mjhjmtu

Given intput:

Via: 1.1.1.1 
not relevant line 
keyword + some text
...
not relevant line N
keyword + some text
...
not relevant line N
Via: 2.2.2.2
not relevant line 
keyword + some text
...
not relevant line N
keyword + some text
...
not relevant line N
Via: 3.3.3.3
not relevant lines
Via: 4.4.4.4
not relevant
Via: 5.5.5.5
not relevant line 
keyword + some text
...
not relevant line N
keyword + some text
...
not relevant line N
not relevant line N
...

Required output:

Via: 1.1.1.1 
keyword + some text A
keyword + some text A
Via: 2.2.2.2
keyword + some text B
keyword + some text C
Via: 5.5.5.5
keyword + some text D
keyword + some text E

keyword string can occur N times in any Via block, or may not occur at all. In the output I need only those Via blocks where keyword occurs together with keyword strings belonging to them. The closest answer I found is here, but I can't make it into what I need.

With sed:

sed -n '/^Via:/ x; /keyword/p; d; ; /keyword/H; $ x; /keyword/p; ' input.txt

Or, if you want keyword anchored at the beginning of line:

sed -n '/^Via:/ x; /nkeyword/p; d; ; /^keyword/H; $ x; /nkeyword/p; ' input.txt

Awk solution:

awk '/^Via:/ f=1; r=$0; kw=0; next 
 f && /keyword/ printf "%s%sn",(!kw)? r ORS:"",$0; kw++ ' file

• /^Via:/ - capturing line starting with Via: into r variable. Set flag f=1 into "active" state indicating processing of a certain Via block




• kw - flag denoting the number of "keyword" lines under each Via block




• f && /keyword/ - while processing lines under Via block - consider only lines matching keyword pattern