I have lists i can export of detailed call records for customers and it shows up in 3 columns, col1 end time, col2 call-id, col3 start time.

some lists are 800k lines long.
I’ve written a script that does the job but it’s slow… very slow.

I’m looking at a way to improve the run speed.

This works, but it’s very slow.

#!/bin/bash
IFS=$'n';
starts=(); 
ends=(); 
cids=(); 
#check if cdr.tmp is the same as cdr.txt, if not, condition the .txt file for use.
if [ "$(cat cdr.tmp)" != "$(cat cdr.txt)" ]; then 
  >cdr.tmp
  if [ ! -f cdr.tmp ]; then
    touch cdr.tmp
  fi; 
  line=0; 
  for i in $(cat cdr.txt); do
    echo working on line $line;
    random=$(tr -dc A-Za-z0-9 </dev/urandom | head -c 36; echo); 
    echo $i | sed "s/missing_externaltrackingid/$random/g" >> cdr.tmp; 
    ((++line)); 
  done; 
  sed -i 's/ 2024,//g' cdr.tmp;
  cat cdr.tmp > cdr.txt;
fi; 
#timestamp to epoch
ts2epoch() {
  orig="$1"
  orig="$(echo $orig | sed 's/th//g')";
  orig="$(echo $orig | sed 's/rd//g')";
  orig="$(echo $orig | sed 's/nd//g')";
  orig="$(echo $orig | sed 's/1st/1/g')";
  orig="$(echo $orig | rev | cut -d '.' -f2 | rev)";
  orig="$(echo $orig | tr -d '"')";
  epoch=$(date -d "${orig}" +"%s")
  echo "$epoch"
}; 
#epoch to timestamp
epoch2ts() {
  echo $(date -d @$1 +%Y-%m-%d_%H:%M:%S)
}; 
#add pertinent values to arrays.
for line in $(tac ~/cdr.txt); do 
  id=$(echo "$line" | cut -d ',' -f2); 
  start=$(echo "$line" | cut -d ',' -f3); 
  end=$(echo "$line" | cut -d ',' -f1); 
  cids+=("$id"); 
  echo adding $id to cids array >> ~/bin/concurrent.log; 
  starts+=($(ts2epoch "$start")); 
  echo adding $(ts2epoch "$start") to start times >> ~/bin/concurrent.log; 
  ends+=($(ts2epoch "$end")); 
  echo adding $(ts2epoch "$end") to end times >> ~/bin/concurrent.log; 
done; 
#calculate earliest time and latest time
begin_ts="$(echo ${starts[0]})"; 
end_ts="$(echo ${ends[-1]})"; 
#create a sequence of times with 30 second intervals, then for each interval, check to see how many calls were active during that instance
for time in $(seq $begin_ts 30 $end_ts); do  
  active=0; 
  for i in ${!cids[@]}; do 
    if [ $time -gt ${starts[i]} ] && [ $time -lt ${ends[i]} ]; then
      ((++active)); 
    fi; 
  done; 
  echo $(epoch2ts $time):$active active. >> ~/bin/concurrent.log
  echo $(epoch2ts $time):$active active. >> cdr-active.txt;
done;