Selective extraction from patterned occurence (distant repeated lines) using awk or sed

Question

A file has a line pattern where a line that does not contains numbers occurs twice in the file.

Energy initial, next-to-last, final = 
 -87057.8200168321  -87208.7223900015  -87208.7197287879
Energy initial, next-to-last, final = 
 -87208.7197287879  -87208.7223900015  -87310.7197287879

I want to extract value -87310.7197287879 which is the value at the end of the line following the 2nd occurrence of the line that does not contain numbers.

How can I do this with Awk or Sed ?

Answer 1

awk '/^Energy initial, next-to-last, final =/ 
    { if (secondline==0) { secondline=1; next; } else { getline; print $3; } }' inputfile

-87310.7197287879

Answer 2

An alternative in case you have such patterned occurrence happening once or more than once in your file:

$ awk '/^Energy initial, next-to-last, final =/ {a[NR]=1} 
       {if(NR >= 3 && a[NR-2] == a[NR] && a[NR] == 1) {toprint=1; next}}
       {if (toprint) {print $3; toprint=0}}' infile

No matter, Hauke's solution is far less ugly.

Answer 3

Just a quicky:

> awk '{print $3}' /tmp/test.txt | grep -m1 -vE '[a-z]'
-87208.7197287879

The awk command prints the 3rd item ($3) from /tmp/test.txt and do a grep that does NOT match any letters ([a-z]) which eleminates the lines with text and stop after 1 selected lines (-m1)