Awk: Splitting file on nth occurence of delimiter, wrong first split file

Question

I want to split a text file like the one pasted below (sorry for the length), on every n occurence of ">". For example, every 2nd occurrence of ">", but I need to able able to change that number.

test_split.txt:

>eeefkdfn
a
a
a
>c 4ufjdhf
b
b
b
b
>
c
c
> c
d
d
d
d
d
>3
>cr
>c3
e
e
e
e
e
> 5
f
f
f
f
>cr
g
g
g
g
> cr dkjfddf
h
h
h
h

So I want to have output files this these (only showing the first two):

file_1.txt:

>eeefkdfn
a
a
a
>c 4ufjdhf
b
b
b
b

file_2.txt:

>
c
c
> c
d
d
d
d
d

etc.

Question:

I have been trying to achieve that result using this awk command:

awk '/^>/ {n++} { file = sprintf("file_%s.txt", int(n/2)); print >> file; }' < test_split.txt

And instead of the desired result, I am getting correct output (split) files, except for the first one, which only contains one occurence of ">" (instead of two), like this:

cat test_0.txt

>eeefkdfn
a
a
a

cat test_1.txt

>chr1 4ufjdhf
b
b
b
b
>
c
c

Any idea why that is? Thank you!

Answer 1

This seems more simple:

awk 'BEGIN{i=1}/^>/{cont++}cont==3{i++;cont=1}{print > "file_"i".txt"} file

Will gives you the expected result:

$ cat file_1.txt
>eeefkdfn
a
a
a
>c 4ufjdhf
b
b
b
b

$ cat file_2.txt
>
c
c
> c
d
d
d
d
d

Explanation

BEGIN{i=1} : File counter initialization.

/^>/{cont++} : To count every > found.

cont==3{i++;cont=1} : To increase the file counter and initialize the cont var every third appearance of the > char which becomes first again.

{print > "file_"i".txt"} : Direct the output to the expected file.

Answer 2

You can use this awk for dynamic control over number n where file will be split on nth occurrence of > in input data:

awk -v n=2 'function ofile() {
   if (op)
      close(op)
   op = sprintf("file_%d.txt", ++p)
}
BEGIN {
   ofile()
}
/>/ {
   ++i
}
i > n {
   i=1
   ofile()
}
{
   print $0 > op
}
END {
   close(op)
}' file

Here is one liner in case you want to copy/paste:

awk -v n=2 'function ofile() {if (op) close(op); op = sprintf("file_%d.txt", ++p)} BEGIN{ofile()} />/{++i} i>n{i=1; ofile()} { print $0 > op }' file