate 24 when the timer 22 expires, as described above. Note that it is impossible for the bar key BK to be pressed continuously at a time interval shorter than the timer time T.
This is because there is only one Berkey BK, so no matter how quickly you hit it repeatedly, the timer interval is T.
(time that is considered simultaneous). As shown in Figure 4 c or g, A input and B input
When key codes are input alternately to the input with a time difference longer than the timer time T, the key code applied to the (A) input and the key code applied to the B input are stored in registers 18 and 19, respectively, at the rising timing of each. Ru. However, since the time difference is longer than the timer time T, the key codes will not be stored in both registers 18 and 19 at the same time. Therefore, the key code of input (A) stored in register 18 is timer 2.
2 is applied to bus X via gate 24, after which register 18 is cleared. In addition, the key code of the input (B) stored in the register 19 is stored in both timers 2 and 2 when the timer 23 expires.
When the condition of the AND circuit 38 to which signals obtained by inverting the outputs of the outputs 2 and 23 are inputted by the inverters 36 and 37 is satisfied, the gate 27 is made conductive and the signals are supplied to the bus X. Note that a short pulse is generated from the differentiating circuit 39 one cycle of the clock pulse _φ10 after the time T of the timer 23 ends, and the OR circuit 4
Register 19 is cleared via 0. As shown in Figure 4d, if the rising edge of the (A) input key code is earlier than the rising edge of the (B) input key code, and the time difference is shorter than the timer time T, first, the (A) input The key code is stored in the register 18 in response to the rising edge of the key code. This causes the timer 22 to be set. timer 22
The set output "1" is applied to the AND circuit 41.
When the key code of the input (B) rises during the time T of the timer 22, the key code is stored in the register 19. As a result, the output of the OR circuit 21 becomes “1”
and is added to the other input of the AND circuit 41. Therefore, the output of the AND circuit 41 becomes "1", and an enable pulse is applied to the gate 26 via the OR circuit 42. This makes register 1
A key code a obtained by converting the key code of input (A) stored in 8 by a key code conversion circuit 25 is supplied to the bus X via a gate 26. At this time, the output of the timer 22 is inverted from "1" to the inverter 3.
The output of the AND circuit 38 becomes "0" due to the output "0" of the gate 23, and the output of the AND circuit 30 becomes "0" due to the output "0" of the inverter 32 which is an inversion of the output "1" of the timer 23, and the output of the AND circuit 38 becomes "0". and 27 are both non-conductive. Therefore, the key codes of input (A) and input (B) which are considered to have been pressed at the same time are blocked, and the converted key code a is given to bus X instead. Output “1” of OR circuit 42
is delayed by one cycle of clock pulse _φ10 by a delay flip-flop, and the timer 22, 23 registers 18, 19 are all cleared by the delayed output via OR circuits 34, 35, 40. As shown in Figure 4e, (B) input key code is
(A) If it rises earlier than the input and the time difference is shorter than the timer time T, first register 19
The key code is stored and the timer 23 is set. If a key code is given to the input (A) and the key code is stored in the register 18 before the time T of the timer 23 ends, the output of the OR circuit 20 becomes "1".
The timer 22 is then set. The condition of the AND circuit 44 inputting the output “1” of the timer 23 and the output “1” of the OR circuit 20 is satisfied, and the OR circuit 42
An enable pulse is applied to gate 26 via. As a result, the key code a converted by the key code conversion circuit 25 is supplied to the bus X in the same way as described above. As shown in FIG.
7 becomes conductive, and the key code of the register 19 is supplied to the bus X as is. Through the processing of the simultaneous keystroke processing circuit 13 as described above, the key code bus The key codes indicating the pressing of the bar key BK or the key codes indicating the pressing of the bar key BK are sequentially supplied in accordance with the keystroke order. In FIG. 1, the key code of the key code bus X is supplied to a key set detection section 45 and a post-input key code register 46. The key set detection unit 45 detects which of the key sets G1 to G5 and g1 to g4 the key code applied to the key code bus X belongs to, and generates a set code. Key set G1 to G5
As shown in FIG. 2, the keys on the keyboard section 10 are classified, and the sets g1 to g4 are used to classify the converted key codes a resulting from simultaneous keystrokes. The key code of each key in the key group 11 is shown using the same symbol as the key top display in FIG.
The key code of the bar key BK is indicated by BK, and the key code a obtained by converting each key code of key group 11 (A input) is indicated by a symbol with a dash symbol attached to the key top display in Fig. 2, and each key set G1 ~
G5, the set code of g1~g4 is the same code G1~
G5, g1 to g4, the relationship between the set code detected by the key set detection unit 45 and the key code is as shown in Table 2.

[Table] The set code table storage unit 45A of the key set detection unit 45 stores the input/output relationship shown in Table 2 in advance, and reads out the set code corresponding to the input key code. Incidentally, if the specific bit portion of the key code is assigned to the set code, the storage section 45A is unnecessary. The set code output from the key set detection section 45 is supplied to the post-input set code register 48. The output of the last input key code register 46 is input to the first input key code register 47. The output of the last input set code register 48 is input to the first input set code register 49. The post-input key code register 46 and post-input set code register 48 respectively store the key code just supplied to the key code bus X and its set code. The first input key code register 47 and the first input set code register 49 contain the second input key code register 4.
6 and the key code supplied to the key code bus X immediately before the key code stored in the post-input set code register 48 and its set code are stored, respectively. That is, immediately before a new key code and set code are read into the post-input key code register 46 and post-input set code register 48 by pressing a key on the keyboard unit 10,
The key code and set code (of the key pressed immediately before) that had been stored in the registers 46 and 48 are now transferred (shifted) to the first-input key code register 47 and the first-input set code register 49. ing. Therefore, the post-input register 4
The shift from 6, 48 to the first input registers 47, 49 is performed every time a key is pressed on the keyboard section 10. In this way, the stored codes in the later input registers 46, 48 and the earlier input registers 47, 49 are sequentially switched each time a key is pressed. From the later input registers 46 and 48 to the earlier input registers 47 and 49
Shift control of the sequential feed to can be performed in any manner according to existing digital technology. For example, when a new key code is supplied to the key code bus
One pulse (for example, 1 microsecond width) is generated from the one-shot circuit 51, and the output of this one-shot circuit 51 is added to the load control input of the first input registers 47 and 49, and the code stored in the second input registers 46 and 48 is sent to the first input register. 47 and 49, respectively. Further, the output of the one-shot circuit 51 is delayed by a very short time (for example, 1 microsecond) by a system lock pulse φ in a delay circuit 52, and the output of this delay circuit 52 is input to the load control input of the rear input registers 46 and 48. In addition, the key code currently being supplied to the key code bus X and its set code are read into the post-input registers 46 and 48. Note that the clock pulse φ is faster than _φ10 . As mentioned above, in principle, the post-input register 4
The contents of registers 6 and 48 are sequentially shifted to the first input registers 47 and 49, but as will be described later, when the clear signal CLR is applied from the clear signal generation logic 53, the contents of all registers 46 to 49 are cleared. Every time this clearing is performed, the shift from the later input registers 46, 48 to the earlier input registers 47, 49 is stopped. In this example, there are often two strokes (the key is pressed twice)
A clear signal CLR is generated every time. In other words, when a key is pressed twice, the code of the first key pressed is stored in the first input registers 47 and 4.
9, and the code of the next input (pressed) key is stored in the post-input registers 46 and 48, respectively. When the necessary processing is completed based on the contents of the first input and second input registers 46 to 49, a clear signal CLR is generated, and the register 4
The contents of 6 to 49 are cleared. Therefore, the empty contents of the input registers 46 and 48 after being cleared (usually) on the third key press are transferred to the first input register 4.
7, 49, so the first input register 4
The stored contents of 7 and 49 are also empty, and the code of the third key pressed is stored in the post-input registers 46 and 48. When the fourth key is pressed, the code of the third key is shifted to the first input registers 47 and 49, and the code of the fourth key is stored in the second input registers 46 and 48. After the necessary processing, the clear signal CLR is generated again.
Thus, generally every two keystrokes registers 46 to 4 are
9 is given a clear signal CLR. However, as explained later, in exceptional cases, the clear signal is sent on the third keystroke.
CLR may occur. The set codes stored in the post-input set code register 48 and the pre-input set code register 49 are input to the key function determining unit 55 via the key function converting unit 54, respectively. The key function determination unit 55 determines the function of the input key based on the set code of the first input key and the set code of the second input key, that is, based on the relationship between the input keys. With this invention, the same key can be used for various functions (Hiragana input,
This key function determination unit 55 automatically determines which function the currently input key was input (used) based on the relationship between key inputs. I am trying to determine the function of the key. Key function determining section 55
The functions determined by the key are: hiragana input, number input, punctuation/symbol input, 2-stroke kanji input, 3-stroke kanji input, 4-stroke kanji input, katakana input, and converted kana input. Was it input? Here, key set G1~
Since the keys G5, g1 to g4 are used in common for each set, it is sufficient to determine the key function based on the set code rather than the key code. The key function converter 54 converts the contents of the set code given from the registers 48 and 49 according to the function determined by the key function determiner 55, and inputs the converted set code to the key function determiner 55. Specifically, when the first two keystrokes in the four-stroke kanji input function are confirmed, the set code for the remaining two keystrokes is converted by the key function converter 54, but in other cases, the set code is converted by the key function converter 54. The data is passed through as it is without being converted in step 54 and is input to the key function determination section 55. A signal representing the function determined by the key function determining section 55 is input to the character information generating section 56. Key codes are supplied to the character information generating section 56 from the later input key code register 46 and the earlier input key code register 47, respectively. The character information generating section 56 uses the function determined by the key function determining section 55 and the register 4.
From the combination of two key codes given from 6 and 47, input characters, etc. (i.e. input hiragana, numbers, punctuation/symbols, kanji, katakana,
or converted kana).
The character information generated from the character information generation section 56 is input to the kana-kanji automatic conversion section 57. In the kana-kanji automatic conversion section 57, if the input character information is converted kana information, it is converted into a predetermined kanji, but if it is other information, it is passed through as is. The character information output from the automatic kana-kanji conversion unit 57 is supplied to an output printer (not shown) or a display device (not shown), and the key-input characters, symbols, etc. are finally output (printed or Is displayed. A detailed example of the key function conversion section 54, key function determination section 55, character information generation section 56, and clear signal generation logic 53 is shown in FIG. A detailed example of the automatic kana-kanji generator 57 is shown in FIG. Overview of Fig. 5 A more detailed explanation of each circuit shown in Fig. 5 will be given later in the operation explanation for each example of kana input, kanji input, etc.
An overview of the configurations of Nos. 3 to 56 will be described below. The key function converter 54 is connected to a code converter 58 into which the last input set code and the first input set code supplied from the registers 48 and 49 are respectively input, and a signal indicating that the first two keystrokes of four-stroke kanji input have been made. 2S. The code conversion section 58 stores the signal 2 in the temporary storage section 59.
When S is stored, the values of the first input set code and the second input set code are converted as shown in Table 3, but when the signal 2S is not stored, the input code is output as is without conversion. do. still,
The temporary storage section 59 consists of a delay flip-flop 60 which delays the signal 2S by a predetermined time, and a set-reset type flip-flop 61 which is set by the output of the delay flip-flop 60. Third
The table shows a set code conversion table in the code conversion unit 58.

[Table] In Table 3, the set codes after conversion are shown as 4G1 to 4G5. In addition, if set codes g1 to g4 are input when keys are pressed simultaneously with the bar key, this set code g is omitted in Table 3.
1 to g4 are always passed through the code converter 58 without being converted. The converted set codes 4G1 to 4G5 indicate the remaining two keystrokes in the four-stroke kanji input. The key function determining section 55 includes a key function ROM (ROM is an abbreviation for read-only memory) 62.
The set code output from the code converter 58 is supplied to the address input of the key function ROM 62. The key function ROM 62 produces one of KR, MR, HR, 2S, 4S, and SF output depending on the combination of the last input set code and the first input set given to the address input. Of course, if the combination does not match, no output will be generated. Note that an error signal (not shown) may be generated in response to an incorrect combination. The key function is determined by the key function ROM and AND circuits 63, 64, etc. provided on the output side thereof. The key function determination table, that is, the input/output table of the key function ROM 62 is shown in Table 4. "Output" in the table indicates the determined function.

[Table] MR is a signal indicating that the input function for hiragana, numbers, punctuation/symbols, or two-stroke kanji has been determined. The signal MR read from the ROM 62 is input to AND circuits 63 and 64.
The signal MR output from 3 is used as a signal indicating the above function. The signal MRS outputted from the AND circuit 64 is a signal indicating that the three-stroke kanji input function has been determined. SF is a signal indicating that the first two keystrokes (ie, automatic shift operation) for inputting three-stroke kanji have been performed. This signal SF is stored in a temporary storage section 65, and its storage output is applied to the other input of the AND circuit 64 and also to the other input of the AND circuit 63 via the inverter 66. Therefore, if the signal SF is stored first, the signal MR read out from the ROM 62 is changed to MRS, indicating that 3-stroke kanji is input; however, if the signal SF is not stored, the signal MR read out from the ROM 62 is The entered MR indicates that it is an input of two-stroke kanji or hiragana without being changed. KR is a signal indicating that the katakana input function has been determined. HR is a signal indicating that the converted kana input function has been determined. 2S is a signal indicating the first two keystrokes of a four-stroke kanji input, and 4S is a signal indicating the remaining two keystrokes of a four-stroke kanji input. Clear signal generation logic 53 includes an OR circuit 67 and a delay circuit 68. The OR circuit 67 has
Function determination signal output from key function determination section 55
MR, MRS, KR, HR, 2S, and 4S are input. That is, when any function other than the automatic shift operation SF is determined, "1" is output from the OR circuit 67. Output “1” of OR circuit 67
is delayed by a small amount of time (for example, one to several cycles of clock pulse φ) in delay circuit 68, and the clear signal
CLR is obtained. Therefore, when any function other than automatic shift SF is determined, the clear signal CLR is generated after a certain period of time, and the registers 46 to 49 are
All memory codes are cleared. Therefore, normally, all registers 46 to 49 are cleared every two keystrokes. In the case of automatic shift SF, a clear signal CLR is generated based on the signal MRS at the third key press, and the temporary storage section 65 storing the signal SF is also cleared by this clear signal CLR. The output of the OR circuit 67 is the delay flip-flop 9.
0 to the temporary storage section 59 of the key function conversion section 54.
It is also applied to the reset input R of the flip-flop 61 within. When the signal 2S is read from the key function ROM 62, "1" is output from the OR circuit 67 and applied to the reset input of the flip-flop 61. The signal 2S is delayed by a delay flip-flop 60 in the temporary storage section 59 and then applied to the set input of the flip-flop 61, so that the signal 2S is
The flip-flop 61 is set after a certain period of time (one bit time) after the data is read from the ROM 62. Thereafter, when a certain function is determined, the flip-flop 61 is immediately reset by the output "1" of the OR circuit 67 (that is, the memory of 2S in the temporary storage section 59 is cleared).
It is assumed that the delay flip-flops 60 and 90 have the same delay time, but the flip-flop 61 has set priority. The character information generating section 56 includes a character code ROM section that stores in advance the codes of all kana characters, kanji, etc. that can be input with this character input device. The character code ROM section includes main ROM 69, corresponding to each function determined by the key function determination section 55,
It is equipped with a 3-stroke kanji ROM 70, a hiragana ROM 71, a converted kana ROM 72, and a 4-stroke kanji ROM 73. In the character code ROM section, only one ROM (one of 69 to 73) corresponding to the function determined by the key function determining section 55 is enabled for reading operation. In other words, the signal MR is the main
The signal MRS is input to the enable input of ROM69, and the signal KR is input to the enable input of ROM70.
is the signal to the enable input of Katakana ROM71.
HR is the enable input of conversion kana ROM72,
The signal 4S is input to each enable input of the 4-stroke kanji ROM 73, and the signal “1” is input to the enable input.
Only one ROM to which the is input is in a readable state. Registers 46 and 47 (first
The post-input key code and the first-input key code are input from (Fig.). Signal from key function determining section 55
A predetermined character code is read out from one ROM which is in a readable state by MR to 4S in accordance with the combination of two address signals applied to the address input, that is, the last input key code and the first input key code. However, as mentioned below, 4-stroke kanji ROM
73 has four address inputs, including the key codes for the first and second keystrokes stored in the register 75 and the third and fourth keystrokes supplied from registers 46 and 47. A predetermined Kanji code is read out based on the code. Character code
The character codes read from each ROM 69 to 73 of the ROM section are led to the output bus 77 via the multiplexer 74, and are sent to the kana-kanji automatic conversion section 57 (sixth
Figure). The last input key code and the first input key code are input to the register 75 from the registers 46 and 47. A signal 2S output from the key function determining section 55 is supplied as a load signal for this register 75. Therefore, when the first two key strokes of the four-stroke kanji input are performed, the register 75 enters the writing state, and the first key code and the second key code stored in the registers 46 and 47 are stored, respectively. do. The first keystroke code and the second keystroke code stored in the register 75 are 4-stroke kanji.
It is added to the first keystroke address and second keystroke address of the ROM 73. Subsequently, when the third and fourth keystrokes for inputting four-stroke kanji characters are made, a signal 4S is generated from the key function determining section 55, and the four-stroke kanji ROM 73 becomes readable. At this time, since the key codes for the third and fourth keystrokes are stored in the subsequent input key code register 46 and the first input key code register 47, respectively, the four-stroke kanji ROM 73 is in a readable state by the signal 4S. When this happens, the four address inputs of this ROM73 include the first one for the four-stroke kanji.
All the key codes for the keystrokes to the fourth keystroke are available, and a predetermined kanji code is read out from the ROM 73 based on these four keycodes. Signal 4S is input to a delay flip-flop 76 whose delayed output is applied to the clear input of register 75. Therefore,
Immediately after the four-stroke kanji code is read from the ROM 73, the register 75 is cleared. Of course, other
Register 46 in ROM69, 70, 71, 72
Only the last-input and first-input key codes from 47 and 47 are input as addresses, and the output of register 75 is not added. The main ROM 69 stores in advance codes representing hiragana characters, codes representing numbers, codes representing punctuation marks/symbols, and codes representing two-stroke kanji characters, corresponding to the combination of first-input keys and second-input keys. I remember. Since the two-stroke key combinations for inputting hiragana, numbers, punctuation/symbols, and two-stroke kanji are not duplicated, they can be stored together in one ROM 69. First, an example of the hiragana table in the main ROM 69 is shown in Table 5.

[Table] In Table 5, at the intersection of the first input key shown in the horizontal column and the second input key shown in the vertical column, the code of the hiragana character read by the combination of both keys is written.
Although the hiragana themselves are listed in Table 5, these are for the sake of ease of understanding, and it goes without saying that the code (for example, JIS) corresponding to the hiragana is actually stored in the ROM. . In Table 5, circled hiragana indicate lowercase letters. As can be seen from Table 5 and FIG. 2, two-stroke pseudo Roman alphabet input is used for kana input. For example, "ma"
is input by two strokes of the key with "M" displayed on it and the key with "A" displayed on it. Next, Table 6 shows the numbers in main ROM69, punctuation marks,
A part of the storage table for symbols and two-stroke kanji codes is shown.

[Table] 5 5 5 5 5 5 5 5
5 5 5 5 5 5

Claims (1)

[Scope of Claims] 1. A device for inputting characters one by one by successive two to four keystrokes, comprising: a keyboard section; and whether each key on the keyboard section is pressed individually or a specific key within the keyboard section; A key code generating means that generates different key codes depending on whether the keys are pressed at the same time, and a character code that generates a character code based on a combination of two to four key codes successively generated by the key code generating means. and if the key code corresponding to a predetermined one of the first keystroke or the second keystroke is a simultaneous keystroke of the specific key and a key in a predetermined key set of the keyboard section, the mode is a four-keystroke input mode. Based on the combination of the four key codes from the first key press to the fourth key press, a character code indicating one character input by four key presses is generated, and the key code combination of the first key press and the second key press is the same as the four key presses. In the case of a predetermined combination that does not correspond to the key code combination of the first key press and the second key press in the key press input mode, the two key press input mode is determined and one character is input by two key presses based on the combination of two consecutive key codes. processing means for generating a character code indicative of the 2-keystroke input mode or the 4-keystroke input mode, and generates character codes corresponding to the 2-keystroke input mode and the 4-keystroke input mode, respectively. A character input device characterized by being able to. 2. The keyboard section includes a key group consisting of a plurality of keys and a specific key, and the key code generating means includes a first circuit that detects a keystroke on the key group and generates a key code of the pressed key. and a second circuit that detects the pressing of the specific key and generates the key code;
It is determined whether or not a key in the key group and the specific key are pressed at the same time, and if they are pressed simultaneously, the key codes generated by the first and second circuits are blocked and the first and second circuits are pressed. 2. The character input device according to claim 1, further comprising a simultaneous keystroke processing circuit that generates a keycode having a content different from a keycode in the circuit corresponding to a keystroke in the key group. 3. The simultaneous keystroke processing circuit is a circuit that determines simultaneous keystrokes if one of the keys in the key group or the specific key is pressed within a very small predetermined waiting time after the other key is pressed first. A character input device according to claim 2. 4. The character input device according to claim 1, 2, or 3, wherein the specific key is a bar key arranged in front of the key arrangement of the key group in the keyboard section. 5 In the character input device according to claim 1 or 2, where the number of keystrokes required to input one character is different, the keystroke pattern is set so that the key code combinations of the first keystroke and the second keystroke do not overlap. and the processing means determines whether the character input is by two keystrokes or four keystrokes from the combination of the first keystroke and second keystroke keycodes successively generated by the keycode generation means. and a second means for generating a code indicating the input character in response to a combination of 2 or 4 key codes corresponding to the determined number of keystrokes. input device. 6. The first means includes a register means for sequentially storing the key codes successively generated by the key code generating means, and a combination of the number of keystrokes required to input one character and the keycode combination of the first keystroke and the second keystroke. a memory for storing in advance a relationship between the two keystrokes and reading a signal for determining whether a character is input by two keystrokes or four keystrokes from the key code combination of the first keystroke and the second keystroke stored in the register means; and clearing means for clearing the memory of the register means based on at least the last keystroke of the number of keystrokes determined by the memory, and the second means corresponds to each number of keystrokes. and selects only the character code memory corresponding to the number of keystrokes determined by the memory of the first means, and selects the character code memory corresponding to the number of keystrokes determined by the memory of the first means. 6. The character input device according to claim 5, further comprising means for reading a character code corresponding to a key code combination in said register means from said selected character code memory when a key code is stored in said register means. 7. The second means is means for outputting a combination of key codes corresponding to the number of keystrokes determined by the first means as a code indicating one input character. Character input device according to item 5. 8. a keyboard section, a key code generating means for generating a different key code depending on whether each key of the keyboard section is pressed individually or at the same time as a specific key in the keyboard section; A character code is generated based on a combination of 2 to 4 key codes generated by the key code, and the key code corresponding to a predetermined one of the first keystroke or the second keystroke is the combination of the specific key and the keyboard section. If the key is pressed simultaneously with a key in a predetermined key set, it is determined that the mode is 4-key input mode, and 4-key input is performed based on the combination of 4 key codes from the 1st keystroke to the 4th keystroke. A character code indicating a character is generated, and if the key code combination of the first key press and the second key press is a predetermined combination that does not correspond to the key code combination of the first key press and second key press in the 4-key press input mode, 2-key press input is performed. In a character input device comprising a processing means that determines the mode and generates a character code indicating one character input by two keystrokes based on a combination of two consecutive key codes, the kana to be converted into kanji, that is, the conversion When inputting kana, the keystrokes are distinguished from inputting hiragana and katakana, and when a converted kana is inputted, the processing means generates a character code for the converted kana that is distinguished from hiragana and katakana; Furthermore, the character code string generated from the processing means is input, and if the character code string includes a converted kana character code, a combination of the converted kana character code and the converted kana character code or kanji code before and after it is input. Determine words that include a given kanji,
A character input device comprising a kana-kanji automatic conversion unit that outputs a kanji code corresponding to the converted kana character or converted kana character string based on this determination. 9. In the character input device as set forth in claim 8, the kana characters are predetermined 2 characters other than the specific key.
Two keystrokes are input by consecutively pressing one key, and for hiragana, katakana, and converted kana, the same-sounding kana are input using the same keystroke pattern, but the first or second keystroke is the same as the specific key. A character input device in which the above three types of kana input are distinguished depending on whether keys are pressed simultaneously or not. 10 The kana-kanji automatic conversion unit has a memory that stores in advance a word containing a large number of kanji characters corresponding to all the kana readings of the word or a combination of a part of the kana readings and the remaining kanji characters. 8. The character according to claim 8, wherein the converted kana character code appearing in the character code string generated from the processing means and the kanji code before and after the converted kana character code are read out from the memory as address inputs. input device.