Import question
Fairlight
fairlite at fairlite.com
Mon Mar 20 13:01:21 PST 2006
Y'all catch dis heeyah? Don Coleman been jivin' 'bout like:
> I have an import process that imports a text file via Windows Task scheduler
> every 4 minutes. Following is the beginning of one line copied from that
> text file which will not import.
>
> 06914995:02:47:42 PM: :SMC: (copied/pasted from the imported text file)
>
> This is a colon field-delimited file. My filePro import process stops
> reading data after the apparent blank space after the "PM:". My hex editor
> says this field contains a hex 00 NUL and each variable after this field
> shows as a blank value. The import process then continues as expected
> beginning with the next record (NL). This problem arose after a software
> update to the alien application creating the file. They have been unable to
> explain or correct the issue on their end so it looks like I have to program
> around it. This happens in a limited number of random records (1-10/day)
> out of up to 25,000 records in the file. Any suggestions on how to program
> around this so the file does not have to be manually edited. In order to
> fix presently you must open the file with a text editor and delete that
> "invisible" character. It then imports correctly.
You just want to strip out every NULL in every line?
#!/usr/bin/perl
foreach my $line (<>) {
$line =~ s/\000//g;
print(${line});
}
exit;
Save that program, mode it 755, pipe your data through it or call it with
the data file path as an argument (either works), and redirect the output
to a new file.
I'd have used sed, but I can't honestly remember how to represent a null to
sed off the top of my head. Same concept, I just don't use sed much since
I'm in perl so often. I have $0.01 that says Bob Stockler gives you an
awk solution that does the same thing, and JPR can readily produce the sed
version. :)
mark->
More information about the Filepro-list
mailing list