source: trunk/gsdl/perllib/plugins/GBHPlug.pm@ 1020

###########################################################################
#
# GBHPlug.pm -- plugin for processing gb encoded html
# A component of the Greenstone digital library software
# from the New Zealand Digital Library Project at the 
# University of Waikato, New Zealand.
#
# Copyright (C) 1999 New Zealand Digital Library Project
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program; if not, write to the Free Software
# Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
#
###########################################################################

# creates simple single-level document from .htm or .html files 
# (case-insensitive match on filenames). Adds Title metadata 
# taken from <title> tags if found otherwise first 100 characters
# outside of tags.
# will also attempt to include images which it will search for in
# the same directory as the document itself (it will also search 
# directories relative to that directory).

# this plugin currently does nothing with href links so relative links 
# may become broken.


package GBHPlug;

use BasPlug;
use sorttools;
use util;
use unicode;
use cnseg;
use gb;


sub BEGIN {
    @ISA = ('BasPlug');
}

sub new {
    my ($class) = @_;
    $self = new BasPlug ();

    return bless $self, $class;
}

sub is_recursive {
    my $self = shift (@_);

    return 0; # this is not a recursive plugin
}


# return number of files processed, undef if can't process
# Note that $base_dir might be "" and that $file might 
# include directories
sub read {
    my $self = shift (@_);
    my ($pluginfo, $base_dir, $file, $metadata, $processor) = @_;

    my $filename = &util::filename_cat($base_dir, $file);
    my $absdir = $filename;
    $absdir =~ s/[^\/\\]*$//;

    return undef unless ($filename =~ /\.(html?(\.gz)?)$/i && (-e $filename));

    my $gz = 0;
    if (defined $2) {
    $gz = $2;
    $gz = 1 if ($gz =~ /\.gz/i);
    }

    print STDERR "GBHPlug: processing $filename\n" if $processor->{'verbosity'};

    # create a new document
    my $doc_obj = new doc ($file, "indexed_doc");

    if ($gz) {
    open (FILE, "zcat $filename |") || die "GBHPlug::read - ERROR: zcat can't open $filename\n";
    } else {
    open (FILE, $filename) || die "GBHPlug::read - ERROR: can't open $filename\n";
    }
    my $cursection = $doc_obj->get_top_section();

    my $text = "";
    my $line = "";
    my $title = "";
    while (defined ($line = <FILE>)) {
    $text .= $line;
    }

    # convert to unicode
    $text = &unicode::unicode2utf8(&gb::gb2unicode($text));
    # segment the Chinese words
    $text = &cnseg::segment($text);

    # we'll use the worthless alarm thingy to temporarily replace 
    # '\n' so we'd better check it doesn't occur naturally
    if ($text =~ /\a/) {
    print STDERR "GBHPlug::read - 'WARNING '\a' character occurs in text!!\n";
    }

    # remove line breaks
    $text =~ s/\n/\a/g;

    # see if there's a <title> tag
    if ($text =~ /<title[^>]*>([^<]*)<\/title[^>]*>/i) {
    if (defined $1) {
        $title = $1;
    }
    }
    # if no title use first 20 characters
    if ($title eq "") {
    my $tmptext = $text;
    $tmptext =~ s/<[^>]*>//g;
    $title = substr ($tmptext, 0, 20);
    }

    if ($title ne "") {
    $doc_obj->add_utf8_metadata ($cursection, "Title", $title);
    }

    # remove header rubbish
    $text =~ s/^.*?<body[^>]*>//i;

    # and any other unwanted tags
    $text =~ s/<(\/p|\/html|\/body)>//g;

    # fix up the image links
    $text =~ s/(<img[^>]*?src\s*=\s*\"?)([^\">]+)(\"?[^>]*>)/
    &replace_image_links($absdir, $doc_obj, $1, $2, $3)/ige;

    # put line breaks back in
    $text =~ s/\a/\n/g;

    # add the text
    $doc_obj->add_utf8_text($cursection, $text);

    # assume that any metadata passed to this plugin is already utf8
    foreach $field (keys(%$metadata)) {
    # $metadata->{$field} may be an array reference
    if (ref ($metadata->{$field}) eq "ARRAY") {
        map { 
        $doc_obj->add_utf8_metadata ($cursection, $field, $_); 
        } @{$metadata->{$field}};
    } else {
        $doc_obj->add_utf8_metadata ($cursection, $field, $metadata->{$field}); 
    }
    }

    # add OID
    $doc_obj->set_OID ();

    # process the document
    $processor->process($doc_obj);

    return 1; # processed the file
}

sub replace_image_links {

    my ($dir, $doc_obj, $front, $link, $back) = @_;

    my ($filename, $error);
    my $foundimage = 0;
    
    $link =~ s/\/\///;
    my ($imagetype) = $link =~ /([^\.]*)$/;
    $imagetype =~ tr/[A-Z]/[a-z]/;
    if ($imagetype eq "jpg") {$imagetype = "jpeg";}
    if ($imagetype !~ /^(jpg|gif|png)$/) {
    print STDERR "GBHPlug: Warning - unknown image type ($imagetype)\n";
    }
    my ($imagefile) = $link =~ /([^\/]*)$/;
    my ($imagepath) = $link =~ /^[^\/]*(.*)$/;

    if (defined $imagepath && $imagepath =~ /\w/) {
    # relative link
    $filename = &util::filename_cat ($dir, $imagepath);
    if (-e $filename) {
        $doc_obj->associate_file ($filename, $imagefile, "image/$imagetype");
        $foundimage = 1;
    } else {
        $error = "GBHPlug: Warning - couldn't find image file $imagefile in either $filename or";
    }
    }

    if (!$foundimage) {
    $filename = &util::filename_cat ($dir, $imagefile);
    if (-e $filename) {
        $doc_obj->associate_file ($filename, $imagefile, "image/$imagetype");    
        $foundimage = 1;
    } elsif (defined $error) {
        print STDERR "$error $filename\n";
    } else {
        print STDERR "GBHPlug: Warning - couldn't find image file $imagefile in $filename\n";
    }
    }

    if ($foundimage) {
    return "${front}_httpdocimg_/${imagefile}${back}";
    } else {
    return "";
    }
}

1;