grepping lines above and below a pattern found
mike
say i have this in a text file
"Use this form to
post your messages.
Remember that
it can be viewed by
millions of people worldwide.
For questions about posting,
check our FAQ"
what is the most commonly used way to print ,say , 2 lines above and 2
lines below a certain pattern found such that my output is
"post your messages.
Remember that
it can be viewed by
millions of people worldwide.
For questions about posting, "
for the matched pattern "viewed" .
thanks...
Gunnar Hjalmarsson
> say i have this in a text file
>
> "Use this form to
> post your messages.
> Remember that
> it can be viewed by
> millions of people worldwide.
> For questions about posting,
> check our FAQ"
>
> what is the most commonly used way to print ,say , 2 lines above
> and 2 lines below a certain pattern found
Don't know which way is most common. If the file isn't too big, an
easy approach is to slurp all the lines into an array. Otherwise you
can read the file line by line in a while loop, and keep the two
latest lines saved temporarily.
Which ways have you considered?
--
Gunnar Hjalmarsson
Email: http://www.gunnar.cc/cgi-bin/contact.pl
Paul Lalli
Assuming the file is not so huge that putting it all in memory at once
will hurt you....
#!/usr/bin/perl
use strict;
use warnings;
{
local $/;
open FILE, 'file.txt' or die "Cannot open file: $!";
my $contents = <FILE>;
my $word = 'viewed';
my @matches = $contents =~ /((?:.*\n){2}.*\b$word\b.*\n(?:.*\n){2})/g;
foreach (@matches){
print "$_\n";
}
}
Paul Lalli
Abhinav
mike wrote:
Is this allowed ?
system "grep -A2 -B2 $pattern $file";
Thanks
Abhinav
David K. Wall
Might not work well on an OS without grep. :-)
I'm tired of the question, anyway. How's this for a fairly general
solution? The interface could probably be improved -- and I suppose
it would be easier to just use Tie::File and loop through the lines.
use strict;
use warnings;
print join "...\n",
lines_around(
qr/\bviewed\b/i,
*DATA,
{before => 2, after => 2}
);
sub lines_around {
# pattern in filehandle with options...
my ($pattern, $filehandle, $options) = @_;
$options->{before} = defined $options->{before}
? $options->{before}
: 1;
$options->{after} = defined $options->{after}
? $options->{after}
: 1;
# return matching line, default 1
$options->{match} = defined $options->{match}
? $options->{match}
: 1;
# return first match only, default 0
$options->{first_only} = defined $options->{first_only}
? $options->{first_only}
: 0;
my (@buffer, @matches);
while (my $line = <$filehandle>) {
shift @buffer if @buffer > $options->{before};
push @buffer, $line;
next unless $line =~ /$pattern/;
pop @buffer unless $options->{match};
my @matched_lines;
push @matched_lines, @buffer;
push @matched_lines, scalar <DATA>
for 1 .. $options->{after};
return join('', @matched_lines) if $options->{first_only};
push @matches, join '', @matched_lines;
@buffer = ();
}
return @matches;
}
__DATA__
Use this form to
post your messages.
Remember that
it can be VIEWED by
millions of people worldwide.
For questions about posting,
check our FAQ
Use this form to
post your messages.
Remember that
it can be VIEWED by
John W. Krahn
my $above = 2;
my $below = 2;
my @buffer;
while ( <> ) {
@buffer = ( @buffer, $_ )[ -( $above + 1 ) .. -1 ];
/viewed/ && print @buffer, map scalar <>, 1 .. $below;
}
John
--
use Perl;
program
fulfillment
Tad McClellan
------------------------------
#!/usr/bin/perl
use warnings;
use strict;
use Tie::File;
my $filename = 'text.file';
tie my @array, 'Tie::File', $filename, autochomp => 0 or
die "could not tie file $!";
foreach my $i ( 0 .. $#array ) {
print @array[ $i-2 .. $i+2 ] if $array[$i] =~ /viewed/;
}
untie @array;
------------------------------
--
Tad McClellan SGML consulting
ta...@augustmail.com Perl programming
Fort Worth, Texas
Brad Baxter
perl
-ne'/viewed/&&($x=$.);push@a,$_;$.>5&&shift@a;$x&&$.-$x>1&&last}{$x&&print@a'
file
Yes, it diverges from spec 2 lines from the end, but I didn't think
anybody would care. :-)
Brad
Clyde Ingram
"Tad McClellan" <ta...@augustmail.com> wrote in message
news:slrnc8dq1m...@magna.augustmail.com...
> mike <s999999...@yahoo.com> wrote:
>
> > what is the most commonly used way to print ,say , 2 lines above and 2
> > lines below a certain pattern found such that my output is
> tie my @array, 'Tie::File', $filename, autochomp => 0 or
Nice .. didn't know about this.
The OP might have some use for my inelegant meanderings below, on this
theme.
(Kindly forgive the overlong posting. I expect others have far more elegant
solutions too.)
Trying a "grep" on own source for pattern "[Tt]i." (running on Windoze XPee)
with:
trial_cgrep.pl [Tt]i. 2 trial_cgrep.pl
shows (with "+" marking matched records):
(2) :use warnings;
(3) :use strict;
(4)+:use Tie::File;
(5) :use English;
(6) :
(24) :chomp @file_list;
(25) :
(26)+:die "$nr_lines: nr_lines should be zero or positive integer"
(27) : unless ($nr_lines =~ /^\d+$/);
(28) :
(36) :for my $filename (@file_list) {
(37) :
(38)+: tie my @array, 'Tie::File', $filename, autochomp => 0 or
(39)+: die "could not tie file $!";
(40) :
(41) : # Build a look-up table of record numbers already printed, so that
we
(72) : }
(73) :
(74)+: untie @array;
(75) :}
(76) :
FWIW, the source is below.
Regards,
Clyde
#!/usr/bin/perl
use warnings;
use strict;
use Tie::File;
use English;
my $usage = "$PROGRAM_NAME pattern nr_lines file1 file2 file3 ...";
my $help = "
pattern: Search pattern.
nr_lines: Number of records to show, before and after matching
records.
file1 ...: List of files to search.
";
my $MIN_NR_ARGS_EXPECTED = 3;
my $nr_args_supplied = 0 + @ARGV;
die "Expected at least $MIN_NR_ARGS_EXPECTED arguments. " .
"Got only $nr_args_supplied.\n" .
"Usage: $usage\n$help" unless ($nr_args_supplied >=
$MIN_NR_ARGS_EXPECTED);
my $pattern = shift;
my $nr_lines = shift;
my @file_list = @ARGV;
chomp @file_list;
die "$nr_lines: nr_lines should be zero or positive integer"
unless ($nr_lines =~ /^\d+$/);
foreach (@file_list) {
die "$_: No such readable file"
unless (-r $_);
}
for my $filename (@file_list) {
tie my @array, 'Tie::File', $filename, autochomp => 0 or
die "could not tie file $!";
# Build a look-up table of record numbers already printed, so that we
# do not print the same record twice
my %printed_nr = ();
foreach my $i ( 0 .. $#array ) {
if ($array[$i] =~ /$pattern/) {
for my $j ($i-$nr_lines .. $i+$nr_lines) {
next if $j < 0; # Avoid falling off start of array
next if $j > $#array; # .. and end of array
next if $printed_nr{$j}; # Skip record if already printed
# Prepare print prefix for record being printed
my @prefix_parts = ();
# Omit the filename if only one supplied.
push (@prefix_parts, "$filename ") if (@file_list > 1);
# Record number. e.g. (42)
push (@prefix_parts, "(" . ($j+1) . ")");
# "+" indicator, if this record matches the pattern, else
SPACE
push (@prefix_parts, (($array[$j] =~ $pattern) ? "+" : "
" ));
my $prefix = join("", @prefix_parts) . ":";
print $prefix . $array[$j];
$printed_nr{$j} = 1;
}
}
}
untie @array;
}
ENDS
Anno Siegel
That's a good way to put Tie::File to use. It doesn't address two
problems that come into view when the basic functionality is there:
- Dealing with matches that are too close to the beginning or end of
the file to fill a whole window
- What to do with overlapping windows
The OP hasn't addressed these issues, and I don't want to suggest
solutions (some have been suggested in this thread). I just want
to point out that this solution, pretty as it is, isn't the whole
story.
Anno
Tad McClellan
>
> "Tad McClellan" <ta...@augustmail.com> wrote in message
> news:slrnc8dq1m...@magna.augustmail.com...
>> mike <s999999...@yahoo.com> wrote:
>>
>> > what is the most commonly used way to print ,say , 2 lines above and 2
>> > lines below a certain pattern found such that my output is
>
>> tie my @array, 'Tie::File', $filename, autochomp => 0 or
>
> Nice .. didn't know about this.
I knew about it because I re-read the Perl FAQs from time to time. :-)
How do I change one line in a file/
delete a line in a file/
insert a line in the middle of a file/
append to the beginning of a file?
> my $nr_args_supplied = 0 + @ARGV;
^^^
^^^
The LHS already ensures scalar context for @ARGV, so you don't need that.
> my $pattern = shift;
> my $nr_lines = shift;
> my @file_list = @ARGV;
my( $pattern, $nr_lines, @file_list ) = @ARGV;
> chomp @file_list;
You are expecting command line arguments with newlines in them?
> die "$_: No such readable file"
> unless (-r $_);
That should probably be:
die "$_: No such readable file"
unless (-f $_ and -r _);
Brad Baxter
>
> perl
> -ne'/viewed/&&($x=$.);push@a,$_;$.>5&&shift@a;$x&&$.-$x>1&&last}{$x&&print@a'
> file
>
> Yes, it diverges from spec 2 lines from the end, but I didn't think
> anybody would care. :-)
It also has a problem with overlapping windows.
perl
-ne'/viewed/&&\!$x&&($x=$.);push@a,$_;$.>5&&shift@a;$x&&$.-$x>1&&last}{$x&&print@a'
file
(\! to hide !$ from the shell ... YMMV)
Brad
Clyde Ingram
> I knew about it because I re-read the Perl FAQs from time to time. :-)
>
> > my $nr_args_supplied = 0 + @ARGV;
> ^^^
> The LHS already ensures scalar context for @ARGV, so you don't need that.
Ah, quite right.
> > my $pattern = shift;
> > my $nr_lines = shift;
> > my @file_list = @ARGV;
>
> my( $pattern, $nr_lines, @file_list ) = @ARGV;
Oh yes . . .
> > chomp @file_list;
>
> You are expecting command line arguments with newlines in them?
Er. . ., no
> > die "$_: No such readable file"
> > unless (-r $_);
>
> That should probably be:
>
> die "$_: No such readable file"
> unless (-f $_ and -r _);
Um, certainly. . .
Thanks for the improvements, Tad.
That was just a quick knock up script. (I did say it was inelegant.)
Always nice to be shot down so constructively.
Regards,
Clyde
David Combs
David K. Wall <dw...@fastmail.fm> wrote:
>Abhinav <matrix_...@yahoo.dot.com> wrote:
>
>> mike wrote:
>>
>>> say i have this in a text file
>>>
>>> "Use this form to
>>> post your messages.
>>> Remember that
>>> it can be viewed by
>>> millions of people worldwide.
>>> For questions about posting,
>>> check our FAQ"
>>>
>>> what is the most commonly used way to print ,say , 2 lines above
>>> and 2 lines below a certain pattern found such that my output is
>>>
>>> "post your messages.
>>> Remember that
>>> it can be viewed by
>>> millions of people worldwide.
>>> For questions about posting, "
>>>
>>> for the matched pattern "viewed" .
>>>
>>> thanks...
>> Is this allowed ?
>>
>> system "grep -A2 -B2 $pattern $file";
>
>Might not work well on an OS without grep. :-)
Might not work well on an OS without GNU grep. :-)
David
David Combs
John: Perhaps you could add a few comments explaining
your (as usual) nifty code?
Once you've done that, maybe it can be inserted
in to some tutorial part of the pod-files --
looks like it'd be useful, and educational, for
lots of people.
Thanks,
David
Anno Siegel
[...]
> my $above = 2;
> my $below = 2;
>
> my @buffer;
>
> while ( <> ) {
> @buffer = ( @buffer, $_ )[ -( $above + 1 ) .. -1 ];
> /viewed/ && print @buffer, map scalar <>, 1 .. $below;
/viewed/ && print @buffer,
grep defined, map scalar <DATA>, 1 .. $below;
...in case of matches near the end of the file.
> }
Anno