Search Results

Use D in command mode. It deletes the rest of the line from the cursor position.
Command C does the same but then switches to insert mode after deleting rest of the line so that you can start...

One solution using awk, without converting the original input lines into intermediate format.

awk -F"_" '{++a[$1$3]} END{for(i in a){print i" "a[i]}}' input_file

Output:
hotbbb 2
coldbbb 1...

Not exactly. To explain further, following is the input data.

1,775,947,633
4,738
7,300
16,610
15,279
0
0

Expected output is (sum of all the numbers):
1775991560

I am trying to...

I need to sum up the values in field nr 5 in a data file that contains some file listing. The 5th field denotes the size of each file and following are some sample values.

1,775,947,633
4,738...

Or you may use negation in your original command as follows:

awk '!NR%2{$0="some text "$0}1' file

grep -v "OK" is another option to exclude lines with string "OK".

In your awk command, try printing entire line instead of just first field using just print or print $0

Thanks for your inputs. I tried the following and it worked. I needed to read from variable and not from a file.


RECORD="F1~F2~F3"
echo ${RECORD} | IFS="~" read VAR1 VAR2 VAR3
echo $VAR1 $VAR2...

I wanted to use specific variable names for better readability. For example, the actual field name is more readable than an array name with index. Hence did not go for array.

Another approach I...

A record contains 50 fields separated by "~". I need to assign each of these fields to different variables. Following is the shell script approach I tried.

...

Following code snippet would help.


DB=$(expr substr $1 1 1)
case $DB in
'B') DIR=/database/banking
;;
'R') DIR=/database/retailer
;;
*) echo error
;;

I'm on AIX. I have triggered an infinite loop process (to keep looking for input file availability for further process). At present only I can kill the process.

In case my colleague wants to kill...

Are you on solaris? If so then try nawk

Just another way


if [ $(echo $amt - $opn_amt | bc | grep ^-) ]; then
echo "correct"
else
echo "Wrong"
fi

Change your awk to

awk '{a[NR]=$3;b[NR]=$4;row[NR]=$0} END {for (i in a) printf("%s\t%.3f\n", row[i],(1-(b[i]/(7*24*60*a[i]))))}' infile

Or you can modify your awk statement to as below

awk '{for(i=2;i<=NF;i++) {if (substr($i,1,1)=="#") next; else print $1,$i}}'

awk -F"," '{OFS=",";$++NF="unix"}1' file1 file2


or

awk '{print $0",unix"}' file1 file2

Could you paste the entire script? Ensure you have the hashbang line at the beginning to denote the shell.. Example: #!/usr/bin/ksh

Save it as a script -> .ksh or .sh and execute it

Change only the directory name assigned to variable DIR. Leave the remaining as is. substr should be on the line and hence $0 for awk to read the line.

@ Akshay, you are right. Good catch. But I'm not sure if it is a valid scenario to have a variable like lck_

I presume the variable TABLE would vary and will be attached to lck_ as in


lck_T1...

DIR=/directory/with/2000/files
for FILE in $(ls $DIR)
do
echo "insert into tabella ('"$(awk '{print substr($0,1,7)}' $DIR/$FILE)"','"$(cat $DIR/$FILE)"');"
done


Beware of ' appearing within...

Could the process id change between two statements?

Try using a file with fixed file name such as /tmp/hosts.tmp instead of /tmp/$$

Could you please add more details on your statement
where the first column is a substring of the string of file.

What is the rule to fetch the substring? Is it first n characters or first 2...

Replace the if statement with

if [ "$(echo eval '$lck_'$TABLE)" != "" ]; then

sed 's/ | /,/g' infile

DIR=/your/directory
for FILE in $(ls $DIR)
do
OLD_DT=$(echo ${FILE##*_} | cut -d. -f 1)
NEW_DT=$(expr substr $OLD_DT 5 4)$(expr substr $OLD_DT 1 4)
RENAMED_FILE=$(echo...