JPS5880730A - Inputting of data and control reference signal into memory unit of graphic japanese character input unit - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for storing data and control reference signals in a memory unit of a Chinese character input device. In this case, the kanji input device has an input keyboard for inputting the kanji string forming the kanji or compound word, a display device for actually displaying the entered kanji or compound word, and a central control device. .

How to input Chinese text into a text system (word processor) or printing or recording device:
Already proposed. In one of these methods, kanji or idioms are manually entered using phonetic identifiers representing initials and initials. Homophones 1 (to avoid ambiguity caused by phonetic identifiers of radicals, tone identifiers,
A glyphic identifier, such as a triangular shape, is also input as required. With this proposal, the fixed code required for input in conventional methods does not have to be used. The skill requirements placed on the operator of the input device are therefore considerably reduced.

However, communication is always established between the operator and the input device via the display area 7, which is realized in the form of a multi-word simultaneous display, which makes the use of the device easier, but also because of this. Due to the learning effect, input speed and recording speed are extremely high.

For this type of device, reducing the ambiguity that arises when manually inputting phonetic identifiers is an essential requirement. There are around 10,000 different kanji characters in use in modern China, and these kanji characters can be combined to create around 5,000 characters.
0 word ripeness (7) This is obvious from the fact that words are created. The essential problem is that there are a large number of homonyms, words that sound exactly the same but have different meanings. Normally, radicals, tones, or shapes are used to distinguish these homophones. Radicals are characters that can be said to be etymological, and are often located on the left side or at the top of kanji. Although it can be pronounced as an independent kanji, when it becomes a radical,
It expresses only the ancient kana meaning, regardless of the sound of the kanji to which it belongs. However, part 1 斤 can also form an independent kanji. The second means used to distinguish homonyms is tone. There are four different tones. High and low pitch (1st tone), low pitch (2nd tone), low concave tone (6th tone)
, in a high-descending tone (fourth tone). In the above proposal, the tone identifier is entered via an input keyboard.

The final means of uniquely defining a kanji is its glyph characteristics. A glyph identifier represents information about the structure of a Kanji character. All kanji are made up of 8 brushstrokes, and these strokes, or kuruma, are
It is classified as a middle class and is written in a certain order and in a certain direction. The car-shaped glyphic identifier is represented by a four-digit code. According to the aforementioned proposal, the monomorphic identifier is entered via the numeric keys of the keyboard.

In this proposal, the above identifiers are input together, so only the initial and final identifiers are input? Compared to inputting
Ambiguity is significantly reduced. Another advantage is that all the idioms, that is, the kanji that form the idiom, are input sequentially and sequentially by the initials and finals.

When this proposal is applied to a recording device, the data storage method becomes a major problem because the data group is enormous. According to the basic principle mentioned above, for one kanji, 2 Bytes are required to store the initial and final identifiers, the same < 2 Bytes are required to store the monomorphic identifier, and < 2 Bytes are required to store the initial and final identifiers of radicals. An additional 2 Bytes are required. Therefore, the storage capacity required for 10,000 characters is 60KB.
It becomes yte. ``In addition, 210 Kbytes are required to store the phonetic identifiers of 50,000 idioms.

For example, if a mozuik printing mechanism with a rask consisting of 32×24 rask points is used as the printing mechanism, 960 KByto is required to print all the kanji characters. Therefore, in this type of recording device memory unit, D 1
A memory total of N-1' 1230 K Bytes is required. This not only imposes a considerable cost burden, but also increases the cost required for circuit configuration and has a negative impact on processing speed.

It is an object of the invention to provide a method for storing data and control reference signals. The method should then significantly reduce the demand for storage space, increase the operating speed and in particular considerably shorten the access time. In doing so, the advantages associated with the input of phonetic identifiers must also be fully maintained, such as ease of operation, high input or recording speed, and beautiful recorded fonts.

This problem is solved by the features of patent claim 1.

According to the method of the present invention, ambiguity can be completely removed without using double memory when storing a huge group of data that inevitably appears when processing Chinese characters. At the same time, in a way that increases operating speed, reduces access time, and maintains visually appealing fonts.
Information 7 necessary for printing with the MODEIK printing mechanism can be used.

In the configuration of the present invention, the combination of kanji that forms a compound word is
They are grouped and stored as a combination of address numbers in the first memory. These address number sets are read when the phonetic identifiers of multiple kanji characters forming a compound word are input and compared with the address numbers corresponding to the input identifiers. This makes it possible to select phrases easily and quickly. □ Another construction is based on the idea that all kanji are composed of one or more subpatterns that vary in size and position. In this configuration, information in the third memory is stored, and this information 1 (in relation to the size of the character area provided for display or recording)
Be able to select the required partial pattern. in this case,
It is advantageous to provide a total of 250 partial patterns with a maximum of 6 different sizes, and arrange a maximum of 4 partial patterns at 32 positions 1- to form one Kanji character.

Within the scope of the present invention, the addresses for the fourth memo IJ'& sub-control are divided into blocks corresponding to one segmented area of the character area. A fourth memory stores information for the display or printing mechanism. If this division is performed according to the frequency with which a partial pattern appears in a predetermined segmented area within the character area, the demand for storage space in the fourth memory can be reduced without reducing pixel information that determines display or recording quality. .

Finally, within the scope of the invention, it is possible to assign to every Chinese character stored in the first memory a code combination belonging to one Chinese character, for example a code combination according to a telegraph code.

With this configuration, in the field of teletype or telegraph technology,
For example, the method according to the invention can be used advantageously for so-called code transmission.

Next, the present invention will be described in detail with reference to the drawings.

The embodiment of FIG. 1 has five notes IJ SP i~sp.
5, a control device ST, an input key
Board FiT, display device D1 and printing mechanism DW7% are shown. The necessary control commands and phonetic and glyph identifiers G are entered via the input keyboard ETY. Phonetic identifiers include the initials PV of kanji and radicals,
These are the final PE and tone B. The display device is a multi-word interval display, and displays the most frequently used kanji or compound word corresponding to the input identifier. In the case of compound words, of course the display starts from the first kanji input.

If there are a considerable number of homonyms, they are displayed on the display in order of frequency of use by appropriate control commands. The printing mechanism DW is turned on by another control command to print the kanji or idiom at the first display location on the display. Advantageously, the printing mechanism DW is a monochrome printing mechanism. Text as required
A system can be provided in which the kanji or idioms to be printed can be stored.

A first memory SP1 to a fourth memory IJSP4 are provided in the memory unit. 1st and 2nd
Memo 1JsP1. SF3 is used to select kanji and compound words, and the 6th and 4th memo usp3. sp4 is used to compose kanji or idioms, and to control the display device and printing mechanism. The memory sp5 is a read/write memory and is provided as a temporary storage memory for inputting and outputting data and instructions.

As mentioned above, 10,000 kanji are input using the identifiers PV, PK, B, and G, so 1ooo is stored in the memory SPi.

Identifiers PV, PE, B for all kanji of the character.

G will be remembered. All kanji (identifiers corresponding to 1 kanji) are stored in order of frequency of use in Chinese, and each has a number from 1 to 1ooooot at the address ADH.
It is said as. Each address includes the initial PV, rhyme full
PE, the phonetic identifier representing tone B4 is 2 Bytes,
The phonetic identifier representing the initial and initial of the radical is 2 By
te, and the glyph identifier G of each kanji is also 2 Bytes.
Each is memorized. Each time an identifier is entered, a search process begins in the first memory to find the address where the Chinese character corresponding to this identifier is stored. As the number of input identifiers increases, the search reference signal increases, so ambiguity caused by homophones (homonyms) is avoided to a considerable extent. At least significantly reduced.

Furthermore, this ambiguity can also be avoided by inputting the initials PV and finals PW'4 not in units of characters but in units of phrases. For this purpose, a second memory SP2 is provided. The address ADH is detected in the first memory sp1 through the search process, but the address ADH is detected in the second memory SP2.
is controlled by this address. 2nd memory SP2
Addresses 1 to 10000 store combinations that actually appear of the kanji characters stored in the first memory SP1. All addressable areas of the second memory IJSP2 are divided into four groups 01-G4. Each group stores combinations of the same number of kanji, and phrases each consisting of the same number of kanji.

Group G1 stores combinations of two kanji, or two kanji idioms, group G2 stores combinations of three kanji, group G3 stores four kanji, and group q4 stores five kanji. Kanji combinations are memorized. In each group 01 to G4, combinations of the same number of kanji, that is, compound words, are also arranged in order of frequency of use. In the second memo IJ S P 2, instead of the identifier, a set of address numbers, the first memo IJ S
Only the number combinations of addresses in which each kanji is stored in Pi are stored.

In the second memo IJ S P 2, 2 Bytes are required to memorize a phrase consisting of two kanji, and 5 Bytes are required for a 6-kanji phrase. The information read from the first memory sp1 and the second memory SP2 are compared with each other. In this case, the selected group of address number pairs, defined by the first address found in the first memory and therefore by the first kanji of the idiom, is the second kanji forming the idiom. is compared with the address found in the first memory when inputting the address. The fifth memo IJSP5 is provided for temporary storage necessary for this purpose. In this way, the address, that is, the address number of the kanji characters forming the compound word, is uniquely and quickly detected. For more information on this process,
This will be explained later with reference to the embodiment shown in FIG.

In addition to selecting the desired kanji or idioms, displaying these kanji or idioms on a display device or recording them on a record carrier is also a major problem. The data and information necessary for that purpose are stored in the sixth memory SP3 and the fourth memory SP3.
It is stored in memory SP4. The basic idea is that all kanji can be broken down into partial parts. In other words, one kanji can be thought of as being formed by placing one or more partial batons of different sizes in different positions.

According to the invention, in the sixth memo IJ S P 3:
1-information is stored that indicates the partial button necessary for displaying the kanji and the position of the partial button within the character area. This information then corresponds to the kanji addressed in the first and second memories, and the addresses 1 to 1oooo
It is stored using the number at t. The character area is provided for display or recording. In order to display all the Kanji characters, n partial batons are required, and one Kanji character is formed using up to four partial batons. In this case, it is necessary to provide several partial buttons of different sizes in relation to the size of the character area. As a result of actual use, 2
It has become clear that it is advantageous to provide partial patterns of ~6 different sizes, with an average of 4 types. Therefore, since there are 250 partial patterns / of one type, 021,000 partial patterns can be used to form one kanji.

Therefore, numbers 1 to 100 are stored in the sixth memory SP3.
Up to four partial button numbers TZNg can be stored in each address ADH determined by 00.

This partial pattern number can define one partial pattern 0 out of 1000. Further, the partial pattern can take P different positions within the character area. Therefore, in addition to the partial pattern number TZN, information regarding its position within the character area must also be assigned to each character pattern. This information i
It's called TZL. Information TZ corresponding to P-62 positions
Advantageously, it is stored in the sixth memory SP3 in encoded form as an identification number from 1 to 62. From the above, each address of the third memory SP3 has at least 2 l3yte.

11 Bits are provided for the partial bang number TZN, and 5Bi, t are provided for the partial slam position information TZL. Therefore, when one kanji consists of the maximum four partial patterns, 81 bytes are stored. Information representing the partial baton number TZN and the partial baton position TZL is read out in accordance with the address number of the desired kanji,
It is temporarily stored in memory SP5. At the same time, this information
It is also used to control memory 5P47J. Information for controlling necessary pixels in the display device or printing mechanism DWO is stored in n=1000 addresses of the memory sP4. The information read out from the fourth memo IJ S P 4 is used to display or record the partial pattern forming one kanji character in a manner not shown.

As will be explained in detail later with reference to FIGS. 6 and 5, the information stored in the fourth memo IJ SP 4 t(C) relates to all character areas. Using information TZL indicating the position of the partial pattern, it is possible to shift to the reference point of the partial pattern area.

Next, the method of the present invention will be explained in detail with reference to FIG. Figure 2 shows two kanji “1ha011”
A case is shown in which an idiom consisting of rθn'' is input.

First, the first kanji's voice initial pv ni ri and rhyme final P W niao
is input to the first memory SPi. In the first memory sp1, the Kanji characters corresponding to these phonetic identifiers are stored in order of frequency of use, and a search process is started to detect these Kanji characters. The first input identifier (pv=h,
PE, ==ao) corresponds to 28 kanji.

In other words, there are 28 homonyms. Therefore, 28 addresses are output as address "numbers. Since each kanji character is arranged in order of frequency of use, these address numbers are read out as a number string such as 99,150°348, etc., and stored in memory sp5. Temporarily memorized.6th memo IJ
SF 3 is controlled by this address number, and the information stored therein representing the partial button number and partial button position is read out. Further, using the information stored in the fourth memory, kanji characters are formed by combining partial patterns. These pieces of information are input to the display device DvC, where C is the one to five most frequently used Kanji characters, that is, Kanji characters corresponding to Kanji numbers 99, 150, 348, etc., are displayed.

If you input the second kanji after inputting the first kanji,
In other words, in the example in Figure 2, the initial PV2r and the final P F2
= en If Y is input, the corresponding search process begins in the first memory spI.

In this case, 26 homophones corresponding to the input identifier are detected, and their addresses are read out in order of frequency of use, that is, as address numbers 13,800°, 1006, etc. This address number is also temporarily stored in the fifth memory SP5. At the same time, in the second memo IJ S P 2, the first address of the first kanji character "hao", the shuffle address number 99, controls the corresponding address area. In this case, since the input phrase consists of two characters, group G1 is selected. The address number set stored in group G1 is compared with the address number set temporarily stored in the fifth memory, in this example 99/13. In this example, as a result of the comparison, the 6th most frequently used address number set starting with the kanji character "hao" is detected as uniquely corresponding to the input phrase "haoren". Ru.

At this time, the 6th memo! Addresses ADH for controlling JSP3'a', that is, addresses 99 and 16, are also uniquely determined. The 6th memory sp3 contains the kanji “ha, o”
Information representing the structure of ``'' and °'ren''' is stored at addresses 99 and 13. One of these days +
The information corresponding to haOn is two partial buttons YEN with partial button numbers TzNo and TzNq, and their position information is TZLp and TZLr, respectively.

For example, the partial baton numbers for Kanji IIhaoI+ are 55 and 4.
1, and the position numbers are 1 and 7. In this example, the second kanji "r e n" consists of one partial bang, its number is TZ Nn and its position is TzLm. Information DIi to DI3 representing each partial button is stored in the fourth memory S.
It is found at the corresponding addresses TZNo, TZNq, and TzNn in PJ. This information is stored in the memory SP
At the same time, the display device or printing mechanism D
It is also sent to W. The printing process is started by a special control command, a print command.

Although the above explanation relates to a two-character compound word j7-), a compound word consisting of six to five letters is processed in the same way. Even in that case, note 1JsP corresponds to the input identifier.
Idioms detected in i.

As a set of address numbers, only the set of address numbers stored at the corresponding address in the memory SP2 is compared. It goes without saying that this address number set belongs to a predetermined group within the address, corresponding to the number of kanji characters forming the idiom.

The information stored in the sixth memory represents the partial stamps necessary to form a certain kanji and their position within the character area. As shown in FIG. 6, this information also represents a partial button area provided for display or recording, that is, an area having a size of 2%, 2%, 2% of the character area.

Of course, it is also possible to provide the partial batten area 2 with a different size. The partial button area shown here is just one example of dividing a character area.

In addition to the partial baton number, the information in the memo IJ SF 3 also represents the position where the partial baton should be placed within the character area when forming a kanji. According to the present invention, if the position of the partial button is related to a predetermined reference point within the character area, partial buttons of various sizes can be addressed in the same way as for the complete character area. I can do it.

The example in FIG. 4 shows 32 different positions that a partial button can take within a character area ZF constructed on a raster consisting of 24×32 raster points or pixels. A text area of this type can be realized, for example, by a mosaic printing mechanism or a corresponding display.

In the embodiment of the fourth section, 62 position points are distributed from L1 to L32, and among them, for example, the position point L1' and the reference point can be used.

As shown in FIG. 5, a kanji character consisting of up to four partial bangs can be displayed according to the above-mentioned principle of distribution of partial bang positions. FIG. 5a shows four partial flaps Tz1 to TZ4 of different sizes. The partial button TZi is set as shown in FIG.
It is uniquely defined in the format shown in . Similarly, other parts of butter T゛z2. TZ3. TZ4 also has a partial baton number TzN
2 and TZN 3. In this case, the shapes of the partial flaps TZ3 and TZ4 are the same.

In addition to the partial pattern number TZN, the sixth memory sp3 also stores information TZL representing the positions occupied by the partial patterns Tz1 to TZ4 in the character area. This information TZ
When displayed or recorded, the partial patterns TZ1 to 'rz4 are brought together in the manner shown in FIG. 5 to form one kanji. For that purpose, the 6th memory SP
3, information TZL = 1 is stored in addition to the partial pattern number - bow 4ZNl for the partial baton TZi, position information TZL 'l = 17 for the partial baton TZ2, and position information for the partial baton TZ3. TZL 3 = 22 belongs to the partial button TZ4, and position information TZL = 26 belongs to the partial button TZ4.

This information is located at the positions Ll, L17°L22 shown in Figure 5.
．． Represents L26. Also, like the information for point-like display or point-like recording from the fourth memory sp4, it is supplied to the display device and the printing mechanism DW via the memory SP5'i. For example, when controlling the display or printing elements of a mozuiku printing mechanism in the form of a raster or a trench matrix,
For display or recording purposes, each partial button can be assembled onto a character area in the manner shown in FIG. 5 to form a single character. If the partial butterfly 'rz1y is placed at the position point L1, the partial butterflies 7TZ'l, 'rz3. TZ4 is the position point L17°L22. Allocated to L26.

As described above, the information DI is stored in the fourth memory SP4 as so-called pixel information.

With this information, the display element of the display device or the printing element of the printing mechanism DW is controlled in order to reproduce the partial stamp. Each partial button has a size and size related to the entire character area.
Accordingly, each exists only in one part of the character area. Therefore, in an advantageous embodiment, it is proposed to divide the addresses for controlling the individual display or egg elements into blocks corresponding to parts of the character area. This is done by dividing the character area into six to four equally sized segmented areas. 4 of the same size
One of the two segmented areas corresponds to one quadrant of the character area.

According to the research results, approximately 20% of partial slams exist across all four quadrants (that is, the entire character area), and approximately 60% of
% is in the two quadrants on the left, approximately 60% is in the two quadrants on the three sides, and the remaining 20% is in just one quadrant. Therefore, the above numbers (1000 partial bangs, 24X32
Character area consisting of raster points = 96 Bytes) Y
Based on this, the storage capacity of the fourth memory sp4 is 96K.
53K instead of Byte('1000X96)
Become a Flyte. Therefore, the fourth memo IJ S P 4
About half of the storage space can be saved.

The following data is temporarily stored in the fifth memo IJSP5. This means that all control commands entered via the input keyboard ETA since the last print command, the 10 most frequently used kanji (this includes all identifiers and address numbers) of idioms of up to 5 kanji,゛are),
An address number string representing the 10 most frequently used phrases, a partial button for up to 5 kanji characters, and information DI necessary to control the display device or printing mechanism DW.
, is. In 95% of cases, the time required from inputting one identifier by operating a key on the keyboard until one kanji is displayed on the display device is 5 m sec.
shorter. In the case of inputting kanji that are used very rarely, this time is approximately 100 mθθC. However, even in this case, there is no appreciable impairment in input speed or recording speed.

[Brief explanation of drawings]

FIG. 1 is a block diagram of a memory unit used in implementing the method according to the invention, and FIG. 2 is a block diagram of the memory unit of FIG.
Figure 3 is a diagram showing the details of the unit, Figure 3 is a diagram showing the size of the partial batons that make up the kanji, Figure 4 is a distribution map of the position points of the partial batons, and Figures 5 a and b are composed of four partial batons. This is a diagram showing the Kanji character. . SP1 to SP4...memory, SF3...temporary memory, 8T/control device, ET...human keyboard, D/
・Display device, DW ・Printing mechanism, PV...Voice initial, PE
... Finale, B... Tone, G... Glyph identifier, TZ
i~TZ4...Partial button, TZN...Partial button number, TZL...Partial button position, DI...Display
Print element control information. FIG 5 -180- 21koz41 2N3 礪

Claims (1)

[Claims] 1. A kanji input device having an input keyboard for inputting a string of kanji forming a kanji or compound word, a display device for actually displaying the input kanji or compound word, and a central control device. In the method of storing data and control reference signals in the memory unit of the device, in the first memory (SPI) of the memory unit, the phonetic identifiers of Chinese characters, i.e. the initials ( pv)
, final (pg), and tone (B) are stored, and a search is performed in the first memory (SPI) corresponding to the input identifiers (PV, PJ B, G), and at that time, The address (ADH) of the detected kanji is made to correspond to the address number (1 to 104), and furthermore, from this address (1 to 10'), the second memory (SF2) and the sixth memory ( SF3) is controlled 1-, and the second memory (SF3) is configured to control at least two
The phrases are arranged in order of frequency of use of the first kanji, and the phrases in each addressable area in the second memory (8P2) are memorized.
It is divided into groups (01 to G4) according to the number of characters, and in these groups CG1 to G4), phrases consisting of 2 to 4 kanji characters are stored in a 1/lc arrangement in order of frequency of use. The sixth memory (SF3), the first memory (S
Address (1 to 1o) corresponding to the address number of PI)
In 4), information (TZN. TZL) regarding the partial stamps (TZ) forming a kanji is stored, and information (TZN, TZL) in the sixth memory (SF3) is stored in the fourth memory (SF3). )Y control and the fourth memory (SF3
) to the display device (D) or output printing device (DW)'& controlling information (DI)? A method of inputting data and control reference signals into a memory unit of a Chinese character input device, the method comprising: storing data and control reference signals in a memory unit of a Chinese character input device. 2. The method according to claim 1, wherein code combinations belonging to each kanji and code combinations according to teletype or telegraph codes are stored as identifiers in the first memo IJ (SPI). . 6. Kanji group (Gl~) in second memory (8P2)
The combination of a4) is stored as a combination of numbers (1 to 1000 ()) forming the address of the first memo IJ (SPI), and is also assigned to the first character of one group (01 to G4). After addressing one of the areas in the second memory (SPI) by the address (e.g. 99),
Read out a combination of one kanji group (01-G4) and assign this combination to another identifier (PV, PB, B, G,
, l), the first memory (SPl)
(6) Compare one after another with the address of another kanji (for example, 99/13) detected in 1. , and if a positive ratio 1 bite result is obtained, the combination detected at that time (e.g. 99/13)
is used to address the sixth memory (SF3) and the information (TZN ) read from the addressed area (e.g. 99.13) of the third memory (SF3)
'a'. Used for addressing the fourth memory (SF3),
Through this information (TZN) 7<, the information (DI) 'aj
3. The method according to claim 1, wherein the image is sent to a display device (-D) or a printing mechanism (DW). 4. The first memory (f9P) in the third memory (SF3)
1) A partial button (T
first information (TZ) for determining at least one of
N) and partial slam (TZ) in the character area (ZF) provided in the display device (D) or printing mechanism (DW).
7. A method according to any one of claims 1 to 6, characterized in that second information (TZL) (4) for determining the position of is stored. 5. First information (TZN) in the sixth memory (SF3)
), select one of n different partial buttons and select 1
5. The method according to claim 4, wherein the second information (TZL) causes selection of one of p different positions within the character area. 6. In order to display all kanji characters, n = 1000 partial buttons should be provided, and in order to display only one kanji character, a maximum of 4 buttons should be placed at P232 positions within the character area. Claim 4 with two partial batons (TZ)
or the method described in paragraph 5. Store the data (one out of n, one out of p) for displaying one character consisting of Z partial stamp (TZ) in the third memo IJ (SF3) as an identification number,
This identification number is used as an address for the fourth memory (SF3), and the display device (D, ) and printing mechanism (DW) display and print information (DI ) The method according to any one of claims 1 to 6, wherein the method stores: 8. Divide the address for controlling the fourth memory (SF3) into address blocks corresponding to the segmented areas of the character area, and at this time, how many partial buttons can be accommodated in a predetermined segmented area. 8. The method according to claim 7, wherein the division is performed according to the following.