I have a file with many lines, like this:
one 3456 orange
two 6690 grape
three 5570 apple
four 6683 pear
five 8847 cherry
six 9035 banana
So I write a awk script to catch this output:
apple banana cherry
It looks like this:
awk '/three/ { a = $3}; /six/ { b = $3}; /five/ { c = $3} END {print a" " b" "c}' <file
But this doesn't look the best way becuase I keep using $3 - how do I catch this variable to reuse it?
There's nothing particularly wrong with doing it the way you are. If you use a variable, it will have to be repeated. However, you could use an array to accumulate your values.
Here is an example of using a simple variable:
awk '{var = $3} /three/ {a = var}; /six/ {b = var}; /five/ {c = var} END {print a, b, c}' file
Here is an example using an array:
awk '$1 ~ /^(three|six|five)$/ {arr[++c] = $3} END {for (i = 1; i <= c; i++) {printf "%s%s", delim, arr[i]; delim = OFS}; printf "\n"}' file
You don't need to use redirection. AWK can accept filenames as arguments.
In a print statement, commas substitute the value of OFS (by default a space) so you don't need to use " ".
In the array version, you can easily change the regex since it's all in one place.
banana banana banana
c should be i as in arr[i] in the END block. I'll edit my answer.apple cherry banana - is it possible using array to print in the correct order?