The environment variable that influences the operation of PowerBSORT is shown below.
Environment variable | Meaning |
---|---|
Specification of character string used to specify internationalization information | |
Specification of character collating sequence used | |
Specification of width of character classification, character variable, and multiple byte character | |
Specification of language of message data base used | |
Specification of directory path name of temporary file |
Environment variable | Meaning |
---|---|
FIFO (first-in, first-out) option | |
Unicode UCS-2 format byte order | |
Unicode UTF-32 format byte order | |
Whether or not BOM is skipped |
LANG=locale, LC_ALL=locale
The default value of LC_*.
Refer to the manual of OS for details of environment variable LANG and LC_*.
LC_COLLATE=locale
Specify locale of the collation data base made by the localedef command.
Sort and merge use the collating sequence defined in locale.
LC_CTYPE=locale
Specify the width of the character classification, the character translation, and the multiple byte character.
LC_MESSAGES=locale
Specify the language of the message data base used.
For instance, the application can use the message data base for English and the message data base for Japanese properly.
TMPDIR=directory path name[:...]
Specify the directory path name of the temporary file.
Two or more directory path names can be specified by delimiting it by colon (:).When BSORT_TMPDIR of a startup file is omitted, it becomes effective.
BSORT_FIFO={ON | OFF}
The FIFO (first-in, first-out) option is specified.
Specified value | Meaning |
---|---|
ON | When the value of the specified key field is the same, those records are output in input order. This is effective for the sort function. The specification of the fifo (first-in, first-out) option is disregarded when combined with the merge function, the copy function, the record summation option, and the suppression option. |
OFF | The order of records with the same key may be different in the output file compared to the input file. |
Note
When this specification is omitted, the following specification becomes effective.
FIFO option (-f) in bsort command
fifo operand of -option option in bsortex command
optionfunc of BSRTPRIM structure in BSORT function
BSORT_UCS2TYPE={BIG_ENDIAN | LITTLE_ENDIAN}
Specify the Unicode UCS-2 format byte order.
Valid when the data format is set to UCS-2 using the PowerBSORT command line interface.
Specified value | Meaning |
---|---|
BIG_ENDIAN | Sets Big Endian. |
LITTLE_ENDIAN | Sets Little Endian. |
Note
When this specification is omitted, the specification of BSORT_UCS2TYPE of a startup file becomes effective.
BSORT_UTF32TYPE={BIG_ENDIAN | LITTLE_ENDIAN}
Specify the Unicode UTF-32 format byte order.
Valid when the data format is set to UTF-32 using the PowerBSORT command line interface.
Specified value | Meaning |
---|---|
BIG_ENDIAN | Sets Big Endian. |
LITTLE_ENDIAN | Sets Little Endian. |
Note
When this specification is omitted, the specification of BSORT_UTF32TYPE of a startup file becomes effective.
BSORT_UNICODEBOM={ON | OFF}
Specify whether or not the Unicode byte order marker, BOM, is skipped.
Valid when the input file is a text file and the code set of that file is Unicode (UCS-2, UTF-32, or UTF-8).
If a BOM exists in the input file and is skipped, a BOM is still inserted in the output file. If multiple input files are specified, a BOM exists in the first input file specified, and the BOM was skipped, a BOM is inserted in the output file.
Specified value | Meaning |
---|---|
ON | Skips the BOM. Note that, if the head of the file is anything other than a BOM, it is not skipped. |
OFF | Does not skip the BOM. |
Note
When this specification is omitted, the specification of BSORT_UNICODEBOM of a startup file becomes effective.