"Property not found" error on big text import

[urbanist1100]

I'm trying to import a big text file (which has really
long records - almost 20,000 columns). I'm only really
concerned with col. 1-80 and 2223-2550. Anyhoo, I've
written in my own specs in advanced mode under the import
text file wizard, and everything seems fine until I
hit "finish". Then I get the error message: "Property not
found", and the whole thing stops and puts me back into
the wizard. Any ideas?

urbanista

 

[John Nurick]

Hi Urbanista,

IIRC there's a limit on the length of record that the built-in import
routine can handle. If you have a fixed width file with 20,000
characters per record you're probably over it, and if it's a delimited
file with 20,000 fields per record I'm sure you are.

I'd tackle this by eliminating the unwanted data at the text file level
and then importing what's left. For a fixed-width or simple delimited
file, download the Windows version of the Gnu textutils from
http://unxutils.sourceforge.net/ and use the "cut" utility. The command
line for a fixed-width file will be something like this:

cut -c1-80,2223-2550 D:\folder\file.txt > D:\folder\new.txt

(use -f for a simple delimited file).

If it's a delimited file whose text fields are quoted so they can
contain delimiter characters or linebreaks things get a bit more
complicated: post back if you have trouble.