Alastair :
>>
....
The other problem is to do with converting data between Dos and Windows: I
have
many instances of European names with characters outside the normal US/UK
standards.
..... does anybody have any experience and/or hints how I can overcome the
problem?
<<
I am moving my app from DOS (3.1a) to Windows (6.1a and 6.5++).
I decided not to convert but rewrite and redesign DB.
Only the data are kept from old DOS DB and moved to the new Windows DB, and
TEXT and NOTE fields may contains extended ASCII char.
What I did is :
1) In RBase 3.1, for each DOS table, I unload data in 1 file for each DOS
table :
OUTPUT FileName
UNLOAD DATA FROM TableName AS ASCII
OUTPUT screen
2) In RBW, for each Windows Table, I reload data :
LOAD TableName FROM Filename AS ASCII
3) In RBW, I have a conversion program (first version was writen by Scott
Salibury in 1998) that search in the Windows DB structure every TEXT/NOTE
fields, and parse each char and find the corresponding ANSI char.
You must take care about :
- table's structure may be different in DOS and WINDOWS DB
- Delimit and Quote settings may be differents,
- If you use CONSTRAINT in the Win DB, you may have problem to reload some
rows that were corrects with RULES
This works fine for tables up to 30.000 rows.
If you want the conversion program, please send a private mail.
It works for french char.
You'll have to add danish, german ... char.
Hope this help,
J.M. GRATIAS, Logimatique, France
