Search Results

must be in a diff path - probably /usr/bin/ksh93 - don't have Solaris in front of me.
Try it with /usr/bin/bash...

What Solaris version are you on? Do uname -a

Hello asavaliya,

Could you please try following and let me know if this helps.

awk 'NR==1{for(i=1;i<=NF;i++){split($i, A,"@");B=B?B OFS A[1]:A[1];C=C?C OFS A[2]:A[2]};print B ORS C;next}...

Try alsoawk '/E_TIM/ {CNT++}
{LN[CNT]; HD[$1]; MX[CNT,$1]=$2}
END { for (i in HD) printf "%10s", i; print "";
...

In case Perl is an option, here's a program that does that:


$
$ cat -n f25
1 E_TIM 16,
2 ETE 15,
3 EOND 26,
4 EEC 81,
5 E_1 un,
6 E_2 ...

one way of doing it:
tac file.xml | awk '
/vsDataEUtranCellFDD/ {id=1}
/vsDataEUtranFreqRelation/ {id=2}
/vsDataEUtranCellRelation/ {id=3}
/VsDataContainer id/ {sub(/ id=/,"_id_"id"=")}1' |...

Replace tac with tail -r or sed '1!G;h;$!d', see if that works

As per your input.. try this..


awk '{if($3 ~ /A=2/ && NF == 5){a++;$(NF+1)=a}}1' file

awk '$3 == "A=2" {++arr[$2$3]}{printf $0;if (arr[$2$3] ~ /./) printf "\t" arr[$2$3]; print "";}' infile

date '+%Y%m%d.%k00'

20121114.1200

COUNTER=1

while [ $COUNTER -lt 48 ]
do
file_name=$( date --date='-'"$COUNTER"' hour' +"%Y%m%d.%H00" ); COUNTER=`expr $COUNTER + 1`
touch $file_name
done

And, depending on its controlling terminal's settings, it may not be able to write to it either.

If that script needs to run non-interactively, it's stdin, stdout, and stderr must be redirected.
...

All in awk:

awk '{ $1 = $1 ".txt" ;if (a[$1]) { a[$1] = a[$1] "\n"; }; a[$1] = a[$1] "Fi Na=" $2 " no=" $3;} END { for (i in a) { print "Hello\nprl\n" a[i] "\nbye\nq" > i } }' A.txt

for key in `awk ' { print $1 } ' A.txt | sort | uniq`
do
echo "Hello\nprl" > ${key}.txt
awk '/'$key'/ { print "Fi Na=" $2 " no=" $3 } ' A.txt >> ${key}.txt
echo "bye\nq"...

The below code should work with data like in your example:
$ cat A.txt
RL03 RL03_A_1 RL03_B_1 RL03_C_1
RL03 -119.8 -119.5 -119.5
RL07 RL07_A_1 RL07_B_1 ...

also:
awk '{
if ($1 in a) ao[ac++]=$1;
a[$1]=$1;
++l[$1];
cnt[$1]=NF - 1;
for (i=2; i<=NF; i++) {
arr[$1 (l[$1]) (i-1)]=$i;
}
}
END {
for (i=0; i<ac; i++) {
...

You can use variables inside awk.

date_var=$(date +%Y_%m_%d_%H_%M)

nawk -v VM="$date_var" '{ print "Hello_"VM"-"$0}' $A1 > $l1

awk 'BEGIN{split("_4A_1,_4B_1,_4C_1",a,",")}{s=$0".txt";print "Hi\nWelocme" > s;
for(i=1;i<=3;i++){print "File Name ="$0""a[i]" no=10" > s}
print "Bye" > s}' file

If you want to create in...

awk -v f2='X Y Z' 'BEGIN{split(f2,f2a)} {for(i=1;i<NF;i++) print $1,f2a[i], $(i+1)}' myFile

try this not able to run, hope it gives output :)


awk '{ for(i=1;i<=3;i++){if(i==1){print $1,"X",$2}else {if (i==2) { print $1,"Y",$3} else{print $1,"Z",$4}}}}' file

Easy (but you have to add a condition if you want to manage lines with only one field):
awk '{for (i=2;i<=NF;i++) {print $1 OFS $i}}' A.txt > B.txt

Try egrep like this:
$ cat file
Hello -101 -102
Hi -101 -102 0
Hello 0 0
Bye -101 -90 0
Final 0 0
Hello -80
$ egrep 'Hello [-0]' file
Hello -101 -102
Hello 0 0
Hello -80

I think we're having a language barrier problem. Your egrep command line is looking for any line containing any of the following strings in you input file: "ock", "Ock", "oCk", "OCk", "ocK", "OcK",...

Does this work?


grep -i '(OCK)' $fileA.txt> $fileB.txt
Guessing that the OP doesn't realise the significance of parens in an ERE.

Please check your field positions.
Cheers,
Ranga:-)

awk '{ if($3 == $5) $(NF+1)="Yes"; else $(NF+1)="No" }' input > output