If it were me I would start with looking at the national character
source in derby and the java interfaces they use.
I think a reasonable first step would be to look at implementing
native sort order on a database wide basis for all existing character
types. I believe rick took a look at enabling the national character
type but in the end did not think that it was the right way to go.
I think it makes sense to base the sorting on existing java technology
rather than build anything special in derby.
The easiest would be to requre the native sort order to be defined
at database creation time. I believe the most straight forward
implementation would build on disk indexes also in native sort order
so making the decision at runtime would require alternate code paths,
while making the decision at db create time I believe would limit
the the path to compare routines in the datatypes.
I think it is important if at all possible to not slow down existing
db's that don't need this special sorting.
Mamta Satoor wrote:
Hi,
I hope this is not a very basic question but I need help understanding
language support and Derby database in general. In my little research
through the Derby documentation and derby dev list, I have noticed
terminologies like locale, territory, codeset, codepoint, locale based
sorting, codepoint based sorting, etc? How do I educate myself about
this area of the database? Any help will be highly appreciated.
Mamta
