Kbase 19805: How to convert an English database to a double-byte database
| Autor |
Progress Software Corporation - Progress |
| Acesso |
Público |
| Publicação |
5/19/2000 |
|
<iso8859-1> and any double-byte codepages such as <cp949> are incompatible, and not convertible either.
It's wrong to load into a double-byte database by using a dump file
dumped from an English database in <iso8859-1> codepage.
How to convert:
1. convert the English Database to UNDEFINED by using PROUTIL
2. convert the UNDEFINED Database to cp949 by using PROUTIL
4. rebuild the index of the new converted cp949 Database by using PROUTIL
If the customers just want some of the data from the English Database,
they could finish the above conversion, connect to the new converted
cp949 Database with -cpinternal/-cpstream as cp949, and generate the dump
file through database admin. Then they could load it to their cp949
database.
However, there is a simpler workaround as long as there are no characters with values greater than 127, i.e., the code page is all English.
If customer happened to have dumped the .d files from an English database and is trying to load into a double-byte database. The .d file has a trailer indicating it was dumped as iso8859-1.
There will be a conversion error (error 6063) loading it into the Korean cp949 database. There is no conversion from ISO 8859-1
with its Latin accented characters to the Korean code page which
does not have those characters.
If the user deletes the entire trailer, then the file would load
fine.