Tomi NA a écrit :
2006/11/23, Arnaud Lesauvage <thewild@xxxxxxxxxxx>:
Arnaud Lesauvage a écrit :
> Brandon Aiken a écrit :
>> It also might be a big/little endian problem, although I always thought that was platform specific, not locale specific.
>>
>> Try the UCS-2-INTERNAL and UCS-4-INTERNAL codepages in iconv, which should use the two-byte or four-byte versions of UCS encoding using the system's default endian setting.
>
>
> Guys, it worked !!!!
> UCS-4-INTERNAL was the right choice !!!
>
> I love you all !
>
> (now I just have an out of memory problem, but that's going
> to be a new thread)
Guys, it did not work !!! :(
I thought it worked because postgres seemed to be loading
the file and failing at the end with an "out of memory"
error, but in fact I think the conversion remove all
end-of-line characters (one line of 1.5GB was too much for
COPY...).
Still searching !
It will take you a day or two to get started, and then a day or two to
get the job done, but you really might want to look into kettle or
some other ETL tool to do the job.
It looks to me like you're trying to screw in a screw using a hammer.
Yes, I might try something else.
I was thinking that others would probably run into this
problem sometime, and that our investigations might help them.
I think I'll forget about this COPY stuff and just export
with DTS through PostgreSQL ODBC Unicode.
--
Arnaud
