For a simple character set named
MYSET
, sorting rules are specified in
the
configuration file using MYSET
.xml<map>
array
elements within <collation>
elements.
If the sorting rules for your language are too complex to be
handled with simple arrays, you must define string collating
functions in the
ctype-
source file in the MYSET
.cstrings
directory.
The existing character sets provide the best documentation and
examples to show how these functions are implemented. Look at
the ctype-*.c
files in the
strings
directory, such as the files for
the big5
, czech
,
gbk
, sjis
, and
tis160
character sets. Take a look at the
MY_COLLATION_HANDLER
structures to see how
they are used. See also the
CHARSET_INFO.txt
file in the
strings
directory for additional
information.