At 07:39 AM 4/08/2005 +0000, you wrote:

>Hi... I have to be able to store 'unusual' macrons, eg in the word
>'shogun' there is a line above the 'o'. The 'o' with a line above it
>is unicode character \u014D
>
>I've pasted this value into a column manually using IBexpert; while
>editing the char is correct, and when I exit editing mode and commit,
>the value looks like a normal 'o'.
>If I edit it again it looks correct (the one I originally put in), but
>if I exit IBexpert and restart, it now looks like a normal 'o', even
>when editing.
>
>This could be an IBexpert problem, but I'm not sure if it's been
>stored correctly.

The latter. After committing, IBExpert is just showing you what has been
stored in the database or, rather, your environment's interpretation of that.

>In my java application I retrieve the value via JDBC but when it's
>displayed in an edit box it doesn't show as a normal 'o' /or/ the
>special macron, but as an 'undisplayable' block character. I know my
>application can display the special macron ok because I can paste the
>correct one in the edit box and view it ok. (It normally then gets
>stored in the db.)

What you are not telling us is what the character sets of the client and
the destination column are.

>So how can I tell if the database is storing it correctly?

Difficult to say, until we know the conditions under which the characters
are being stored and retrieved and whether you are inspecting the results
using a viewing interface that actually supports the images for those
characters.

>Or could both IBexpert and JDBC be corrupting the character to/from
>the database?

Both IBExpert will be doing whatever the conditions dictate. So provide
more information about the conditions....

./heLen