gawk - delete line in file

[frog_123]

Hi,

I want to delete lines which contain the word "Fehler" but for some reason my programm isn't working:

BEGIN { FS="|"
		OFS="|"
		SUBSEP="|"
		anzArg = 2
		if (ARGC < anzArg) {
			print "Zuwenig Argumente"
			print "gawk -f makeEGNinput.awk inputfiles"
			nosource = "true"
			exit 1
		}
	}

{ # Main
	if (FNR == 1) {
		outfile = (FILENAME ".load")
		print ("Outfile: " outfile)
	}
	# Delete path
	split($5,arrDatei,"/")
	# Delete line with  <Fehler> 
	if ($5 != "<Fehler>"){
		n = asort(arrDatei,arrCount)
		$5 = arrDatei[n]
	}
	#print $0
	print $0 > outfile
}

My input file looks like this:
00036425|06|2010|0001|/usr/swas/pmscp/betr/eis/teis/EGN_00036425_20101013091045.pdf
00041332|06|2010|0000|<FEHLER>
00036425|06|2010|0001|/usr/swas/pmscp/betr/eis/teis/EGN_00036425_20101013091045.pdf

I can somebody help me?

Thanx
frog_123

[Irithori]

Does it need to be gawk?

grep -v FEHLER inputfile
should fullfill your requirements.

[frog_123]

thanks a lot for that simple solution!!

[clowenstein]

In any case, you were testing for "Fehler" while expecting "FEHLER".

[knuno]

Or how about sed?
sed /FEHLER/d inputfile > outputfile
or if you want to change the inputfile directly:
sed -i /FEHLER/d inputfile

[knuno]

And to correct myself, I see your inputfile contains slashes. Just use any delimiter not in your inputfile, for example:
sed #FEHLER#d inputfile > outputfile

Knut

[scm]

And to correct your correction, sed doesn't really care about slashes in its input stream, it's already parsed them out of the pattern. It does, of course, care about slashes in the pattern argument.