swish-e/prog-bin/doc2txt.pm

package doc2txt;
use strict;

=pod

=head1 NAME

doc2txt - swish-e sample module to convert MS Word docs to text

=head1 SYNOPSIS

    use doc2txt;
    my $doc_record_ref = doc2txt( $doc_file_name );

    # or by passing content in a scalar reference
    my $doc_text_ref = doc2txt( \$doc_content );


=head1 DESCRIPTION

Sample module for use with other swish-e 'prog' document source programs.

Pass either a file name, or a scalar reference.

The differece is when you pass a reference to a scalar
only the content is returned.  When you pass a file name
an entire record is returned ready to be fed to swish -- this
includes the headers required by swish for indexing.


=head1 REQUIREMENTS

Uses the catdoc program.  http://www.fe.msk.ru/~vitus/catdoc/

You may need to adjust the parameters used to call catdoc.

You will also need the module File::Temp available from CPAN if passing content
to this module (instead of a file name).  I'm not thrilled about how that
currently works...


=head1 AUTHOR

Bill Moseley

=cut

use Symbol;


use vars qw(
    @ISA
    @EXPORT
    $VERSION
);

# $Id: doc2txt.pm,v 1.2 2002/05/27 06:35:32 whmoseley Exp $
$VERSION = sprintf '%d.%02d', q$Revision: 1.2 $ =~ /: (\d+)\.(\d+)/;

require Exporter;
@ISA    = qw(Exporter);
@EXPORT = qw(doc2txt);

my @InfoTags = qw/Title Subject Author CreationDate Creator Producer ModDate Keywords/;

my $catdoc = 'catdoc -a';  # how cat doc is called. Rainer uses catdoc -s8859-1 -d8859-1


sub doc2txt {
    my $file_or_content = shift;


    my $file = ref $file_or_content
    ? create_temp_file( $file_or_content )
    : $file_or_content;


    my $content = `$catdoc $file`;

    return \$content if ref $file_or_content;

    # otherwise build the headers

    my $mtime  = (stat $file )[9];

    my $size = length $content;

    my $ret = <<EOF;
Content-Length: $size
Last-Mtime: $mtime
Path-Name: $file

EOF

    $ret .= $content;

    return \$ret;
    

}


# This is the portable way to do this, I suppose.
# Otherwise, just create a file in the local directory.

sub create_temp_file {
    my $scalar_ref = shift;

    require "File/Temp.pm";

    my ( $fh, $file_name ) = File::Temp::tempfile( UNLINK => 1 );

    print $fh $$scalar_ref or die $!;


    close $fh or die "Failed to close '$file_name' $!";

    return $file_name;
}
    

1	package doc2txt;
2	use strict;
3
4	=pod
5
6	=head1 NAME
7
8	doc2txt - swish-e sample module to convert MS Word docs to text
9
10	=head1 SYNOPSIS
11
12	use doc2txt;
13	my $doc_record_ref = doc2txt( $doc_file_name );
14
15	# or by passing content in a scalar reference
16	my $doc_text_ref = doc2txt( \$doc_content );
17
18
19
20
21	=head1 DESCRIPTION
22
23	Sample module for use with other swish-e 'prog' document source programs.
24
25	Pass either a file name, or a scalar reference.
26
27	The differece is when you pass a reference to a scalar
28	only the content is returned. When you pass a file name
29	an entire record is returned ready to be fed to swish -- this
30	includes the headers required by swish for indexing.
31
32
33	=head1 REQUIREMENTS
34
35	Uses the catdoc program. http://www.fe.msk.ru/~vitus/catdoc/
36
37	You may need to adjust the parameters used to call catdoc.
38
39	You will also need the module File::Temp available from CPAN if passing content
40	to this module (instead of a file name). I'm not thrilled about how that
41	currently works...
42
43
44	=head1 AUTHOR
45
46	Bill Moseley
47
48	=cut
49
50	use Symbol;
51
52
53	use vars qw(
54	@ISA
55	@EXPORT
56	$VERSION
57	);
58
59	# $Id: doc2txt.pm,v 1.2 2002/05/27 06:35:32 whmoseley Exp $
60	$VERSION = sprintf '%d.%02d', q$Revision: 1.2 $ =~ /: (\d+)\.(\d+)/;
61
62	require Exporter;
63	@ISA = qw(Exporter);
64	@EXPORT = qw(doc2txt);
65
66	my @InfoTags = qw/Title Subject Author CreationDate Creator Producer ModDate Keywords/;
67
68	my $catdoc = 'catdoc -a'; # how cat doc is called. Rainer uses catdoc -s8859-1 -d8859-1
69
70
71	sub doc2txt {
72	my $file_or_content = shift;
73
74
75	my $file = ref $file_or_content
76	? create_temp_file( $file_or_content )
77	: $file_or_content;
78
79
80	my $content = `$catdoc $file`;
81
82	return \$content if ref $file_or_content;
83
84	# otherwise build the headers
85
86	my $mtime = (stat $file )[9];
87
88	my $size = length $content;
89
90	my $ret = <<EOF;
91	Content-Length: $size
92	Last-Mtime: $mtime
93	Path-Name: $file
94
95	EOF
96
97	$ret .= $content;
98
99	return \$ret;
100
101
102	}
103
104
105	# This is the portable way to do this, I suppose.
106	# Otherwise, just create a file in the local directory.
107
108	sub create_temp_file {
109	my $scalar_ref = shift;
110
111	require "File/Temp.pm";
112
113	my ( $fh, $file_name ) = File::Temp::tempfile( UNLINK => 1 );
114
115	print $fh $$scalar_ref or die $!;
116
117
118	close $fh or die "Failed to close '$file_name' $!";
119
120	return $file_name;
121	}
122
123