DUC

Extract

This class opens extracts, both from files and from strings, 
so that the user will be able to get extract information quickly
and easily.

MEAD::Extract should do much the same stuff as this class.

Methods description

Methods code

    my ($self, $extract_string) = @_;

    $self->{extract_string} = $extract_string;

    my $extract_tree = XML::TreeBuilder->new;
    $extract_tree->parse($extract_string);
    my $extract_node = $extract_tree->find_by_tag_name("multi-e");

    my $node;
    my $i = 0;
    my @nodes = $extract_node->look_down("_tag", "s");
    foreach my $node (@nodes) {
	$i++;

	## document ID's can be called either of these names...
        my $DID = $node->attr("docref");
	unless ($DID) {
	    $DID = $node->attr("docid");
	}

        my $SNO = $node->attr("num");


        my $WCNT = $node->attr("wdcount");

	## TODO: AJW 8/28
	## get the wordcount somehow, possibly splitting the words...

        ##die unless ($DID and $SNO and $WCNT);
	unless ($DID && $SNO) {
	    die "Couldn't find (at least one of) 'docid' or 'num' " .
		"in the $i-th sentence node";
	}

	$self->{DID_list}[$i] = $DID;
	$self->{SNO_list}[$i] = $SNO;
	$self->{WCNT_list}[$i] = $WCNT;
    }

    ##cleanup.
    $extract_tree->delete;

    my ($self, $sentence) = @_;
    return $self->get_DOCID_for_sentence($sentence);

    my ($self, $sentence) = @_;
    return $self->{DID_list}[$sentence];

    my ($self, $sentence) = @_;
    return $self->{SNO_list}[$sentence];

    my ($self, $sentence) = @_;
    return $self->{WCNT_list}[$sentence];

    my $self = shift;
    return scalar(@{$self->{DID_list}}) - 1;

    my $self = shift;
    unless ($self->{text}) {
    $self->{text} = $self->{extract_string};
    $self->{text} =~ s/\<.+?\>//g;
    $self->{text} =~ s/\n+/  /g;
    }
    return $self->{text};

    my ($class, $filename) = @_;
    my $self = {};

    bless $self, $class;

    open EXTRACT, $filename;
    
    my $extract_string;
    while (<EXTRACT>) {
	$extract_string .= $_;
    }
    close EXTRACT;

    $self->_really_open_me($extract_string);
    
    return $self;

    my ($class, $extract_string) = @_;
    my $self = {};
    
    bless $self, $class;

    $self->_really_open_me($extract_string);

    return $self;

General documentation