A perl programmer would prefer a HASH to represent the parsed data, rather than an array. Assuming that this is what you want, here is a one-liner to get you started:

>perl -E "my %h = map {split ':',$_} split /\s+/, $ARGV[0]; 
   say qq|$_\t$h{$_}\n|  for sort keys %h"
 "THREAD_ID:1bf1d698 CDR_TY
40815144127   DEL_TIME:240815144127   OA_ADDR:5.0.OTSDC   PRE_TRANS_OA
+:5.0.OTSDC  DA_ADDR:1.1.966555696176   PRE_TRANS_DA:1.1.966555696176 
+"
[download]

The data does have some inconsistencies that are not handled by the regex .. this is just to get you started. You will need to develop the regex to handle the pathological data.

Update:If you run into trouble parsing the pathalogical data, please post the code you tried here, and explain your problems.
Monks here will gladly explain and help correct code, provided you display some effort.