[SOLVED] Parse data

mzh · 11-06-2012, 03:50 AM

Hi
How would you parse the following data in order to extract the "cycle" and the "heat"?

Code:

CYCLE:  3022 TIME: 429.654 TIME LEFT:  3.92D  GRAD.:     3.170 HEAT:-21965.53
CYCLE:  3023 TIME: 335.979 TIME LEFT:  3.92D  GRAD.:     3.817 HEAT:-21965.80
RESTART FILE WRITTEN,      TIME LEFT:  3.91D  GRAD.:     4.586 HEAT:-21965.55
CYCLE:  3025 TIME: 383.594 TIME LEFT:  3.91D  GRAD.:     2.877 HEAT:-21965.67
CYCLE:  3026 TIME: 308.740 TIME LEFT:  3.91D  GRAD.:     3.230 HEAT:-21966.36

What I'm after is a way of generating a list of x/y pairs, such as:

Code:

3022 -21965.53
3023 -21965.80
3024 -21965.55

I would preferably use Bash, but probably this requires a programming language.

millgates · 11-06-2012, 03:55 AM

May I ask how did you get the "3024" value in your output? It's not in your input file.

colucix · 11-06-2012, 04:06 AM

A simple awk one-liner should do the trick:

Code:

awk '/CYCLE/{sub(/HEAT:/,"",$NF); print $2, $NF}'

pixellany · 11-06-2012, 07:09 AM

Code:

sed -r -n '/CYCLE/s/CYCLE:  ([0-9]*).*HEAT:(-[0-9]*\.[0-9]{2})/\1 \2/p'

(not totally robust---eg assumes that we always have ": " after CYCLE)

mzh · 11-07-2012, 02:33 AM

Quote:

Originally Posted by millgates

May I ask how did you get the "3024" value in your output? It's not in your input file.

Sorry I was not explicit on this. You're right, this is in part exactly the question, "how to get the 3024"?

millgates · 11-07-2012, 02:54 AM

Well, I don't kno how the file is organized exactly, but I would try to modify colucix's example like this:

Code:

awk '/CYCLE:/{c=$2-1}/HEAT:/{c++;sub(/HEAT:/,"",$NF);print c,$NF}'

mzh · 11-07-2012, 11:30 AM

Quote:

Originally Posted by millgates

Well, I don't kno how the file is organized exactly, but I would try to modify colucix's example like this:

Code:

awk '/CYCLE:/{c=$2-1}/HEAT:/{c++;sub(/HEAT:/,"",$NF);print c,$NF}'

hm, yes that counter solves it thanks