Print Lines above and below string

[coding_new]

Hi Monks,

I have a question reqarding printing x lines above and below a string found in a file. Currently I have the following code which identifies the string and prints the string and one line below. How would we modify this to print lets say 5 lines above and below the string?

I have searched but cant seem to find anything that does both. Most of what I found print either lines after or before when a string variable is found in the file. I am also not sure the way I have it is the best way to move forward with... Thanks for your input.

#!/usr/bin/perl
use strict;

$hits=0;
$string_to_find="string";
$file = "filename";

open   (LOGFILE, $file);

while($line = <LOGFILE>) {
if ($line =~ /$string_to_find/i) {
$hits++;

print "$line";
print scalar <LOGFILE>;

}
}
close(LOGFILE);

[jethro]

Well, if you found examples that print lines before the hit and examples that print lines after the hit, you could just combine them

Usually to print lines before a hit, you create a line storage where the previous 5 lines are stored. When you find the string print the previous lines from storage. To print 5 lines after a hit, just set a counter to 5 when you find it and print lines as long as the counter (that you decrement) is greater than 0. Both methods can be combined without any problems

But the simple algorithms have one drawback: They will reprint lines if these are in the vicinity of two hits. This might be just what you want. If not, use the following method: Have a storage of 5 lines and a counter you set to 10. Always print from the storage, but compare the searchstring with the line you just read. If you get a hit set the counter to 10. Again print and decrement

while ($line = <LOGFILE>) {
  push @storage, $line);
  if ($line =~ /$string_to_find/i) {
    $counter=10;
  }
  if (@counter-- and @storage>=5) {
    print shift @storage;
  }
}

Untested

[thenaz]

So you want the context of the matched string. You will need to keep a rolling window of the last few lines you've seen. Actually, if your input isn't huge, just slurp up into an array, like so (untested):

open LOGFILE, "<$file";
@lines = <LOGFILE>;
for my $i (0..$#lines) {
  next unless $lines[$i] =~ /$string_to_find/i;
  my $a = $i - 5 < 0 ? 0 : $i - 5;;
  my $b = $i + 5 > $#lines ? $#lines : $i + 5;
  print $lines[$_i] for my $_i ($a..$b);
}

[coding_new]

Thanks for the input. Few questions. My file sizes will be around 2.5 megs max. Not sure if that's a problem or not. Also once I ran the script you provided, I got an error on the last line

Here is the error: syntax error at ./readlog5.pl line 12, near "$_i ("

[ww]

++ for "Close!" ...but "untested" lets typos slip by:

open (LOGFILE, "<$file") or die "Can't open $file";
my @lines = <LOGFILE>;
my $lines; my $i;
for $i(0..$#lines) {
    next unless $lines[$i] =~ /$string_to_find/i;
      my $a = $i - 5 < 0 ? 0 : $i - 5;;
    my $b = $i + 5 > $#lines ? $#lines : $i + 5;
    for $i($a..$b) {
        print $lines[$i];
    }
}

[Anonymous Monk]

what $#lines does? would anyone please me

[Anonymous Monk]

Here you go $#lines $# -> Where is $# fully documented?

[Samy_rio]

this is one of the way. try this

#!/usr/bin/env perl

use strict;

my $hits=0;
my $string_to_find="string";
my $file = "filename";

open   (LOGFILE, $file);
my @cont = <LOGFILE>;
close(LOGFILE);

for(my $i = 0; $i <= $#cont; $i++) {
my $line = $cont[$i];

if ($line =~ /$string_to_find/i) {
  my $st;
  ($i <= 5) ? ($st = 0) : ($st = $i - 5);
  my $ln = $i - 1;

  my $eln = $i + 1;
  my $en = $i + 5;
  ($en > $#cont) ? ($en = $#cont) : ();

  print @cont[$st..$ln];
  print $line;
  print @cont[$eln..$en];

}
}

[coding_new]

That code provided me with what I needed. Thanks for your help.

[Gulliver]

wgrep.pl as listed at this link will do it.

http://www.unix.com/solaris/33533-grep-display-few-lines-before-after.html

Or just use grep at the command line or qx() with the -B and -A options to specify the number of lines before and after.

[duyet]

On the command line: grep -n5 "search_string" filename

If you want to use it in your code

my $result = `/usr/bin/grep -n5 "search_string" $file`;
print $result;

[coding_new]

Thanks for the input, but when I tested out that code, it simply added 5 blank lines before and after my string.

[duyet]

Sorry i first thought you wanted to print empty lines. Try the "grep" command, as in my updated post