JPH0628059B2 - Character processor - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial application] The present invention relates to a character processing device for inputting a kanji / kana mixed sentence by kana / kanji conversion.

[Prior Art] Currently, a character processing device such as a Japanese word processor generally inputs kanji-kana mixed sentences using kana-kanji conversion.

Conventionally, the kana-kanji conversion dictionary was often stored in an external memory (floppy disk, hard disk).
However, due to improvements in conversion methods such as batch conversion, the number of dictionary accesses has increased. The method of storing in the internal memory is the mainstream.

The internal memory is readable / writable RAM that is volatile (the contents are lost when the power is turned off), and non-writable but non-volatile (the contents are erased even when the power is turned off) There are two types of ROMs that are commonly used.

If the dictionary is stored in RAM, the dictionary will be loaded from external memory to RAM as preparation before Kana-Kanji conversion, but it has the drawback of requiring a long loading time.
Since AM is more expensive than ROM, it also has the disadvantage of higher cost. Therefore, the dictionary is generally stored in the ROM.

In addition, recently, the demand for the conversion rate of Kana-Kanji conversion has become more sophisticated, so that the dictionary tends to have a large capacity.

Since the dictionary has a large capacity in this way, for an individual person, a large number of words that will never be used are stored in the dictionary. It is unavoidable that such a useless word is useless and wasteful of memory, but it is also a cause of erroneous conversion, and some words are more than appreciated by some operators.

For example, as a result of the increase in the number of words, it is assumed that the surname "Kobe" is stored in the dictionary and the name "Ko" is also stored in the dictionary. If an operator tries to convert "Kobe Port" with such a dictionary and then types "Kobeko", the Kana-Kanji conversion is usually "Last name + First name".
Since the process of converting the pattern of No. 1 with priority is incorporated, there is a high possibility that it will be converted to "Kobe Isao" against the operator's intention.

In such a case, if "Gong" is changed to "Port", learning will work and "Kobe Port" will be correctly converted from the next time. However, as long as the surname "Kobe" exists in the dictionary, the intention of "Kobe city" becomes "Mr. Kobe", the intention of "Kobe product" becomes "Mr. Kobe", and misconversions occur here and there. ,
It impairs operability.

In the first place, the surname "Kobe" is a familiar surname for ordinary people, and the solution of deleting it from the dictionary is the most effective because it has no trouble. However, the dictionary is usually ROM
Since it exists above, it is not possible to delete words directly.

As a method of deleting a word on the ROM, a method of storing the existing address of the word on the dictionary ROM and ignoring the word at that position during conversion and not using it, reading of the word to be deleted, notation, part of speech, etc. A method is conceivable in which word information is stored, and if a word that matches the word information is present in the dictionary at the time of conversion, the word is ignored and not used.

[Problems to be Solved by the Invention] However, deleting a word by the above-described method of storing word information requires a large amount of memory each time one word is deleted, resulting in a high cost, and at the time of conversion. Matching with word information has the drawbacks that the processing becomes complicated and the conversion time becomes long.

In addition, when deleting words by the method of storing addresses,
The deleted word dictionary becomes unusable when the contents of the dictionary are updated. Therefore, an operator who uses a plurality of types of machines must create a deleted word dictionary for the machine language, which is problematic in terms of data compatibility.

A method of storing both the address and the word information is also conceivable, but at that time, there is a drawback that the specification memory becomes considerably large and the cost is too high.

[Means (and Action) for Solving Problems] In order to solve the above-mentioned problems, the character processing device of the present invention has an input means for inputting a kana character string, a word reading, and a notation. The first non-rewritable dictionary means storing the dictionary identification information for identifying the dictionary and the first dictionary means while referring to the first dictionary means. And a conversion means for converting the kana character string input by the input means into a notation of a word in which the kana character string is read, and a word that should be invalidated at the time of conversion by the conversion means, cannot be rewritten. A first pointer which is a pointer in a valid dictionary and dictionary identification information of the non-rewritable dictionary, and reading of the invalid word and word information of the invalid word including notation Corresponds to and And a second pointer which is a pointer in the first dictionary means of a word to be invalidated in the conversion by the conversion means among the words stored in the first dictionary means. Invalid word pointer storage means for storing; load means for loading the first pointer as the second pointer from the second dictionary means to the invalid word pointer storage means; Determining means for determining whether or not the dictionary identification information stored in the dictionary means and the dictionary identification information stored in the second dictionary means match.
When the determination means determines that they match,
The first pointer loaded by the loading means is stored as the second pointer in the invalid word pointer storage means, and when it is determined that there is a mismatch, the first pointer is stored.
Of the dictionary means to find a pointer in the dictionary means for a word whose reading and word information matches the word stored in the second dictionary means, and the invalid word pointer storage means obtains the pointer. A first control means for controlling the pointer to be stored as a second pointer; and a pointer in the first dictionary means for a word whose reading is the inputted kana character string in the conversion by the converting means. It is determined whether a pointer that matches the obtained pointer is stored in the invalid word pointer storage means, and if it is determined that the pointer is stored, the word corresponding to the pointer is excluded from the conversion target. And a second control means for controlling.

EXAMPLES The present invention will be described in detail below with reference to the drawings.

FIG. 1 is an example of the overall configuration of the present invention.

In the illustrated configuration, the CPU is a microprocessor, performs arithmetic operations for character processing, logical judgments, etc., and has an address bus AB, a control bus CB, and a data bus D.
Via B, each component connected to those buses is controlled.

The address bus AB transfers an address signal indicating a component to be controlled by the microprocessor CPU. Control bus CB is a microprocessor CPU
The control signals of the respective components to be controlled by are transferred and applied. The data bus DB transfers data between the constituent devices.

Next, the ROM is a fixed read-only memory.
Microprocessor CP described later with reference to FIGS. 0 to 14
U control procedure and kana-kanji conversion dictionary DIC
Remember.

The RAM is a writable random access memory having a structure of 1 word 16 bits, and is used for temporary storage of various data from each constituent element. DELD is a deleted word dictionary and stores words deleted from the dictionary DIC.
SWTBL is a search word table, and is a table for temporarily storing the existence positions of words required during Kana-Kanji conversion.

KB is a keyboard, which is a character / symbol input key such as an alphabet key, a hiragana key, or a katakana key, and
Various function keys such as a conversion key, a word deletion key, and an execution key for instructing various functions for the character processing apparatus are provided.

DISK is an external memory for storing the document data and the deleted word dictionary DELD. Document, delete word dictionary DEL
D is stored as needed, and the stored data is recalled when needed by a keyboard instruction.

CR is a cursor register. The CPU can read and write the contents of the cursor register. The CRT controller CRTC described later displays a cursor at a position on the display device CRT corresponding to the address stored here.

DBUF is a display buffer memory that stores a pattern of data to be displayed.

CRTC is cursor register CR and buffer DBUF
It plays the role of displaying the contents stored in the display CRT.

The CRT is a display device using a cathode ray tube or the like, and the display pattern of the dot configuration and the display of the cursor on the display device CRT are controlled by the CRT controller.

Further, CG is a character generator, which stores patterns of characters and symbols to be displayed on the display device CRT.

The character processing device of the present invention comprising the above-described components operates in response to various inputs from the keyboard KB. When an input from the keyboard KB is supplied, first, an interrupt signal is sent to the microprocessor CP.
The control signal is sent to U, the microprocessor CPU reads out various control signals stored in the ROM, and various controls are performed in accordance with these control signals.

FIG. 2 is a diagram showing an example of conversion operation by the device of the present invention. 2-1 shows the screen when the reading column "Kobeko" is input. The cursor is displayed next to the input reading row. When the conversion key is pressed here, a screen 2-2 is displayed. In 2-2, the reading column “Kobeko” is “Kobe Isao”
Has been converted to. This is not the desired conversion, so delete the word "Kobe" (surname) from the dictionary, and then enter "Kobeko" again 2-3
Will be displayed. If you press the conversion key again here, 2-4
Is displayed, and this time it is correctly converted to "Kobe Port".

FIG. 3 is a diagram explaining an operation for deleting a word. Reference numeral 3-1 indicates an initial screen, and when the word delete key is pressed in this state, a screen 3-2 is displayed. The word deletion is started in 3-2, and the input window for the deleted word is displayed. When the operator inputs the reading of the word to be deleted, "Kobe", the reading of the deleted word is displayed in the window.
-3 screen is displayed. When you press the conversion key further, "Kobe" is converted to "Kobe" and displayed in the window.
It becomes the screen of 3-4. When the enter key is pressed here, the notation and reading of "Kobe" are taken in, the part-of-speech input window opens, and the screen 3-5 appears. When the operator inputs the part-of-speech "surname", the screen of 3-6 appears. When the enter key is pressed here, the word with the reading "kobe" notation "Kobe" part-of-speech "surname" is deleted from the dictionary, and the end message is 3
It is displayed as shown in -7.

FIG. 4 is a diagram showing the structure of the dictionary DIC. Dictionary is R
Present on OM. Therefore, the contents cannot be changed.

The dictionary stores "dictionary version" at the beginning. This is data that is updated when the contents of the dictionary are changed, and dictionaries with exactly the same contents have the same dictionary version. Whether or not the address part of the deleted word needs to be updated as described later is managed by this dictionary version.

Word data is stored following the dictionary version. Each word data consists of "reading", "notation" and "part of speech".

"Yomi" stores word reading information, for example, "Kobe" for "Kobe". The code uses the lower byte of JIS X 0208 code and is stored as 1 byte per character.

In the "notation", the notation information of the word, for example, if "Kobe" is used, the character "Kobe" is 2 bytes per character JIS X 0208
It is stored using code.

“Part of speech” is the word's part of speech, for example, “Kobe”
“Last name”, “place name”, etc. are stored.

FIG. 5 is a diagram showing an example of data stored in the dictionary DIC. As shown in the figure, the word data of the dictionary is stored in ascending order of reading (lexicographic array).

FIG. 6 is a diagram showing the structure of the deleted word dictionary DELD.

The deleted word dictionary is stored in the external memory. The necessary parts are read from the external memory when the power is turned on, and the RAM
Loaded in.

The deleted word dictionary is divided into three parts.

6-1 is a part for storing the "dictionary version". As the dictionary version, the dictionary version of the dictionary DIC when the address part of the deleted word dictionary is created is stored as it is.

Reference numeral 6-2 is an entity portion, in which readings, notations and parts of speech of deleted words are stored as will be described in detail with reference to FIG.

An address portion 6-3 stores a pointer indicating where the deleted word is on the dictionary DIC as described in detail in FIG.

Note that only the dictionary version 6-1 and the address part 6-3 are always present in the RAM. The entity portion 6-2 normally exists only in the external memory. When it becomes necessary to recreate the address part, the substance part 6-2 is temporarily loaded from external memory into RAM. In other situations, the memory corresponding to the real part is released and is effectively used for another purpose.

FIG. 7 is a diagram showing the detailed structure of the deleted word dictionary entity.

It consists of three fields, "reading", "notation" and "part of speech", and stores the reading, notation and part of speech of the deleted word.

The deleted word should normally be present in the dictionary DIC,
When it exists, the reading, the notation, and the part of speech on the dictionary DIC are stored as they are.

In the figure, the deleted word 1 is "Kobe" (surname) and the deleted word 2 is "Cabinet" (noun).

In addition, even if a word that does not exist in the dictionary DIC is described as a deleted word, it does not matter. Deleted words that do not exist are simply ignored.

FIG. 8 is a diagram showing a detailed configuration of the deleted word dictionary address part.

In the deleted word dictionary address section, for each deleted word stored in the deleted word entity section, the word DI
It remembers where it exists in C.

The deleted word dictionary address part is data that exists for speeding up the kana-kanji conversion process. Even without this data,
It can be recreated at any time from the entity part and the dictionary DIC.

For example, since the deleted word 1 is “Kobe” (surname) according to FIG. 7, the address where “Kobe” (surname) on the dictionary DIC exists is stored as the first entry. Similarly, for the deleted word 2, the address where "Cabinet" (noun) exists is stored as the second entry.

The deleted word dictionary address part is data depending on the version of the dictionary DIC, and the dictionary version of the dictionary DIC at the time of creation is stored at the head of the deleted word dictionary. Also, when the version of the dictionary DIC is changed, it is recreated by referring to the data of the deleted word dictionary entity.

FIG. 9 is a diagram showing the structure of the search word table SWTBL.

The search word table is a table that temporarily stores where in the dictionary DIC the words necessary for analyzing the input reading string are in the process of performing the kana-kanji conversion process.

For example, when the input reading string is "Koubekou",
For this analysis, words such as "ko", "kou", "koube", "uu", "uube", "beko", "kou", etc. are necessary, and the location of those words on the dictionary DIC is the field "pointer". Memorized in.

The operation of the above embodiment will be described according to the flow.

FIG. 10 is a flow chart of a part for receiving a key input and performing a process.

Step 10-1 is an address creating process, which initializes the deleted word dictionary as shown in FIG. This process is normally executed only immediately after the power is turned on.

Step 10-2 is a process of fetching data from the keyboard. The type of the key fetched in step 10-3 is determined, and the process branches to each key processing routine.

When the conversion key is input, the process branches to step 10-4, and the conversion process of kana-kanji conversion is performed in step 10-4 as described in detail in FIG. After that, the process branches to step 10-2.

When the word deletion key is input, the process branches to step 10-5, and in step 10-5, the word deletion supplemental processing described in detail in FIG. 14 is performed. After that, the process branches to step 10-2.

If the key is any other key, the process branches to step 10-6, and other processes such as insertion and deletion that are carried out in a normal character processing device are carried out. After that, the process branches to step 10-2.

FIG. 11 is a detailed flowchart of the "address creation process" of step 10-1.

In step 11-1, the "dictionary version" and "address part" of the deleted word dictionary are loaded from the external memory into the RAM.

In step 11-2, the dictionary version of the deleted word dictionary is compared with the dictionary version of the dictionary DIC. If they match, the process returns as it is, but if they do not match, it is necessary to recreate the deleted word dictionary address part, so the process proceeds to step 11-3.

In step 11-3, the address part is first initialized for re-creation, and the dictionary version of the dictionary DIC is set as the dictionary version in the deleted word dictionary. Also, the deleted word dictionary entity is read from the external memory into the RAM.

In step 11-4, one word is deleted from the actual part.

In step 11-5, it is determined whether the processing has been completed for all the deleted words, and if the processing has been completed, the process returns. When the processing is not completed, step 11-
Proceed to 6 to determine the deleted word address.

In step 11-6, it is searched where the same word as the extracted deleted word exists in the dictionary DIC to obtain the address.

In step 11-7, the obtained address is set in the deleted word dictionary address part.

Then, since the next deleted word must be processed, the process branches to step 11-4.

FIG. 12 is a detailed flowchart of the “conversion process” in step 10-4.

In step 12-1, the word search processing described in detail in the thirteenth step is performed in order to register the words necessary for analyzing the input reading string in the search word table SWTBL.

In step 12-2, morphological analysis, syntactic analysis, etc. are performed to analyze the input reading string, and bunsetsu candidates are created.

In step 12-3, the likelihood of each bunsetsu candidate is calculated, which bunsetsu is most likely to be converted is determined, and it is determined as the first candidate.

In step 12-4, a conversion result is created and output based on the determined first candidate.

FIG. 13 is a detailed flowchart of the "word search process" in step 12-1.

In step 13-1, one reading (reading to be searched) of a word necessary for analyzing the input reading string is determined from the dictionary DIC.

In step 13-2, it is judged whether or not there are no more readings to be searched for, and when there is no more reading, the process returns.

In step 13-3, the dictionary DIC is actually searched for the reading to be searched for the address.

The deleted word dictionary address part is searched to determine whether the address found in step 13-4 is described in the deleted word dictionary address part.

In step 13-5, it is judged whether or not there is a matching address, and if it exists, it is considered that the word has been deleted. Therefore, the process directly branches to step 13-1 to proceed to the next search reading process. If it does not exist, the word has not been deleted, so the process proceeds to step 13-6, and the address is registered in the search word table.

FIG. 14 is a detailed flowchart of the "word deletion process" in step 10-5.

In step 14-1, a notation input window for deleting words is displayed on the screen.

In step 14-2, a process of accepting the notation of the deleted word from the operator is performed. The reading entered by the operator and the selected notation are temporarily stored in the internal memory. When the operator inputs the notation and hits the execution key, the process proceeds to step 14-3.

In step 14-3, a window for entering a part of speech is displayed. Here, the part of speech input by the operator is captured in the internal memory. When the operator inputs a part of speech and hits the execution key, the process proceeds to the next step 14-4.

In step 14-4, the reading, the notation, and the part of speech obtained as a result of the processing so far are registered in the deleted word data entity part. Since the actual part does not normally exist in the RAM, it is read from the external memory and registered.

In step 14-5, the dictionary D of the word just deleted
Find the address on the IC.

In step 14-6, the obtained address is also registered in the deleted word data address.

In step 14-7, the word deletion ending process is performed. That is, the entire deleted word dictionary is stored in the external memory, the real part area in the RAM is opened, and the end message is displayed on the window. Delete the window at an appropriate timing and then return.

[Other Embodiments] In the above description, R is used as the memory for storing the dictionary.
Although the case of the OM has been described, the situation is the same for all non-writable memories, and the present invention is applicable. For example, even if it is an optical disk, a CDROM, etc., it is not possible to directly delete the word because it cannot be written, but it can be deleted according to the principle of the present invention.

Further, although the deletion word dictionary has an address as a way of holding, if the way of holding other than the address depends on the version of the dictionary, the situation is the same, so the present invention can be applied. . For example, the same configuration can be performed by storing the serial numbers of words from the beginning of the dictionary.

Also, as a way of holding a deleted word dictionary, reading, notation of words,
Although it has a part of speech, if there is other word information to be stored, it is necessary to store it, and if it is information that is not related to matching, the storage can be omitted. For example, if there are words in the dictionary that differ in frequency and have the same reading, notation, and part of speech, the frequency also needs to be stored in the deleted word dictionary. However, if there are no words that have the same reading, notation, and part of speech and differ only in frequency, there is no need to overdo and store the frequency in the deleted word dictionary.

[Effects of the Invention] As described above, according to the present invention, for a word in a dictionary that cannot be rewritten, a pointer in the dictionary of the word that should be invalidated is stored, and by referring to this pointer, rewriting cannot be performed. In order to invalidate a word in a possible dictionary and obtain this pointer, the pointer in the dictionary and the word information itself are stored for the word to be invalid defined in a dictionary, and the dictionary If the identification information matches, the pointer is loaded and used, and if it does not match, a new pointer is created from the word information and used. You can invalidate a word that should be invalidated, especially if the dictionary versions match, you do not need to create a new pointer and you can easily obtain a pointer to realize invalidation. Has the effect of

[Brief description of drawings]

FIG. 1 is a block diagram of the overall configuration of the present invention, FIG. 2 is a diagram showing an operation example of kana-kanji conversion in the present invention, FIG. 3 is a diagram showing an example of operation of word deletion in the present invention, FIG. 4 is a diagram showing the structure of the dictionary DIC of the present invention, FIG. 5 is a diagram showing an example of words stored in the dictionary DIC of the present invention, and FIG. 6 is a general structure of the deleted word dictionary of the present invention. FIG. 7, FIG. 7 is a diagram showing a configuration of a deleted word dictionary entity part in the present invention, FIG. 8 is a diagram showing a configuration of a deleted word dictionary address part in the present invention, and FIG. The figure which showed the structure of the search word table, FIGS. 10-14 are the flowcharts which show operation | movement of the character processing apparatus of this invention. DISK ... External memory CPU ... Microprocessor ROM ... Read-only memory RAM ... Random access memory DIC ... Kana-kanji conversion dictionary DELD ... Deleted word dictionary SWTBL ... Search word table

Claims (1)

[Claims] 1. An input means for inputting a kana character string, a word reading, and word information of the word including a notation are stored in association with each other, and dictionary identification information for identifying the dictionary is stored. The stored non-rewritable first dictionary means, and referring to the first dictionary means, converts the kana character string input from the input means into a notation of a word in which the kana character string is read. And a first identification point, which is a pointer in a non-rewritable dictionary of a word to be invalidated in the conversion by the conversion means, and dictionary identification information of the non-rewritable dictionary. A second dictionary means for storing the reading of the word to be invalidated and the word information of the word to be invalidated including the notation in association with each other, and among the words stored in the first dictionary means , By the conversion means Invalid word pointer storage means for storing a second pointer which is a pointer in the first dictionary means of a word to be invalidated in the conversion, and the second dictionary means to the invalid word pointer storage means, Loading means for loading the first pointer as the second pointer, dictionary identification information stored in the first dictionary means at startup, and dictionary identification stored in the second dictionary means Determining means for determining whether the information matches, and when the determining means determines that they match,
The first pointer loaded by the loading means is stored as the second pointer in the invalid word pointer storage means, and when it is determined that there is a mismatch, the first pointer is stored.
Of the dictionary means to find a pointer in the dictionary means for a word whose reading and word information matches the word stored in the second dictionary means, and the invalid word pointer storage means obtains the pointer. A first control means for controlling the pointer to be stored as a second pointer; and a pointer in the first dictionary means for the word whose reading is the input kana character string in the conversion by the conversion means. , It is determined whether a pointer that matches the obtained pointer is stored in the invalid word pointer storage means, and if it is determined that the pointer is stored,
A character processing device comprising: a second control unit that controls so as to exclude the word corresponding to the pointer from the conversion target.