[SOLVED] remove a specific word using sed

Issue

I have a simple question

I have this file :

[NOC_S_Z_1]
    callerid=NULL
    secret=NOC_S_Z_1
    context=CallCenterGWs
    type=peer
    host=dynamic

[NOC_S_A_2]
    callerid=NULL
    secret=NOC_S_A_2
    context=CallCenterGWs
    type=peer
    host=dynamic
[5154]
    callerid=<5154>
    secret=5154
    context=SupportGroup
    type=friend
    host=dynamic

[5155]
    callerid=<5155>
    secret=5155
    context=SupportGroup
    type=friend
    host=dynamic
[PRIGW]
    callerid=NULL
    secret=PRI18865
    context=CallCenterGWs
    type=peer
    host=192.168.2.95

I want to use sed or awk command to look for this word callerid=NULL and remove it , but here is the trick , i want it to remove the line above it and the 4 lines below it ! it's a very large file but it's all the same type and the callerid=NULL is randomly sorted in the file

I managed to remove only the callerid=NULL but this is not helping

Any idea would be helpfull.

Solution

Assuming [ always denotes start of new block I would harness GNU AWK for this task following way, let file.txt content be

[NOC_S_Z_1]
    callerid=NULL
    secret=NOC_S_Z_1
    context=CallCenterGWs
    type=peer
    host=dynamic

[NOC_S_A_2]
    callerid=NULL
    secret=NOC_S_A_2
    context=CallCenterGWs
    type=peer
    host=dynamic
[5154]
    callerid=<5154>
    secret=5154
    context=SupportGroup
    type=friend
    host=dynamic

[5155]
    callerid=<5155>
    secret=5155
    context=SupportGroup
    type=friend
    host=dynamic
[PRIGW]
    callerid=NULL
    secret=PRI18865
    context=CallCenterGWs
    type=peer
    host=192.168.2.95

then

awk 'BEGIN{RS="["}NR>1&&!/callerid=NULL/{printf "[%s",$0}' file.txt

gives output

[5154]
    callerid=<5154>
    secret=5154
    context=SupportGroup
    type=friend
    host=dynamic

[5155]
    callerid=<5155>
    secret=5155
    context=SupportGroup
    type=friend
    host=dynamic

Explanation: I inform GNU AWK that row separator is [ then I use printf to output records excluding first one (NR>1, empty string before first [) and those with callerid=NULL inside (negated regular expression). I do prepend [ to record as it is treated as row separator, i.e. not included in record.

(tested in gawk 4.2.1)

Answered By - Daweo

Answer Checked By - Senaida (WPSolving Volunteer)