Retreaving N first sentences from text.

c0d34w4y has asked for the wisdom of the Perl Monks concerning the following question:

Fella Monks...

The code below is a demo of my attempt at
writing a little function to retreave N first
sentences from a string.  I wonder if there
would be any suggestion on your part as to 
other implementations to achieve similar results?

Cheers,
c0d34w4y

my $s = "foo. bar. foobar. foo bar.";
my $p = get_nsentences($s,2);  # $p = "foo. bar.";

print $p;
print "done\n";

## SUBS

# get n first sentences from a given string.
sub get_nsentences {
    my $source_txt = $_[0];
    my $ret_txt = "";

    # i know this looks twisted...
    for($i=$_[1]||1;$i>0 && length($source_txt);$i--) { $source_txt =~
+ s/([^\.]+\.)/(($ret_txt.=$1)&&"")/e; }
    return $ret_txt;
}
[download]

print join(" ", map { sprintf "%#02x", $_ }unpack("C*",pack("L",0x1234
+5678)))
[download]

Comment on Retreaving N first sentences from text. Select or Download Code

Replies are listed 'Best First'.
Re: Retreaving N first sentences from text. by Beatnik (Parson) on Dec 15, 2001 at 03:48 UTC
Maybe Lingua::EN::Sentence or Text::Sentence are modules you can use? Text::Sentences `split_sentence` method : `@sentences = split_sentences( $text );` Greetz Beatnik ... Quidquid perl dictum sit, altum viditur.	[reply] [d/l] [select]
Re: Retreaving N first sentences from text. by Zaxo (Archbishop) on Dec 15, 2001 at 04:05 UTC
Lingua::EN::Sentence has lots of re's: `use Lingua::EN::Sentence qw( get_sentences ); sub nsentences { my ($num, @text) = @_; my $sentences = get_sentences(join " ",@text); join " ", @{$sentences}[0..$num-1]; }` [download] After Compline, Zaxo	[reply] [d/l]
Re: Re: Retreaving N first sentences from text. by c0d34w4y (Acolyte) on Dec 15, 2001 at 04:20 UTC
I knew my code was far from being perfect ;-) Thankx for great suggestions... I'm going to change my code a bit now. I think the Lingua::EN::Sentence module might be of some use. -- `print join(" ", map { sprintf "%#02x", $_ }unpack("C*",pack("L",0x1234 +5678)))` [download]	[reply] [d/l]
Re: Retreaving N first sentences from text. by dws (Chancellor) on Dec 15, 2001 at 03:58 UTC
There was a good article in TPJ a few issues back that included a discussion of the inner workings of Text::Sentence, and how it recognizes sentences.	[reply]
Re: Retreaving N first sentences from text. by dragonchild (Archbishop) on Dec 15, 2001 at 03:49 UTC
Uhhh.. if you know that all sentences are separated by the '.', then why not do something like: `sub get_nsentences { my ($source_text, $n) = @_; my @sentences = split /\.\s/, $source_text; $ret_text = (join '.', @sentences[0..$n-1]) . '.'; return $ret_text; }` [download] ------ We are the carpenters and bricklayers of the Information Age.* Don't go borrowing trouble. For programmers, this means Worry only about what you need to implement.	[reply] [d/l]
Re: Re: Retrieving N first sentences from text. by boo_radley (Parson) on Dec 15, 2001 at 03:55 UTC
broken : `She said "I believe this will be the winter of my discontent. Then again, I may drink some egg-nog. Who knows?".` How many sentences there?	[reply]
Re: Re: Retreaving N first sentences from text. by blakem (Monsignor) on Dec 15, 2001 at 03:59 UTC
You could use the /g regex modifier to overcome the 'only periods' assumption.... `sub get_nsentences { my $text = shift; my $count = shift \|\| 1; join '', ($text =~ /.*?[\.!?]/sg)[0..$count-1] }` [download] This keeps the punctuation with the "sentence" so you don't have to keep track of what it was.... i.e. you can join with '', not with '.' -Blake	[reply] [d/l]