AWK split file every 50 ocurrences of a string

I have a beginner question about awk.

I am using the line below to split a file into c files, using 'MATCH' as my delimiter.

awk 'BEGIN{flag=0} /MATCH/{flag++;next} {print $0 > (flag ".txt")}' file

My file is very long, but it has the form shown below:

MATCH
a
b
c
d
MATCH
a
b

I want to have the above awk line split my file every 50 'MATCH' ocurrences. The current command creates a new file for each 'MATCH' ocurrence. I am sure there is a simple way to achieve this, but I have not figured it out yet. I have tried using the line below with no luck.

awk 'BEGIN{flag=0} /MATCH/{flag++ == 50;next} {print $0 > (flag ".txt")}' file

I appreciate the help and guidance.

Untested, using any awk:

awk '
    /MATCH/ && ( ( (++matchCnt) % 50 ) == 1 ) {
        close(out)
        out = (++outCnt) ".txt"
    }
    { print > out }
' file

Assumptions:

• the number of lines in a MATCH block are not known beforehand
• the number of lines in a MATCH block could vary
• the MATCH lines are to be copied to the output files

Sample input with 9 MATCH blocks:

$ cat file
MATCH
1.1
1.2
MATCH
2.1
2.2
MATCH
3.1
3.2
MATCH
4.1
4.2
MATCH
5.1
5.2
MATCH
6.1
6.2
MATCH
7.1
7.2
MATCH
8.1
8.2
MATCH
9.1
9.2

One awk idea:

awk -v blkcnt=3 '                             # for OP case set blkcnt=50
BEGIN   { outfile= ++fcnt ".txt" }
/MATCH/ { if (++matchcnt > blkcnt) {
             close(outfile)
             outfile= ++fcnt ".txt"
             matchcnt=1
          }
       #  next                                # uncomment if the "MATCH" lines are *NOT* to be copied to the output files
        }
        { print $0 > outfile }
'  file

For blkcnt=3 this generates:

$ head -40 {1..3}.txt
==> 1.txt <==
MATCH
1.1
1.2
MATCH
2.1
2.2
MATCH
3.1
3.2

==> 2.txt <==
MATCH
4.1
4.2
MATCH
5.1
5.2
MATCH
6.1
6.2

==> 3.txt <==
MATCH
7.1
7.2
MATCH
8.1
8.2
MATCH
9.1
9.2

For blkcnt=4 this generates:

$ head -40 {1..3}.txt
==> 1.txt <==
MATCH
1.1
1.2
MATCH
2.1
2.2
MATCH
3.1
3.2
MATCH
4.1
4.2

==> 2.txt <==
MATCH
5.1
5.2
MATCH
6.1
6.2
MATCH
7.1
7.2
MATCH
8.1
8.2

==> 3.txt <==
MATCH
9.1
9.2

If I've understood correctly, the first 50 blocks of a,b,c,d lines should be written to 1.txt, the next 50 to 2.txt and so on.

This can be achieved by building the filename from the integer value of (flag/50) and adding 1 to it (assuming you want the file series to being with 1 and not 0).

The BEGIN block can be removed as variables are set to 0 when first created if no value is given and they are used numerically.

Thus the following should achieve the desired output:

awk '/MATCH/{flag++;next} {print $0 >(int(flag/50)+1 ".txt")}' file

so while this isn't a complete solution, it does showcase how to capture each group of rows with each "MATCH", so once you count off every 50, then print them out in one shot, bearing in mind one needs to trim out the tail "MATCH" and save it for the next round

nice jot 53 | mawk 'NR % 6 != 1 || ($!NF = "MATCH")^_' | 

> mawk '{ printf(" :: input row(s) = %8u\n ::"
> " output row # = %8u\n "
> "-------------------\n %s%s "
> "----END-NEW-ROW----\n\n", NF^!!NF, NR, $!(NF = NF), ORS)
>
> }' RS='(^)?MATCH\r?\n' ORS='MATCH\n' FS='\n' OFS='\f'

 :: input row(s) =        1
 :: output row # =        1
 -------------------
 MATCH
 ----END-NEW-ROW----

 :: input row(s) =        6
 :: output row # =        2
 -------------------
 2
  3
   4
    5
     6
      MATCH
 ----END-NEW-ROW----

 :: input row(s) =        6
 :: output row # =        3
 -------------------
 8
  9
   10
     11
       12
         MATCH
 ----END-NEW-ROW----

 :: input row(s) =        6
 :: output row # =        4
 -------------------
 14
   15
     16
       17
         18
           MATCH
 ----END-NEW-ROW----

 :: input row(s) =        6
 :: output row # =        5
 -------------------
 20
   21
     22
       23
         24
           MATCH
 ----END-NEW-ROW----

 :: input row(s) =        6
 :: output row # =        6
 -------------------
 26
   27
     28
       29
         30
           MATCH
 ----END-NEW-ROW----

 :: input row(s) =        6
 :: output row # =        7
 -------------------
 32
   33
     34
       35
         36
           MATCH
 ----END-NEW-ROW----

 :: input row(s) =        6
 :: output row # =        8
 -------------------
 38
   39
     40
       41
         42
           MATCH
 ----END-NEW-ROW----

 :: input row(s) =        6
 :: output row # =        9
 -------------------
 44
   45
     46
       47
         48
           MATCH
 ----END-NEW-ROW----

 :: input row(s) =        5
 :: output row # =       10
 -------------------
 50
   51
     52
       53
         MATCH
 ----END-NEW-ROW----