kconv(1) 2/4/92 kconv(1)
NAME
kconv - kanji code converter
SYNOPSIS
kconv [-q] [-e] [-a] [-l limit] [-i ifmt] [-o ofmt] [-f outfile] [file]
DESCRIPTION
kconv handles arbitrary conversions between EUC, Shift-JIS, JIS, Old
JIS and NEC-JIS encoded text. The command supports all characters
defined in JIS X 0208-1990, as well as EUC codeset 3 (gaiji)
characters.
By default, the input text in file is converted to EUC (without
expanding half-size katakana) and written to stdout. If file is not
specified, input is taken from stdin. The encoding of the input file
is detected automatically (see NOTES below).
OPTIONS
-o ofmt
Set the output format to ofmt. Valid formats are:
jis - (New) JIS
ojis - Old JIS
njis - NEC-JIS
sjis - Shift-JIS
euc - Extended UNIX Code (default)
-i ifmt
Set the input format to ifmt. The same values of ifmt are
recognized as for the -o option (see NOTES below).
-f outfile
Write output to outfile instead of stdout. The output file is
created or truncated as necessary.
-e Half-size katakana in the input are converted to their full-size
equivalents. This option is always in effect if the output format
is one of the 7-bit codes.
-a Shift out to ASCII instead of JIS-Roman. Some terminals which
lack the JIS-Roman extended character set do not recognize the
default shift-out sequence to JIS-Roman. Setting this option
causes kconv to generate a shift-out sequence to ASCII instead.
The -a option is ignored if ofmt is not jis or ojis.
-l limit
Limit the amount of lookahead for automatic input format
detection to at most limit bytes.
-q Suppress non-fatal messages.
Page 1 Reliant UNIX 5.44 2, 194
kconv(1) 2/4/92 kconv(1)
NOTES
If the -i option is not used, and input is read from a pipe or device,
the automatic input detection saves all characters read for that
purpose in a temporary file. The amount of text saved may be
everything up to EOF. If kconv is used as a non-interactive filter,
either the input type should be specified with the -i option, or the
amount of lookahead should be limited with the -l option, to avoid
writing excessively large temporary files.
DIAGNOSTICS
Exit status is normally 0. Errors return an exit status of 99 and
print a diagnostic on stderr. Non-fatal errors return an exit status
of 1 and print a diagnostic on stderr, unless the -q option is set.
Input format unknown - no conversion done
The input file is not in any recognized format and was copied
through unchanged (non-fatal error).
Identity conversion: ifmt to ofmt
Input and output format types are identical, and the -e option
was not set. The input file is copied through unchanged in this
case (non-fatal error).
Suspect characters in input, check output file contents
An invalid byte sequence was detected during conversion. This
indicates that the input file contains incomplete kanji
characters, partial escape sequences, or byte sequences who's
value does not map into a kanji character. This diagnostic is
also produced if the input file contains EUC codeset 3 (gaiji)
characters, and ofmt is not euc, or if ofmt is ojis or njis, and
the input file contains one of the six kanji appended to
JIS X 0208 in 1983 and 1990 (non-fatal error).
EUC/Shift-JIS ambiguity: use -i option
The automatic input format detection could not distinguish
between EUC and Shift-JIS. The -i option must be used to set the
appropriate input format explicitly.
BUGS
The automatic input format detection cannot always distinguish between
EUC and Shift-JIS. This most commonly happens when the input file only
contains half-size katakana. The -i option must be used to set the
input format in these cases.
ENVIRONMENT
TMPDIR
Directory for temporary file created by automatic input
detection. A temporary file is written only if the -i option is
not used, and the input file is not a regular file or input is
read from a pipe.
Page 2 Reliant UNIX 5.44 2, 194
kconv(1) 2/4/92 kconv(1)
SEE ALSO
kcode(1), fkconv(3K), fkcode(3K).
Page 3 Reliant UNIX 5.44 2, 194