CN1031435A - Morph compiling method and input keyboard thereof - Google Patents

Abstract

A kind of morph compiling method and input keyboard thereof of Chinese character.From the graphic characteristics of Chinese character, concluded and formulated the morph (font key element) of six kinds of stroke points of expression (do not converge, turn round, trident, four forks, five forks, six forks) and quantity and distribution.Utilize that morph design keyboard and coding have that keyboard disk is succinctly eye-catching, coding rule is simple and easy, natural, clear and definite, advantage such as memory capacitance is little.Can be widely used in aspects such as Chinese Computer, chicoder, computer telephone, phone mechanized directory assistance platform, Chinese character translator, retrieval.

Description

Morph compiling method and input keyboard thereof

The present invention relates to the method for Chinese character coding and input keyboard that computer Chinese information is handled, and relate to Chinese character search method.

But encode Chinese characters for computer partial, shape two classes, and wherein font code can be divided into two kinds of stroke class and parts classes.The former can be representative with the patent (GB2100899A) of Li Jinkai, Chinese character is decomposed into stroke sequence encodes, and has the long shortcoming of decomposable process.The latter can be representative with the patent (CN85105556A) of Chen Aiwen, font is decomposed into the parts sequence encodes, and exists because the numerous all difficulties brought of Hanzi component.Stroke and parts that the four-corner number code method adopts Chinese character quadrangular to occur are encoded, owing to all many difficulties of repeated code that exist that can not take Chinese character into account.

It is few and can fully, directly express the coding method of each characteristic of Chinese character to the purpose of this invention is to provide a kind of coded identification.Such method must have advantages such as memory capacitance is little, rule is simple, repeated code is few, stroke is few.

Key problem in technology of the present invention has been to find morph (font key element) that is independent of outside Chinese character stroke and the parts, that can reflect the Chinese character image characteristics, as to represent the stroke point kind, quantity and distribution and the method for extracting morph from Chinese character.

Usually in the process of computing machine key entry contribution, comprised the reading process of operator the operator to original copy.This is a visual identification process, and what this process was noted is the Chinese character image characteristics, seldom considers the detail of stroke, parts, and its velocity ratio thinks over the stroke component detail when writing faster.As seen the starting point of coding method should be placed on the kind, quantity, distribution of the graphic characteristics of Chinese character rather than stroke and parts.Cataloged procedure should not be the simulation of writing process.The situation that it is considered herein that the stroke point is the importance that we discern the Chinese character image characteristics.As " on " word, its feature be two stroke points " " Yu “ Shang " and distribution up and down.The figure that meets these characteristics: go up, All take as " on " word.The influence of stroke length is only worked to the identification (as " soil " and " scholar " " end " and " not ") of indivedual Chinese characters.The inclination angle of stroke is also only to the identification of individual malapropism influential (as the king and the ninth of the ten Heavenly Stems).After abandoning the less feature of these influences, we can be divided into the stroke point six kinds: turn round , Contraband, factory, second, mouth etc.;

Trident is gone into, Ren, worker, end, just wait;

Four fork Shi, Qe, Lv, thirty, well etc.;

Five forks are big, just etc.;

Six fork wood etc.;

Do not converge two, three, eight, river, six etc.;

About the distribution of point can have, about, surround three kinds: ◎.

The kind of point, orientation, quantity and the distribution in Chinese character thereof have determined Hanzi features basically, and we claim that they are the key element of font, are called for short morph.The different azimuth ， of point such as Shang

Certainly also be a key element, but in order to reduce the symbol kind and to improve coding rate, let us abandon the orientation.Only consider kind, quantity and distribution.Kind and quantity are combined also symbolically, obtain the listed morph of table 1.

Table 1 morph table

^*Standby morph, expression " tiltedly " respectively, " to " the trident point.

Wherein " ,=, ≡, " represent not point, quantity 1 ... more than 4.“

" expression trident point, quantity is from 1 ... more than 4." ten ...

" expression four fork points, quantity is from 1 ... more than 4." , Jiong " being the point that turns round, quantity is from 1 ... more than 4.Big and wood is five fork and six forks, and quantity is 1.When the key position was not enough, " wood " was available " greatly, " or " ten, " expression, " greatly " available " ten, " expression." mouth " expression closed four is turned round.

With these morphs, the keyboard rationalizing principle that proposes by DVORAK is arranged on 26 letters cases of QWERTY keyboard and has just obtained the morph keyboard.Fig. 1 is a kind of example of morph keyboard.Wherein omitted morph " wood ", vowel (A, E, I, O, U) key is not arranged morph.Arranged shape slightly key ▲, be used to import shape abbreviation sign indicating number (shape is sign indicating number slightly).Delete keyboard (DEL) in addition and under the Chinese character state, be divided into lower gear: (delete a left side) before deleting and perform: delete back (deleting the right side).The △ key is a slightly sign indicating number key of phonetic, is used for slightly sign indicating number of input Pinyin.

Utilize the encode concrete grammar of (promptly be coding unit) of morph different, multiple scheme can be arranged with purposes with the morph.Utilize the scheme of whole or big portion morph to claim first kind scheme.The scheme of only utilizing a part of morph is the second class scheme.

First kind compiling method example

1. single character code method.Be used for word for word importing national standard " message exchange Chinese character coding set baseset GB2312-80 " to computing machine, coding rule is:

One, the number joint account of the similar point that occurs continuously along the stroke that links to each other and represent with a morph.

Two, not meeting point joint account in groups, the point joint account that is symmetrically distributed in both sides is also with a morph representative.

Three, point does not repeat to count, and promptly no longer considers in the code fetch of point in the back that had calculated in last sign indicating number.

Four, radical " Rolling " with morph "

" representative.

Five, the place is collided not as linking to each other with other stroke in two ends and short tip of casting aside, and thinks to link to each other when point is collided with other stroke with short middle part of casting aside, and makes point and considers.

Six, colliding between the parts (member that in Chinese character, occurs) with relatively-stationary form not as linking to each other.Not making point considers.

Seven, sign indicating number position and order are that 1. " in going up to a left side " gets one yard, and 2. " under the left avertence " gets one yard, and 3. " on the right avertence " gets one yard, 4. " on right time " get one yard, 5. " down to right " get one yard.Have 2. and no longer get 4., do not have and 2. just get 4..

Eight, get foot and do not need to add in addition end key (also not adding space bar) for 4 yards, less than 4 code words are pressed font characteristics form slection element

, one of ◎ makes end key.Choosing when the font characteristics are not obvious

Get final product.

The reason of above-mentioned rule is proposed:

1. the similar point quantity joint account that occurs continuously can increase quantity of information, and makes the usage frequency of each morph even, improves code efficiency.

2. the purpose of not double counting of point is to avoid reducing quantity of information, reduces the uncertainty of coding.

Since the Chinese character that contains radical " Rolling " much and always appear on the first sign indicating number of coding, so select for use one do not appear at first yard morph " " represent radical " Rolling ", make a keystroke input of this radical.

Point, short cast aside and each parts of Chinese character between no matter in clerical type the situation of colliding with other parts often appears in block letter also, seemingly connect like do not connect, time or not when connecting.In order to reduce this uncertain factor, regulation is done without exception not link to each other and is handled.

5. the code fetch order must be looked after the convenience of various structure Chinese character code fetches with effective.Order in the rule is comprising priority level, as the 1. sign indicating number " in going up to a left side " serve as preferential with the point of " in going up " portion, then get the point of left as if the point at supreme middle part.And for example the 2. sign indicating number " under the left avertence " be preferential with a left side, gets the point than the below equally under the situation on " left side ".The 2. sign indicating number can not get " in down " direction.As not having point in " under the left avertence " direction, this sign indicating number is not just got.The 5. " down to right " of sign indicating number serve as preferential with " down " position, " under in " no point sees that again " bottom right " direction has or not point.

6. coding does not adopt space bar to make end key not only to reduce stroke, and makes this coding mix use with the phonetic sign indicating number, does not obscure mutually.When 4 yards of less thaies, use

/ ◎ key is made end key can replenish font information.

Sound, shape two classes coding respectively has characteristics, if can use by double track, both brought into play the strong point that phonetic sign indicating number speed is fast, be convenient to dictate and draft manuscript, has brought into play the advantage that font code can be imported rare characters again.The phonetic sign indicating number can directly be imported with the keyboard of standard based on speech.Each vocabulary finishes with space bar.Therefore which kind of sign indicating number what the rule that occurs according to space bar and vowel of computing machine can be judged input is.When input code meets the pinyin syllable characteristics and space bar is arranged is the phonetic sign indicating number, is the morph sign indicating number when input code does not meet pinyin syllable and when not having space bar.

To " ", "Yes" ... etc. tens high words of frequency of utilization, can stipulate brevity code, add with a font code Or ◎) two keys input.

The disposal route of repeated code:

When computing machine transfers to when finding to have repeated code in the process of internal code (or information exchange code) at input code, remind the operator to note the repeated code word of pointing out on the display with sound and light alarm.The display below is a preface with the usage frequency of each repeated code word when above at one when the repeated code word, lists the repeated code word and indicates sequence number.The operator can key in corresponding sequence number and select a wherein word.If the back is not with numeral, computer understanding is for selecting its 1st word.

Table 2 has provided the example of first kind single character code.

2. the shape abbreviation is encoded.Be used for to the commonly used or specific vocabulary of computing machine input, short sentence so that the coding of short essay.Its compiling method is slightly sign indicating numbers of shape for hitting follow-up 4 morphs of shape speech key ▲ expression earlier, chooses 4 morphs according to the number of vocabulary or key word then.To two-character word, two morphs of initial and end got in each word, totally 4 morphs (mend when not enough with

/ ◎).Three words, first morph got in first and second word, and two morphs of head and the tail got in the 3rd word.Four words, first morph got in every word.Short sentence and short essay are got 4 key words earlier, and then get its morph.For example " adhere to the Four Cardinal Principles ", get key word and " adhere to four fundamental rules ", its code be " ▲=

Mouth Jiong ".Allow the new shape of the definition of performing of user's usefulness ▲ key to omit sign indicating number: the shape of the whole single character codes ▲ correspondence of ▲ words and phrases that (performing) is defined is sign indicating number ▲ (performing) slightly.

3. extra-set character compiling method.Be used for the Chinese character beyond computing machine input national standard " GB2312-80 " baseset.

Hanzi component contained in the extra-set character can select digital watch to go out with its head and the tail morph and repeated code.The morph code of common components sees Table 3.But listed file names with the phonetic sign indicating number that double track uses in the table.

The mutual relationship of Hanzi component in Chinese character can be used morph

, ◎ shows out.At this moment

, ◎ is called constructor, is used for representing the structure of Chinese character.The structure of Chinese character is stratified, is combined into by " standing grain " and " being " as " journey " word, and " being " is combined into " king " by " mouth " again.According to this hierarchical relationship, when being encoded, extra-set character should extra-set character be decomposed into the sequence of parts and constructor composition by following rule:

1. (or back) keys in a constructor that shows these parts and remaining part structural relation before the code of each parts;

Two low level parts (i.e. low level zoarium) before (or back) also need key in a constructor that shows this zoarium and remainder structure;

In coding the order of parts corresponding to these parts the position earlier left back right side, low, order of determining of the order in the outer back earlier in high back earlier successively in Chinese character.

Such sequence has uniqueness.For example form four words that font is different by " first, second, third " three parts, their coding has nothing in common with each other:

In order to represent the beginning and the end of extra-set character coding, first and the last constructive code of extra-set character coding should be " capitalization " (promptly pressing shift simultaneously).

The great advantage that adopts such coding is that it can instruct computing machine to generate the Chinese character dot matrix of extra-set character with the mode of font assembly unit.As long as computer-internal has the font dot matrix of Hanzi component, can determine the assembly unit order according to following assembly unit rule so:

1. two parts elder generations amalgamation that preposition (or back with) constructor is identical;

2. preposition (or back with) constructor two parts inequality wouldn't amalgamation;

3. the parts of amalgamation are participated in amalgamation as new parts.

Receive one group of extra-set character coding as computing machine

First The third, wherein second is different with third preposition structure symbol wouldn't amalgamation.First is identical with the preposition structure symbol of second, is " left and right sides structure ", carries out assembly unit so get the dot matrix of two parts of first, second with left and right sides structure, and the ratio of assembly unit can be selected according to the uniform principle of font by computing machine.(seeing the 5th academic meeting paper of Lin Jingyu " the automatic generation of Chinese character assembly unit ratio " Chinese character information processing research association of China Instruments ﹠ Meters Society), the parts that assembly unit is good " first and second " have before

, with " third " before Identical, so both carry out the assembly unit up and down of dot matrix, obtain at last ^{First and second} _ThirdThe font dot matrix.(see for details and contain " the descriptor code facture of extra-set character " the Chinese information processing international conference collection of thesis of remonstrating with, Beijing 1987.8)

Second class coding example

Utilizing 10 numerals to be coded in has Special Significance (as the dialing coding) on some equipment.Therefore it is corresponding with 10 numerals to choose 10 morphs from 24 morphs: (" mouth " and " " merge into " mouth ")

10 morphs corresponding of table 4 with numeral

These morphs in that some is got in touch on the image or on the implication, therefore are not difficult to remember their corresponding relation with digital.Certainly these morphs are engraved in around the dial (of a telephone) or keyboard on better, can exempt from memory.

Second class coding also can divide single character code, shape, and slightly coding, extra-set character are encoded 3 kinds.

1. single character code rule:

One, along the stroke that the links to each other number joint account of the similar point of appearance continuously, but the maximal value of joint account is no more than the maximal value of such point in selected 10 morphs, and promptly " ten " are no more than 1, " " be no more than 2, "

" be no more than 4, " one " is no more than 3.With a morph representative.

Two, can divide two place's code fetches during greater than maximal value when the joint account value of point number.Number is divided equally during for even number, and first place was littler by 1 than second place when number was odd number.In addition " wood " is decomposed into " ten, ", " greatly " be decomposed into " ten,

".

Three, in maximal value, no point in groups merges counting, and the point that is symmetrically distributed in both sides merges counting and represents with-morph.

Four, collide not regard as with other stroke and link to each other in two ends, short tip of casting aside.

Five, collide between the parts and do not make point and treat.

Six, sign indicating number position and order thereof are: in 1. going up to a left side, 2. under the left avertence, 3. on the right avertence, on 4. right time, 5. under in, 6. bottom right.Have 2. and have 5. and do not get 6., do not have and 2. get 6..

Seven, get foot and do not add end key or space bar in addition for 5 yards, less than is pressed font characteristics form slection element for 5 yards

One of ◎ is as end key, and 1. choosing got final product when the font characteristics were not obvious.

Above rule is similar to first kind single character code, has just done corresponding adjustment owing to the morph kind reduces, and code length increases by 1.

Table 5 is second class single character code examples.

2. shape abbreviation compiling method

The shape of going into to import usefulness as the Chinese of computing machine is sign indicating number slightly, and the shape of its rule and the first kind is slightly yard identical.

Slightly the shape of sign indicating number and the first kind is slightly yard different as the shape of telephone number query and conversion usefulness: can not use

, the ◎ key, and key word choose relevant with the characteristics of phone name in an account book.

Omit sign indicating number as the shape corresponding with telephone number, code length should slightly be longer than telephone number, to reduce repeated code.Generally can extend 1.For example phone number is 5, then the shape used of directory enquiry slightly sign indicating number get 6.Phone number is 6, and then corresponding shape slightly sign indicating number is got 7.When shape when slightly sign indicating number is got 6 key word choose and encode regular as follows:

One, the address of the special use in name in an account book part is got wherein two key words the following person of four words, and two yards of head and the tail (position of taking back in promptly going up is got a yard, and the position that takes over is down got a yard), totally four yards got in every word.

Two, the address of the special use in name in an account book part is got wherein four key words the above person of four words, and first sign indicating number, totally four yards got in every word.

Three, each word was as key word before and after the general address part in the name in an account book was generally got, and first sign indicating number, totally two yards got in every word.

Four, the special use of street name address is got both keyword, and common segment was that a word person (street, road) gets two yards of head and the tail originally, is the first sign indicating number that two word persons (as alleyway, Zhong Lu, South Road) get two words.

Five, the number of back, street or do number can directly to follow back at the street coding.

Six, the inquiry of when telephone set or query station find repeated code is arranged, will reporting to the police, the user dials in next yard position of individual calling again, up to telephone set or query station do not transmit messages warn when inquiring till.

Seven, the street place name that has also has family number in number or after doing number, can dial in family number after hearing the inquiry number of reporting to the police again.

Example:

The following No. 356 Rui 亅 in Hai Yueqiao river

Rui Xia

379635356

The Fang Yin Shi  Lv Tou of sales department of power office

889842

The headmaster's office Rui of Zhejiang University  Ren Shi 

364544

The Ba Ren Ha Shi  Dian of Party School Office of Provincial Party committee

252441

Huai-Hai Zhong Lu 1211 lanes No. 52 Rui Shang Rui 亅 Mouthful

3537801211

The shape that table 6 has provided department store correspondence on the telephone directory of Hangzhou is omited the sign indicating number example.From then on example is found out, the repetition rate of coding quite low (0%) and coding method are simple and easy, easy to learn.

This phone omits the user self-help directory enquiry that sign indicating number can be used for computer telephone and telephone information position with shape.The former is solidificated in the table of comparisons of shape summary sign indicating number with phone number in the telephone set, and the user only need remember coding rule, dials in coding, and telephone set will send corresponding telephone number from trend telephone office.The user can memorize mechanically telephone number.The latter is the computing machine that the user self-help directory enquiry is installed at telephone office's information desk, the user omits code dialing to information desk with shape, the computing machine of information desk is converted to corresponding phone number with this yard, and by computer speech synthesizer answer name in an account book and phone number, the user praises, can by dial 0 please self-help number-searching computing machine be transferred to want on the phone.

3. extra-set character compiling method, the extra-set character compiling method of the situation and the first kind is similar, different persons, only the morph that comprises in the morph code of Hanzi component is limited to 10 listed morphs of table 4.

Positive effect of the present invention is:

1. adopt the morph that can reflect the Chinese character image characteristics be independent of outside Chinese character stroke and the parts to encode, thoroughly avoided the loaded down with trivial details or memory capacitance of stroke or component coding big wait difficult.And people that can Chinese character also can carry out encode Chinese characters for computer.

2. morph quantity is few, can be arranged in easily on keyboard for typewriting, the telephone keypad (or dial).The key face is succinctly eye-catching; Without vowel key and space bar, realized shape sound doubletracks system.

3. coding rule is simple and easy, natural, clear and definite.Without exception.

4. purposes is wide.Not only can be used for general computer Chinese information and handle, can also be used for directory information, retrieval, dictionary etc.

Table 2 first kind single character code for example

The shape sound code example of table 3 common components

The table 5 second class single character code for example

Table 6 phone shape is sign indicating number example (the department store coding in Hangzhou) slightly

The department store, high point

The Zhong Bei department store

The halfway up the hill department store

Department store, blunt river

The new year the department store

New sunshine department store

The innovation department store

Department store forward

The Hang Shi department store

The Xin Feng department store

The Huan Dong department store

To the group department store

The logical department store of China

The Xing Chang department store

The SOGO shop

Annotate: numeral is omited sign indicating number rather than telephone number itself corresponding to shape in the table.

Claims (8)

1, a kind of morph coding method, all or part of of morph (font key element) of kind, quantity and distribution that it is characterized in that adopting six kinds of stroke points of expression (do not converge, turn round, trident, four forks, five forks, six forks) is as coding unit.

2, by the coding method of claim 1 regulation, it is characterized in that representing morph with following symbol:

3,, it is characterized in that selecting for use following morph corresponding with 10 numerals by the coding method of claim 1 regulation:

4,, it is characterized in that along the number joint account of the continuous similar point that occurs of the stroke that links to each other, with a morph representative by the coding method of claim 1 regulation.

5,, it is characterized in that morph " wood " can be expressed as " greatly, by the coding method of claim 1 regulation " reach " ten,

", morph " greatly " can be expressed as " ten, ".

6, by the coding method of claim 1 regulation, it is characterized in that collecting outer Chinese character and use shape

And the sequence that component code is formed is encoded.

7, a kind of keyboard that is used for above-mentioned coding method is characterized in that having arranged the morph symbol on the consonant key position of QWERTY keyboard.

8, a kind of phone dial that adopts above-mentioned coding method is characterized in that digital corresponding with 10 morph symbols and phone dial.