Yrurtj

I am trying to use bash to go through multiple directories (sims) and search for a given string, setting each index of an array to its relevant output, then print that output with comma delimiters for each value and adding a semicolon delimiter to separate each file. Effectively, this should give me a CSV that I can split twice in excel. All the important information is in field 3 each time it finds "Total Energy", which is found in the file named "output" in each directory.

With the set of directories that I am currently testing, each output file should have 2500 energies, but the code currently does not find that (vide infra).

Here is the code I have so far:

#/bin/bash/



saveIFS="$IFS"



#Step 1: Ask user for the range of sims they want

echo "What is the first sim?"

read simcount

echo "What is the last sim?"

read simend



#Step 2: Create the energy files with proper naming conventions and make sure they're empty

energies+="energies${simcount}-${simend}.csv"

fenergies+="final_energies${simcount}-${simend}.out"

touch $energies

touch $fenergies

< $energies

< $fenergies



#Step 3: Go through each directory, print all energies into proper files

while [ $simcount -le $simend ]; do

        echo $simcount

        cd $print'sim'$simcount                                 # Change to the directory of each specified sim

        energy=($(awk '/Total Energy/{ print $3 }' output))     # Print all energies from output into an array

        echo ${#energy[@]}

        fenergy=${energy[${#energy[@]}-1]}                      # Get the last energy in each file

        cd ../                                                  # Go up a directory

        IFS=","                                                 # Change the Internal Field Separator (IFS) to a comma

        echo "${energy[*]};" >> $energies                       # Expand the array of energies into an IFS-delimited list; print it into the new energies file

        echo "$fenergy" >> $fenergies                           # Put the final energy of each sim on a new line in the new final energies file

        ((simcount++))

done



IFS="$saveIFS"



exit 0

This gives the following output:

$ e.sh

What is the first sim?

6

What is the last sim?

15

6

2500

7

1

8

1

9

1

10

1

11

1

12

1

13

1

14

1

15

1

Which means that the loop is finding all 2500 energies the first time around, but each subsequent time through the loop it does not split the output of awk into an array. A representative example of the output this puts in the new file denoted as $energies:

-271.2872230353,-271.3198859908,-271.4166545741,-271.5362409096,-271.6700236287,-271.8068505329,-271.9076587286,...;

-273.2853761106

-273.2855419371

...

-273.2856368361

-273.2857720402

-273.2859963834;

-271.2872230353

-271.3198859908

-271.4166545741

...

To clarify, the first iteration of the loop is successful and outputs the array onto one line with a semicolon separator. All following iterations do not get split into an array (or have an array length of 1) and seem to repeat thousands of times before moving onto the next directory.

I have searched around for a while now, and I cannot figure out why this would happen. I have also tried unsetting energies at the end of each iteration to no avail. So my specific question(s) is(are): why would splitting awk output into an array work the first time through the loop but not any subsequent times? Are there better/more efficient ways of going about this using bash that would be worth checking out?