sed (or awk) to sum values in a csv file

问题

I have a csv file where the lines look like this:

foo,bar,baz,5,6,8
zab,rab,oof,2,2,3

I need a command to transform it summing the last 3 values:

foo,bar,baz,19
zab,rab,oof,7

If possible I need a one liner.

回答1:

Try this for fixed length :

awk -F, '{print $1,$2,$3,$4+$5+$6}' OFS=, file

And for not fixed length :

awk -F, '{for(i=1;i<NF-2;i++)printf("%s,", $i)}{print $NF+$(NF-1)+$(NF-2)}' file

sed is not the right tool for arithmetic.

回答2:

perl -F, -lane '$a+=$_ for(@F[3..5]);print join ",",@F[0..2],$a' your_file

small test below:

> echo "foo,bar,baz,5,6,8" | perl -F, -lane '$a+=$_ for(@F[3..5]);print join ",",@F[0..2],$a'
foo,bar,baz,19

来源：https://stackoverflow.com/questions/15548771/sed-or-awk-to-sum-values-in-a-csv-file