I'm using an external module to generate a text file, in unicode format, placed in  a specific server store. I  import the data into a temporary table with:

IMPORT TABLE "FTSTMP"
FROM "mydata.csv"
IN STORE "resultsfts"';

In earlier versions all worked ok, but since I jumped to 12 build I receive the error:

ElevateDB Error #1011 An error occurred with the value .2 (A conversion error occurred)

from the first line of the file. All seems to leat to the conclusion that the import command does not accept unicode text files as it did.  The start of the file is like this:

FF FE 32 00 2C 00 ....

I observed that now the exported files are without the UNICODE preamble - byte order mark. After I eliminate the FF FE  all import was done ok.

Could you please correct the code to accept proper unicode formatted text files?

Timoscov,

<< Could you please correct the code to accept proper unicode formatted text
files?  >>

This is already on the list.  However, just to note ElevateDB has never
accepted Unicode import files with BOMs.

--
Tim Young
Elevate Software
www.elevatesoft.com

Timoscov Razvan wrote:

>
>
> FF FE 32 00 2C 00 ....
>

There is your problem: the FF FE.
The unicode version of IMPORT and EXPORT does NOT support files with
BOMs.
If you use a Hex Editor to delete those two bytes it will then import
fine.  This is the SQL standard (unfortunately).

You have to be *very* careful not to open any such file in an editor
that will automatically add the BOM.  Also watch out for .SaveToFile.

I am hoping that one day Tim will add the ability to import export
with/without the BOMs  (please, please, please, Tim)

Malcolm


--

For now, you might want to look at the utility I uploaded to the binaries
group under the title, "ANSI-to-Unicode Converter w/ BOM remover".

--
David Cornelius
Cornelius Concepts

"Timoscov Razvan" wrote in message
news:B3697CCD-66C0-49E8-8FD9-D50F76B4EDDB@news.elevatesoft.com...
> I'm using an external module to generate a text file, in unicode format,
> placed in  a specific server store. I  import the data into a temporary
> table with:
>
<cut>
>
> Could you please correct the code to accept proper unicode formatted text
> files?
>