How to store file content sentence by sentence in an array - perl

Question

I want to open a file, and store its content in an array and make changes to each sentence one at a time and then print the output of the file.
I have something like this:
open (FILE , $file);
my #lines = split('.' , <FILE>)
close FILE;
for (#lines) {
s/word/replace/g;
}
open (FILE, ">$file");
print FILE #lines;
close FILE;
For some reason, perl doesn't like this and won't output any content into the new file. It seems to not like me splitting up the array. Can someone give me an explanation why perl does this and a possible fix? Thanks!

Answer 1

split needs a regexp. Change split('.' , <FILE>) to split(/\./ , <FILE>)

Answer 2

Change my #lines = split('.' , <FILE>) to my #lines = split('\.' , <FILE>)
Only . is used in regex to match a single character. So you need to escape . to split on full stop.

Answer 3

#!/usr/local/bin/perl
use strict;
use warnings;
my $filename = "somefile.txt";
my $contents = do { local(#ARGV, $/) = $filename; <> };
my #lines = split '\.', $contents;
foreach(#lines){
#lines is an array which contains one sentence at each index.
}

Answer 4

what i found was in second line of your script missing semicolon(;) that is the error and also your script is not capable of handling content of entire file.It will process only one line. So please find the modification of your script below.If any clarification please let me know.
my $file='test.txt';#input file name
open (FILE , $file);
#my #lines = split('\.' ,<FILE>); this will not process the entire content of the file.
my #lines;
while(<FILE>) {
s/word/replace/g;
push(#lines,$_);
}
close FILE;
open (FILE, ">$file");
print FILE #lines;
close FILE;

Answer 5

You have lots of problems in your code.
my #lines = split('.' , <FILE>) will just read the first line and split it.
split('.' should be split(/\./
my #lines = split('.' , <FILE>) no semicolon terminator.
print FILE #lines; - you have lost all your full stops!
Finally I have to wonder why you are bothered about 'sentences' at all when you are just replacing one word. If you really want to read one sentence at a time (presumably to do some kind of sentence based processing) then you need to change the record separator variable $\. For example:
#!/usr/bin/perl
use strict;
use warnings;
my $file = "data.txt";
open (FILE , $file);
my #buffer;
$/ = '.'; # Change the Input Separator to read one sentence at a time.
# Be careful though, it won't work for questions ending in ?
while ( my $sentence = <FILE> ) {
$sentence =~ s/word/replace/g;
push #buffer, $sentence;
}
close FILE;
.. saving to file is left for you to solve.
However if you just want to change the strings you can read the whole file in one gulp by setting $/ to undef. Eg:
#!/usr/bin/perl
use strict;
use warnings;
my $file = "data.txt";
open (FILE , $file);
$/ = undef; # Slurp mode!
my $buffer = <FILE>;
close FILE;
$buffer =~ s/word/replace/g;
open (FILE, ">$file");
print FILE $buffer;
close FILE;
If you are really looking to process sentences and you want to get questions then you probably want to slurp the whole file and then split it, but use a capture in your regex so that you don't lose the punctuation. Eg:
!/usr/bin/perl
use strict;
use warnings;
my $file = "data.txt";
open (FILE , $file);
$/ = undef; # slurp!
my $buffer = <FILE>;
close FILE;
open (FILE, ">$file" . '.new'); # Don't want to overwrite my input.
foreach my $sentence (split(/([\.|\?]+)/, $buffer)) # split uses () to capture punctuation.
{
$sentence =~ s/word/replace/g;
print FILE $sentence;
}
close FILE;