Re: sqlldr and TAB

Personally I dislike SQL*Loader in many situation. What I prefer to do it blow the entire record into a table as fast as I can and then parse it and distribute it using a stored procedure. It may be a bit slower but it gives far greater control and error reporting ability.

Daniel A. Morgan

Carsten Landeck wrote:

> this works, but when a field is empty it scrambles my data
> with "," it works just fine, but two whitespaces in row are treated as one
>
> so this happens:
> source
> v1 tab tab v3
>
> f1 v1
> f2 v4
> f3 -
>
> "Dmitry" <support_at_ispirer.com> schrieb im Newsbeitrag
> news:9askbp$6kr4a$1_at_ID-75577.news.dfncis.de...
> > Hi,
> >
> > For loading the TAB delimited file use:
> >
> > LOAD DATA
> > INFILE 'import.tab'
> > INTO TABLE imp
> > INSERT
> > FIELDS TERMINATED BY X'09' OPTIONALLY ENCLOSED BY '"' TRAILING NULLCOLS
> > (ObjClass,
> > ...
> > )
> >
> > Best regards, Dmitry
> > --
> > http://www.ispirer.com/chyfo.html - Tool exports data to CSV flat file,
> > generates CREATE TABLE, CREATE INDEX scripts for Oracle and control files
> > for SQL Loader.
> >
> >
> > "Carsten Landeck" <Landeck.c_at_zdf.de> wrote in message
> > news:3ad1b008$0$4943$73bec562_at_personalnews.de.uu.net...
> > > Hi
> > > I have a beginner question:
> > > In the *.ctl file I have the option "FIELDS TERMINATED BY"
> > > I'm using tabulators, but '\t' doesn't work.
> > >
> > > My data-source file is ok
> > >
> > > -> Value TAB value TAB value CR <-
> > >
> > > LOAD DATA
> > > INFILE 'import.tab'
> > > INTO TABLE imp
> > > INSERT
> > > FIELDS TERMINATED BY '\t' OPTIONALLY ENCLOSED BY '"' TRAILING NULLCOLS
> > > (ObjClass,
> > > ...
> > > )
> > >
> > > The result is, that the whole tupel is stored in the first field.
> > >
> > > Has someone a hint?
> > >
> > > Carsten
> > >
> > >
> >
> >
  Received on Tue Apr 10 2001 - 07:23:14 CEST