Searching inforamtion in a text file.

[timely2]

I am wanting to do a search for instances of a given string but I want more then just the line its on, but may three entry lines before and after each time I find the string. The file I look through is an over 5,000,000 line log. I only get the line the string is on with a basic grep and doing a tail and head takes alot of work and time.

Thanks,
Tim Ely

 

[motoslide]

This has been handled before: Here's a reference that should get you started:

faq822-6173

Hope that helps. If that doesn't help, check out the Unix Scripting forum for other examples.

"Proof that there is intelligent life in Oregon. Well, Life anyway.

 

[timely2]

I tried the script they had but I kept getting a bad number error for the math part. I can get the part that does the grep to work and I could get the tail / head part to work but when I tried to have it look at every instance with the number of lines it would error out with the bad number.

 

[taupirho]

If you have Perl, try the following script out:-


#!/usr/bin/perl

# Usage: cgrep [-lines] pattern [files]

$context = 3;

# They might want more or less context.

if ($ARGV[0] =~ /^-(\d+)$/) {
$context = $1;
shift;
}

# Get the pattern and protect the delimiter.

$pat = shift;
$pat =~ s#/#\\/#g;

# First line of input will be middle of array.
# In the eval below, it will be $ary[$context].

$_ = <>;
push(@ary,$_);

# Add blank lines before, more input after first line.

for (1 .. $context) {
unshift(@ary,'');
$_ = <>;
push(@ary,$_) if $_;
}

# Now use @ary as a silo, shifting and pushing.

eval <<LOOP_END;
while (\$ary[$context]) {
if (\$ary[$context] =~ /$pat/) {
print "------\n" if \$seq++;
print \@ary,"\n";
}
\$_ = <> if \$_;
shift(\@ary);
push(\@ary,\$_);
}
LOOP_END

 

[feherke]

Hi

If you accidentally have GNU [tt]grep[/tt] :

-A NUM, --after-context=NUM
Print NUM lines of trailing context after matching
lines. Places a line containing -- between con­
tiguous groups of matches.

-B NUM, --before-context=NUM
Print NUM lines of leading context before matching
lines. Places a line containing -- between con­
tiguous groups of matches.

-C NUM, --context=NUM
Print NUM lines of output context. Places a line
containing -- between contiguous groups of matches.

Feherke.

http://rootshell.be/~feherke/

 

[timely2]

Not sure on the perl how to read that but I was messing with the grep and I found that it does not want to see: grep -n "test" file.log | cut -d: f1

as a number. I broke it up into diffrent parts and seemed to work until you tried to use that part to do math.

Thanks

 

[timely2]

Oh, another part that could effect this is multiple lines will come back when I do a serch.

 

[cspilman]

Maybe something like this:

for line in `awk '/[i][b]search_string[/b][/i]/{START=NR-3 ; FINISH=NR+3} END {while(START <= FINISH){print START ; START++}}' [i][b]filename[/b][/i]`
 do
        cat -n testfile | grep "${line} " 
 done

Keep in mind that I am no programmer and that there are much better ways to do this.

Regards,
Chuck