sum of numbers having different prefixes

[lnbnatera]

hi,

i have two columns as follows:

129 MB
23 KB
100 B
90 GB
1 KB
4 KB
34 MB
29 MB
10 GB

the usual thing that i do is "grep" out using the units to sort things out according to prefixes then add them (with respect to their prefixes. then add the 4 sums that i get.

something like
$ grep 'MB' file.txt > file-MB.txt
$ grep 'KB' file.txt > file-KB.txt
... and so on ...

then,
$ awk -f sum.awk file-MB.txt
<outputs a number>
$ awk -f sum.awk file-KB.txt
<outputs a number>
... and so on ...

then i add all those that i get to have a sum.

is a simpler way to do this in awk in just one shot? something like
$ awk -f new_sum_awk.awk file.txt
<sum>

thanks

 

[PHV]

A starting point for new_sum_awk.awk:
BEGIN{a["B"]=1;a["KB"]=1024;a["MB"]=1024*1024;a["GB"]=1024*1024*1024}
{t+=($1*a[$2])}
END{print "total",t}

Hope This Helps, PH.
Want to get great answers to your Tek-Tips questions? Have a look at FAQ219-2884 or FAQ181-2886