Finding lines matching the Pattern and their previous lines in a file

Shell Programming and Scripting
1

Hi,

I am trying to locate the occurences of certain pattern like 'Possible network disconnect' in a text file. I can get the actual lines matching the pttern using:

grep -w 'Possible network disconnect' file_name.

But I am more interested in getting the timing of these events which are located on the previous lines of those which match the pattern.

I tried some other techniques like:

while read LINE do
....
....
....
done< file_name

But could not extract the desired lines from the text file.

Can anyone help me accomplish desired results?

Thanks a million in advance.

2

Adjust parameter after -B, depending on how many lines before match you want to print:

grep -B5 -w 'Possible network disconnect' file_name
3

Hi bartus11,

Thanks for the reply, but my OS doesn't support -B option to grep.

Mine is SunOS 5.8 Generic_117350-45 sun4u sparc SUNW,Sun-Fire-V440

Regards,
Sagar

4

This will print matching line, then n lines that were before (in random order):

awk -vn=5 '{a[NR%n]=$0}/Possible network disconnect/{for (i in a) print a}' file_name
5

This one prints 5 lines before the pattern + the pattern itself:

awk '/Possible network disconnect/ { print NR }' file_name | while read line
do
awk '{ if(NR >= '$(($line - 5))' && NR <= '$line') print }' file_name
done
6

@bartus, could you please explain your code ?

a[NR%n] : what does the % stand for : is it the modulo operator ?

when you write
for (i in a) print a
[i]what is the initial i value : 1 ? or 0 ?

7 
nawk '{A[NR]=$0; delete A[NR-n]}/Possible network disconnect/{for (i=NR-n;i<=NR;i++) if(A) {print A; delete A}}' n=5 infile
8

This will print without the preceeding lines in random order (lol at that):

awk -vn=5 '/Possible network disconnect/{for (i in a) {c++;c%=n; print a[c]} print} {a[c=NR%n]=$0} ' file_name