Description

Each complete record consists of 3 consecutive lines whose tags are Header:, Category:, and Title:, respectively. When processing a datafile, we want to retrieve complete records and discard lines not belonging to any complete record.

Raw Input

Desired Output

Header: 1
Category: Algorithm
Title: Algorithms in C
Header: 2
Header: 3
Title: Making TeX Work
Category: Publication
Header: 4
Category: Shell Programming
Title: Learning the Bash Shell

Header: 1
Category: Algorithm
Title: Algorithms in C

Header: 4
Category: Shell Programming
Title: Learning the Bash Shell

Script and Comments

Script₁

[ 1] /^Header:/!d
[ 2] $!N
[ 3] /\nCategory:/!d
[ 4] $!N
[ 5] /\nTitle:/!d
[ 6] s/$/\n/

Comments

Only lines belonging to a complete record can survive Steps [1], [3] and [5].
Step [6] adds a newline character after a complete record.