This command works:
csplit really_big_file.txt -f ../dump/really_big_file_ /^H\|756\|/ {*}
but this command:
gzip -dc really_big_file.txt.gz | csplit -f ../dump/really_big_file_ /^H\|756\|/ {*}
yields:
csplit: cannot open '/^H|756|/' for reading: No such file or directory
How to pipe output of gzip into csplit?
In man csplit
it says that file always comes before pattern:
SYNOPSIS
csplit [OPTION]... FILE PATTERN...
So it should be:
gzip -dc really_big_file.txt.gz | csplit -f ../dump/really_big_file_ - /^H\|756\|/ {*}
Example:
$ gzip -dc inputfile.txt.gz
abc
searchstring
def
searchstring
egh
searchstring
$ mkdir split
$ gzip -dc inputfile.txt.gz | csplit -f split/file - /searchstring/ {*}
$ ls -Al split/
total 16
-rw-r--r-- 1 ja users 4 Sep 21 17:53 file00
-rw-r--r-- 1 ja users 17 Sep 21 17:53 file01
-rw-r--r-- 1 ja users 17 Sep 21 17:53 file02
-rw-r--r-- 1 ja users 13 Sep 21 17:53 file03
$ cat split/*
abc
searchstring
def
searchstring
egh
searchstring