CODEWITHSUNDEEP
perl
Gsuz
Gsuz

Reputation: 353

perl find text from file in certain position

130723,-001,1.14,130725,+002,4.20,130731,+006,1.52,130728
130725,+002,4.20,130731,+006,1.52,130728,-003,0.00,130731
130731,+006,1.52,130728,-003,0.00,130731,+003,1.00,130731
130728,-003,0.00,130731,+003,1.00,130731,+000,0.00,130729
130731,+000,0.00,130729,-002,1.00,130728,-001,0.00,130728

the above is part of a log file. Each line in the log file is always the same length and has the same pattern as you can see above. I need to read the file and place in an array all the lines where position 42 to 46 in each line meet certain expectations. In the case above we are looking at the following numbers:

+006 -003 +003 +000 -001

Can someone point me in the right direction?

EDIT :

Thx to Amon for his suggestion.

I ended up with this code for future reference.

open (FILE, $filename) or die "Couldn't open log: $!";
while (<FILE>) { 
if ((split /,/)[8] == "+003"){
push @data, $_ }}

close FILE;

foreach(@data)
{
    print "$_\r\n";
}

I was thinking towards the future if this file gets really big what steps should I take to optimise the process speedwise?

Upvotes: 0

Views: 422

Answers (3)

TrueY
TrueY

Reputation: 7610

Try

perl -F, -ane '$F[7] eq "+003" and push @l,$_; END { print for @l }'<<XXX
130723,-001,1.14,130725,+002,4.20,130731,+006,1.52,130728
130725,+002,4.20,130731,+006,1.52,130728,-003,0.00,130731
130731,+006,1.52,130728,-003,0.00,130731,+003,1.00,130731
130728,-003,0.00,130731,+003,1.00,130731,+000,0.00,130729
130731,+000,0.00,130729,-002,1.00,130728,-001,0.00,130728
XXX

Output:

130731,+006,1.52,130728,-003,0.00,130731,+003,1.00,130731

Upvotes: 0

mattexx
mattexx

Reputation: 6606

While @amon's answer is elegant, you can just use regex:

open FILE, "filename.txt" or die $!;
while (<FILE>) {
    if $_ =~ /^.{41}(\+006)|(-003)|(\+003)|(\+000)|(-001)/
}

Upvotes: 0

Jonathan Leffler
Jonathan Leffler

Reputation: 754590

If you want to do it by column numbers, then substr() is usable with care:

perl -pe '$_ = substr($_, 41, 4) . "\n"' data

Your question asks for columns 42..46, but with an inclusive notation, that selects 5 positions, the last of which is a comma. Specifying 42..46 is perhaps the 1-based half-open range of columnns.

The 41 in the code is 'column 42 - 1' (0-based indexes); the 4 is '46 - 42'. So, for columns [N..M), the formula would be:

perl -pe '$_ = substr($_, N-1, M-N) . "\n"' data

Upvotes: 1

Related Questions