Print between patterns -awk - perl

Question

I have the following input file
-------------------------------------
--- A-TKE spectrum (post-neutron)---
-------------------------------------
2-dim. array: (A = 73 To 162 Step 1) (E = 122 To 198 Step 1)
(The data are written according to the loop structure specified above.
The last loop is the inner-most one. Line breaks are not related to the data structure!
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 1 1 0 0 0 1 0 0 2 0 0 1 0 0 0 0
</A_TKE>
</Results>
This file is part of a bigger file, which contains more number block similar to the above one.
What I want to do is print in a seperate file the numbers that are within the file. As a first try, I thought "Let's print what is between two patters". The START pattern will be --- A-TKE spectrum (post-neutron)--- and the END pattern will </A_TKE>
I used
awk '/--- A-TKE spectrum (post-neutron)---/{flag=1;next}/</A_TKE>/{flag=0}flag{print}' input
To my surprise nothing is printed on the screen. Any idea on why is this happening?
I would also like to now if I can substitue next with a number that will correspond to the line that I want to print after the given pattern.

Answer 1

In perl (because whilst you start with awk, this is tagged as perl too):
perl -ne 'print if m/post-neutron/...m/A_TKE/' input
Which will do what your thingy was trying to.
It works by using the range operator - which is a perl construct which tests if the current file is between two markers.
And the fact that perl can run inline in the same way as awk or sed. -ne says "wrap this code in a while loop" that iterates a line at a time. (e.g. as sed/awk/grep do by default).
Otherwise ... it depends a little on what else is in your file. Because from the sample you gave - the only lines starting with a number are the ones you want. So:
perl -ne 'print if m/^\s*\d+\s*\d+/'
(Or combine the two tests:
perl -ne 'print if m/post-neutron/...m/A_TKE/ and m/^\s*\d+\s*\d+/'
Which will output anything between the two text markers that has number-space-number at the start of line.

Answer 2

Your awk script is almost there. I'd build it like this:
awk '
/<\/A_TKE>/ {exit}
flag && NF==50 {print}
/--- A-TKE spectrum .post-neutron.---/ {flag=1}
' input
You could optimize it of course, but just looking at the logic of it, the idea is that you need do no further processing once you hit your end marker .. and you only want to start printing AFTER your start marker.
Note that I've replaced the brackets in your start marker with dots, since brackets are interpreted as PART OF THE REGULAR EXPRESSION. I believe you may be considering these simple strings to search for, but they are regex.

Answer 3

Some working ideas
Awk can print between two matching lines with this simple command:
awk '/PATTERN1/,/PATTERN2/' file.txt
If you want more control over what happens when the lines are found, you can use the flag approach.
awk '/A-TKE/{f=1;for (i=0;i<7;i++){getline}}/A_TKE/{f=0}f' file.txt
Adjust i<7 to specify how many lines you would like to skip.
Also, in this particular case, you can exploit the fact that you have 50 fields of numbers, which is unlikely to occur anywhere else in the file. Thus, you could also get the numbers of interest out with:
awk 'NF>40' file.txt
NF is a special awk variable that stores the number of fields (columns) in the current record (line).
Why your approach does not work
Your approach does not work, because everything that is between // in awk is treated as a regular expression. In your case you have several - and () in the string, which are not escaped, and thus assume their special meaning and therefore the expression you typed never matches, thus the flag is never set to 1 and awk does not print a line.

Answer 4

Try this simple sed command
sed -n '/A-TKE/,/<\/A_TKE>/p' FileName

Answer 5

If you want to extract rows with just numbers
awk 'NF && !/[a-zA-Z\-]+/' input.file
will generate
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 1 1 0 0 0 1 0 0 2 0 0 1 0 0 0 0
with patterns and smart counters you can write it as
awk '/\/A_TKE/{exit}
f&&s&&s--{next}
f&&NF{print}
/A-TKE spectrum/{f=1;s=7}' input.file
fine tune patterns as you like.