The data formats include the following 3 types; character, numeric and number.
Characters are compared in 8 bit transparent order. When two or more byte characters coexist, each character can be compared by specifying the option. ASCII code, Shift JIS code, and eight unit Japanese kana code can be processed. When the code system of the input file is EBCDIC, the data is sorted in the ASCII code order.
EBCDIC characters are compared in 8 bit transparent order. When the code system of the input file is ASCII, the data is sorted in the EBCDIC code order.
Shift JIS codes are compared by character.
Japanese code used in general-purpose computers of Fujitsu. Characters are compared in 16-bit transparent order.
It is an International character code in which characters all over the world can be described. In PowerBSORT, the following three forms are supported.
UCS-2 form
It is the form that 1 character in Unicode is shown in 2 bytes. UCS-2 and UTF-16 can also be handled.
UTF-32 form
It is the form that 1 character in Unicode is shown in 4 bytes.
UTF-8 form
It is the form that 1 character in Unicode is converted to 8 bit code from 1 to 6 bytes. In PowerBSORT, 1 to 3 bytes are supported.
Character form two digit years
Enables comparison of 2 byte numbers as 2 digit years according to a value specified by the environment variable BSORT_TOPYY.
The character in one byte is replaced with other characters according to information on the given collation order and fields are compared.
An arbitrary bit in 1 byte is taken out and used as a comparing value of a compared field or comparing field. Mask values are used for taking out the bit. The logical product of the mask value and the value of the field become the comparing values.
Fixed-point binary number of which the leading 1 bit is a sign. The forms of the fixed-point binary numbers include the form named little endian used by 8086 family MPU and the form of big endian. When it is simply described as the fixed-point binary number, it is in the form of big endian.
Fixed-point binary number - little endian
It is a field of the fixed-point binary number used by 8086 family MPU. This form is known as little endian.
Fixed-point binary number - system dependent
When the native of the system is big endian, it means it is the fixed-point binary number. When the native of the system is little endian, it means it is fixed-point binary number - little endian.
Unsigned fixed-point binary number
It is a fixed-point binary number without a sign flag. It includes a form named little endian used by 8086 family MPU and a form of big endian. When it is simply described as the unsigned fixed-point binary number, it is in the form of big endian.
Unsigned fixed-point binary number - little endian
It is a field of the unsigned fixed-point binary number used by 8086 family MPU. This form is known as little endian.
Unsigned fixed-point binary number - system dependent
When the native of the system is big endian, it means it is the unsigned fixed-point binary number. When the native of the system is little endian, it means it is the unsigned fixed-point binary number - little endian.
IEEE floating-point binary number
Binary expression with exponent and mantissa. The first bit is a sign.
M floating-point binary number
Floating-point binary number used with a Fujitsu general-purpose computer. Binary expression with exponent and mantissa. The first bit is a sign.
It is an expression form in decimal number that the decimal numbers of 2 digits are stored in a byte.
xx | xx | .. | xx | xS |
x:0x0-0x9 S:[Positive value]0xa, 0xc, 0xe, 0xf [Negative value]0xb, 0xd Note) When other codes than the ones described as positive or negative in the table appear in the position of the sign to be shown in 4 bits, they are handled as the positive sign.
Unsigned packed decimal number
The following two formats are supported in PowerBSORT:
Unsigned packed decimal number (with sign nibble)
This stores numbers, two decimal digits in a single byte, with sufficient length to store the decimal number. The last four bits (nibble) are not signed.
xx | xx | .. | xx | xS |
x:0x0-0x9 S:0xf Note) The value of the last four bits is not checked.
Unsigned packed decimal number (without sign nibble)
This expresses the decimal number without a sign nibble storing numbers as two decimal digits in a single byte.
xx | xx | .. | xx | xx |
x:0x0-0x9
It is an expression form in decimal number that the decimal numbers of 1 digit are stored in the form that can be displayed.
[ASCII code system (NetCOBOL form)]
3x | 3x | .. | 3x | Sx |
3:0x3 x:0x0-0x9 S:[Positive value]0x4 [Negative value]0x5
[ASCII code system (Micro Focus COBOL form)]
3x | 3x | .. | 3x | Sx |
3:0x3 x:0x0-0x9 S:[Positive value]0x3 [Negative value]0x7
[EBCDIC code system]
Fx | Fx | .. | Fx | Sx |
F:0xf x:0x0-0x9 S:[Positive value]0xa, 0xc, 0xe, 0xf [Negative value]0xb, 0xd Note) When other codes than the ones described as positive or negative in the table appear in the position of the sign to be shown in 4 bits, they are handled as the positive sign.
Unsigned external decimal number
It is an expression form in decimal number that the decimal numbers of 1 digit are stored in the form that can be displayed.
[ASCII code system]
3x | 3x | .. | 3x | 3x |
3:0x3 x:0x0-0x9
[EBCDIC code system]
Fx | Fx | .. | Fx | Fx |
F:0xf x:0x0-0x9
External decimal form two digit years
2 byte external decimal numbers are compared as 2 digit years according to a specified value of environment variable BSORT_TOPYY.
Packed decimal form two digit years
2 byte packed decimal numbers are compared as 2 digit years according to a specified value of environment variable BSORT_TOPYY.
1 byte unsigned packed decimal numbers are compared as 2 digit years according to a specified value of environment variable BSORT_TOPYY.
Same as the Unsigned external decimal number.
Leading separate signed number
Expression form in decimal number that the decimal numbers of 1 digit are stored in the form that can be displayed with the sign section in the leading 1 byte.
[ASCII code system]
SS | 3x | .. | 3x | 3x |
3:0x3 x:0x0-0x9 SS:[Positive value]0x2b [Negative value]0x2d
[EBCDIC code system]
SS | Fx | .. | Fx | Fx |
F:0xf x:0x0-0x9 SS:[Positive value]0x4e [Negative value]0x60 Note) When other codes than the ones described as positive or negative in the table appear in the position of the sign to be shown in 8 bits, they are handled as the positive sign.
Trailing separate signed number
Expression form in decimal number that the decimal numbers of 1 digit are stored in the form that can be displayed with the sign section in the trailing 1 byte.
[ASCII code system]
3x | 3x | .. | 3x | SS |
3:0x3 x:0x0-0x9 SS:[Positive value]0x2b [Negative value]0x2d
[EBCDIC code system]
Fx | Fx | .. | Fx | SS |
F:0xf x:0x0-0x9 SS:[Positive value]0x4e [Negative value]0x60 Note)When other codes than the ones described as positive or negative in the table appear in the position of the sign to be shown in 8 bits, they are handled as the positive sign.
Leading overpunch signed number
Expression form in decimal number that the decimal numbers of 1 digit are stored in the form that can be displayed. In the leading 1 byte, signs and numbers are stored.
[ASCII code system (NetCOBOL form)]
Sx | 3x | .. | 3x | 3x |
3:0x3 x:0x0-0x9 S:[Positive value]0x4 [Negative value]0x5
[ASCII code system (Micro Focus COBOL form)]
Sx | 3x | .. | 3x | 3x |
3:0x3 x:0x0-0x9 S:[Positive value]0x3 [Negative value]0x7
[EBCDIC code system]
Sx | Fx | .. | Fx | Fx |
F:0xf x:0x0-0x9 S:[Positive value]0xa, 0xc, 0xe, 0xf [Negative value]0xb, 0xd Note) When other codes than the ones described as positive or negative in the table appear in the position of the sign to be shown in 4 bits, they are handled as the positive sign.
Trailing overpunch signed number
It is the same form as the External decimal number.
