JP2569857B2 - Variable byte length character input control method - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an undefined byte length character input control system for processing a character string in which characters having different byte lengths are mixed in a computer system.

(Prior Art) In recent information processing systems, processing of Japanese data has become indispensable. However, various problems have occurred with respect to the processing of characters of Japanese data that cannot be represented by one byte. These various problems stem from the fact that the code system has been expanded by adding a Japanese code system expressed by one character and two bytes to a one-byte code system using alphanumeric characters as a standard. For example, when processing a character string in which a two-byte character appears after a one-byte character, there is a problem of how to notify a system that processes the character string that the byte length representing one character has changed. , ASCII code '1' and J
There is a problem that the IS kanji code '1' must be recognized as the same number. These problems have a great effect on a system that processes text codes, such as a language processing system.

FIG. 2 is a diagram showing a configuration of a conventional character input control system of indefinite byte length. In the text code 208 generated by the text code generation unit 201, the two-byte character string portion in the text code 208 is surrounded by two control characters having a fixed byte length indicating the start or end of the two-byte character.
For example, in the character string "ABC Japanese XYZ", as shown in Fig. 3, the control that indicates the start and end of a double-byte character between "C" and "day" and between "word" and "X" respectively. Character <KI
>, <KO> are inserted. These two control characters consist of 2 bytes, and are 3F75 and 3F7 respectively in hexadecimal code.
6

Text code 2 generated by text code generator 201
The 1-byte input processing unit 202 given 08 gives a 1-byte character to the 1-byte character attribute identification unit 204 with the given text code 208. The one-byte character attribute identification unit 204 looks up the one-byte code table 206 using the given character code, and obtains an attribute bit table 209 storing attribute information for the one-byte character. Here, temporarily
When the byte input processing unit 202 finds the one-byte code of 3F, the next one byte is read, and if it is 75, the control is transferred to the two-byte input processing unit 203. Otherwise, the one-byte input process is performed. When control is transferred to the 2-byte input processing unit 203, the 2-byte character is identified by the 2-byte character attribute identification unit 205.
The two-byte code table 207 is represented by the two-byte code to obtain an attribute bit table 210 for the two-byte character. Since the control character <KO> indicating the end of the two-byte character string is two bytes, the two-byte input processing unit 203 can identify the control character <KO> without performing any special processing. When KO> is found, the control is returned to the 1-byte input processing unit 202.

Character attributes were identified as follows. For example, when one byte is an 8-bit character and indicates the presence or absence of 32 types of attributes, a byte code table with 256 elements and one element with 4 bytes is prepared, and the character code of “1” is 24 bytes.
If it is 1, the bit indicating the numeric attribute of the 241st attribute bit table of the bytecode table is set to 'ON',
And so on. For 2-byte characters, processing is performed with the table size set to 65536.

(Problem to be Solved by the Invention) In the above-mentioned conventional character input control method of undefined byte length,
Since the byte lengths of the two control characters that indicate the beginning or end of the double-byte character added before and after the double-byte character are fixed, a system that processes text codes containing characters with different numbers of bytes inputs single-byte characters. When entering medium or double-byte characters, it was necessary to accommodate control characters that differed in byte length from the character being processed. In addition, the text code processing system requires two input processing units, a one-byte character input processing unit and a two-byte character input processing unit, and the processing system is complicated and inefficient.

As described above, there is a problem to be solved in the conventional character input control method of indefinite byte length.

(Means for Solving the Problems) The present invention relates to an undefined byte length character input control in a computer system for processing a character string in which English characters represented by one byte and Japanese characters represented by a plurality of bytes are mixed. In the method, for each character with a different byte length mixed in the character string, it indicates that the byte length of the following character changes in the character string and the byte length of the following character in the character string Control character generating means for generating control characters, character attribute information managing means for hierarchizing a byte code table corresponding to each byte constituting the character and managing attribute information for each character for each byte length, different bytes Receives a character string containing mixed-length characters, and places the same character as the immediately preceding character at a virtual position between the characters with different byte lengths in the character string. Text code generation means for generating a text code by inserting the control character indicating the byte length of a character following the character, the text code being generated by the text code generation means; And character attribute identification means for identifying a character attribute defined for each character of the text code with reference to the character attribute information.

(Example) Next, the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing one embodiment of the present invention. The text code generation unit 1 generates a text code to which a control character, which is the byte length of the character before the change, is added when the byte length of the character forming the character string changes in the middle of the input character string. . The character attribute identification unit 2 is a storage area 3
1-byte character attribute information 5 and 2-byte character attribute information 6 therein. Each character attribute information has several byte code tables 7, 8, 9, and 10, and the attribute for each character of each byte length is obtained by layering the byte code table corresponding to each byte of the character. Managing. The byte code table corresponding to the least significant byte of each character includes an attribute bit table indicating the attribute of the character or a byte code table corresponding to the most significant byte of other character attribute information indicating that the byte length of the character changes. The pointer to is stored. The register CT4 indicates the byte code table of the most significant byte of the character attribute information with respect to the byte length of the currently input character. The initial value of register CT4 is 1-byte character attribute information 5
Is a pointer value to the byte code table 7 of FIG.

Consider a byte code table in which one byte is 8 bits and the size of one element is 4 bytes. Here, the code of a 1-byte control character <2S> indicating a change from a 1-byte character to a 2-byte character is 3F, and a code of a 2-byte control character <1S> indicating a change from a 2-byte character to a 1-byte character. To 3F76. At this time, the text code generation unit 1 generates the text code shown in Fig. 4 for the character string "123456 ', where" 1 "," 2 "," 5 ", and" 6 "are 1
It is assumed that the characters are byte characters and the codes are F1, F2, F5, and F6 in hexadecimal. '3' and '4' are double-byte characters,
Assume that the codes are 7BF3 and 7BF4 in hexadecimal, respectively. Control characters <2S> and <1S> are inserted between '2' and '3' and between '4' and '5', respectively. Therefore, by inserting a control character having the same byte length as the character immediately before the byte length changes before and after the 2-byte length character, all the control characters indicating that the byte length of the character changes will be the same as those of the currently input character. Since it is expressed in byte length, the control character can be identified by the byte length of the currently input character.

Next, the processing of the text code shown in FIG. 4 generated by the text code generation unit 1 will be described. FIG. 5 is a flowchart showing the processing of the character attribute identification unit 2. First, the character attribute identification unit 2 substitutes the contents of the register CT4 into the register P15 (step 501). At this time, the pointer value of the one-byte character attribute information 5 to the byte code table 7 is stored in the register CT4, and the pointer value is assigned to the register P15. Then, one byte of the text code is read and assigned to the register Q16 (step 502). Then, the character code F1 of the character “1” is passed to the register Q16, and the F of the byte code table 7 indicated by the register CT4 is
Substitute the contents of the first element into register P15 (step
503) (FIG. 7 (a)).

The content of each bytecode table element is a pointer value to another bytecode table or the bytecode for the most significant byte of the attribute bit table or other character attribute information when a character is a control character indicating a change in byte length One of the pointer values to the table. These are identified as follows. Since the size of each bytecode table is constant and is a multiple of 4 bytes,
By arranging in the storage area 3, the lower 2 bits of the head address of each byte code table can be set to 0. Therefore, assuming that the least significant bit of the byte code table element is the 0th bit and the most significant bit is the 31st bit, if the values of the first and 0th bits are 00 as shown in FIG. Can be identified as a value obtained by adding 3 to the pointer value to the byte-coded table for the most significant byte of the other character attribute information if the pointer bit is (601), 01 is the attribute bit table (602), and 11 is the attribute bit table.

Now, since the byte length of the character for which input processing is performed is 1, the content of the register P15 is not a pointer value to another byte code table, and the character is "1" and is not a control character. Each attribute processing is performed according to the contents of the attribute bit table 12 given to the
(4,505,506) (processing as a number is performed from the meaning of the character '1'). Next, the process returns to step 501, and the same process is performed on the character “2”. At step 503, 1S is stored in the register P15 for the control character 2S.
The 3F-th contents of the byte code table 7 of the byte character attribute information 5 are obtained. Since this content is a value obtained by adding 3 to the pointer value to the byte code table 8 for the upper byte of the 2-byte character attribute information 6, steps 504, 50
After step 5, the processing moves to step 507. Here Redesta P15
Is substituted into the register CT4 (step 507), and the process returns to step 501. Below, register CT4
Holds the pointer value to the byte code table for the upper byte of the 2-byte character attribute information 6. When the upper byte of the two-byte character code of “3” obtained from the text code 11 is obtained, a pointer value to another byte code table 9 is stored in the 7B-th element of the byte code table 8 indicated by the register CT4. In step 504, the process proceeds to step 502. For the lower byte, the attribute bit table 14 is obtained from the F3 contents of the byte code table 10 of the lower byte, and each attribute process is performed (step 506). (FIG. 7 (b)). For the character '4',
Same as '3'. The control character 1S is a control character indicating the start of a one-byte character consisting of two bytes. The 76th content of the bytecode table 9 indicated by the 3Fth element of the bytecode table 8 indicated by the register CT4 includes 1-byte character attribute information. A value obtained by adding 3 to the pointer value to the byte code table 7 of 5 is set, and the register CT4 is again in a state of identifying a 1-byte character by step 507. Hereinafter, characters '5' and '6' are input and processed in the same manner as '1' and '2'.

Although the embodiment described above is limited to one-byte characters and two-byte characters, in a system that handles 1 to N-byte characters, a control character indicating a change in the byte length is provided for each character. The present invention can be easily applied by defining N-1 pieces. For example, in a system that handles 1-3 byte characters, two 1-byte control characters indicate that they will change to 2-byte and 3-byte characters,
It is sufficient to define six control characters of two control characters of two bytes indicating change to one-byte and three-byte characters, and two control characters of three bytes indicating change to one-byte and two-byte characters. . Also, the present invention seems to require many bytecode tables at first glance, but in reality, there are many bytecode tables having exactly the same contents because characters having special attributes are a very limited number of characters. Tend. Therefore, it is possible to solve the problem that a large number of bytecode tables are required by preparing only one bytecode table and devising it so that it can be referred to from different places.

(Effect of the Invention) As described in detail above, the variable byte character input control method of the present invention provides a system for processing a text code in which a single-byte character is being input while a double-byte character is being input. Also, there is no disturbance in processing for control characters. Further, it is not necessary to prepare an input processing unit for one-byte characters and an input processing unit for two-byte characters in the text code processing unit, respectively, and a simple and efficient text code processing unit can be realized.

 The present invention has the above effects.

[Brief description of the drawings]

FIG. 1 is a block diagram showing one embodiment of the present invention, FIG. 2 is a block diagram showing a conventional character input control method of indefinite byte length, and FIG. FIG. 4 is a diagram showing a text code generated by the text code generation unit 1 of the present invention. FIG. 5 is a flowchart showing processing of the character attribute identification unit 2 of the present invention. FIG. 6 is a byte diagram of the present invention. FIG. 7 is a diagram showing types of elements of the code table, and FIG. 7 is a diagram showing an attribute bit table managed hierarchically by the bytecode table of the present invention. 1, 201: text code generation unit, 2: character attribute identification unit, 3: storage area, 4: register CT, 5: 1-byte character attribute information, 6: 2-byte character attribute information, 7, 8, 9,
10… Byte code table, 11,208 …… Text code, 12,13,14,209,210 …… Attribute bit table, 15…
Registers P, 16: Register Q, 202: 1-byte input processing unit, 203: 2-byte input processing unit, 204: 1-byte character attribute identification unit, 205: 2-byte character attribute identification unit, 206
... 1 byte code table, 207 ... 2 byte code table, 601, 602, 603 ... elements of the byte code table.

Claims (1)

(57) [Claims] An indeterminate byte length character input control method in a computer system for processing a character string in which English characters represented by one byte and Japanese characters represented by a plurality of bytes are mixed. Control character generation that generates control characters that have the same byte length as the character for each character with a different byte length, and indicate that the byte length of the following character changes in the character string and the byte length of the following character Means, character attribute information management means for hierarchizing a byte code table corresponding to each byte constituting a character, and managing attribute information for each character for each byte length, and a character string in which characters of different byte lengths are mixed. Received in a virtual position between characters having different byte lengths in the character string, and having the same byte length as the immediately preceding character and following the character. Text code generating means for generating a text code by inserting the control character indicating how many bytes the character has, and receiving the text code generated by the text code generating means, and referring to the character attribute information And a character attribute identification means for identifying a character attribute defined for each character of the text code.