[SOLVED] Shell script to convert values on successive lines into consecutive ranges?

kmkocot · 07-07-2010, 11:54 AM

Hi all,

I have a series of input files formatted like this:

Code:

RTREVF, KOG3266 = 111
RTREVF, KOG3294 = 130
RTREVF, KOG3295 = 177
WAGF, KOG3307 = 107
JTTF, KOG3320 = 174

Each line represents a portion of a data matrix. I want to convert the numbers after the "=" to the range of that partition in the matrix such that the output file looks like this:

Code:

RTREVF, KOG3266 = 1-111
RTREVF, KOG3294 = 112-241
RTREVF, KOG3295 = 242-418
WAGF, KOG3307 = 419-525
JTTF, KOG3320 = 526-699

Can anyone think of an easy way to do this?

Thanks,
Kevin

ntubski · 07-07-2010, 12:21 PM

Code:

awk -vn=0 -F= '{printf("%s= %d-%d\n", $1, n+1, n+=$2);}' < input-file > output-file

grail · 07-07-2010, 08:39 PM

You can just use the print in this case too:

Code:

awk -F= '{print $1"+ "(n+1)"-"(n+=$2)}' infile

Much of a muchness I guess

kmkocot · 07-08-2010, 06:11 PM

Wow! Thanks guys! Both work great!

syg00 · 07-08-2010, 06:38 PM

No concerns if the first record is "1" (or worse, "0") ?.
Corner cases are always a potential problem.

kmkocot · 07-09-2010, 10:59 AM

That's a good thought that I hadn't considered but a previous step in the pipeline throws out any sequences that are less than 25 characters long.

Thanks!
Kevin