bill lam wrote:
> I found unicode is missing in ddins, it needs the following amendment,
>
> SQL_C_BINARY=: _2
> + SQL_C_WCHAR=: _8
>
>
> case. SQL_CHAR;SQL_VARCHAR do.
> (bname)=: ,a=. ln&{."1 brow&{ >(of+i){x
> (blname)=: (nrows,1)$#{.a
> q=. sh;(>:i);SQL_C_CHAR;(vad bname);(#{.a);(<vad blname)
> + case. SQL_WCHAR;SQL_WVARCHAR do.
> + (bname)=: ,a=. ln&[EMAIL PROTECTED]:@(7&u:)"1 brow&{ >(of+i){x
> + (blname)=: (nrows,1)$+:#{.a
> + q=. sh;(>:i);SQL_C_WCHAR;(vad bname);(+:#{.a);(<vad blname)
>
> Chris, please commit the change if there is no error.
>
> However in order to test unicode data for ddins, it needs to
> 1. use unicode databse, eg. access 2000 or later, jdata.mdb in examples is
> only
> access 97.
> 2. use SQLConnectW instead of SQLConnect in ddcon.
>
> - z=. sqlconnect LASTCONNECT=: HDBC;(bs dsn),(bs uid),bs pwd
> + bsw=: (];#)@u:@(7&u:)
> + sqlconnectW=: 'odbc32 SQLConnectW s x *w s *w s *w s' &cd
> + z=. sqlconnectW LASTCONNECT=: HDBC;(bsw dsn),(bsw uid),bsw pwd
>
It should also make it unicode safe for SQL_CHAR SQL_VARCHAR
case. SQL_CHAR;SQL_VARCHAR do.
- (bname)=: ,a=. ln&{."1 brow&{ >(of+i){x
+ (bname)=: ,a=. ln&{.@(1&u:)"1 brow&{ >(of+i){x
--
regards,
bill
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
