source: trunk/gsdl/perllib/mgppbuildproc.pm@ 10113

###########################################################################
#
# mgppbuildproc.pm -- 
# 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.
#
###########################################################################

# This document processor outputs a document
# for mgpp to process


package mgppbuildproc;

use basebuildproc;


BEGIN {
    @mgppbuildproc::ISA = ('basebuildproc');
}

#this must be the same as in mgppbuilder
our %level_map = ('document'=>'Doc',
          'section'=>'Sec',
          'paragraph'=>'Para');

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

    # use a different index specification to the default
    $self->{'index'} = "text";

    $self->{'dontindex'} = {};
    $self->{'indexfieldmap'} = {};
    $self->{'indexfields'} = {}; # only put in the ones that are not specified directly in the index
    $self->{'strip_html'}=1;
    
    return bless $self, $class;
}


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

    $self->{'indexfieldmap'} = $indexmap;
}

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

    return $self->{'indexfieldmap'};
}

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

    $self->{'levels'} = $levels;
}

sub set_strip_html {
    my $self = shift (@_);
    my ($strip) = @_;
    $self->{'strip_html'}=$strip;
}


sub get_gdbm_level {
    my $self = shift (@_);
    
    #if a Section level index is not built, the gdbm file should be at doc 
    #level not Section
    if ($self->{'levels'}->{'section'}) {
    return "section";
    }
    return "document";
}


#sub find_paragraphs {
#    $_[1] =~ s/(<p\b)/<Paragraph>$1/gi;
#}

#this function strips the html tags from the doc if ($strip_html) and
# if ($para) replaces <p> with <Paragraph> tags.
# if both are false, the original text is returned
#assumes that <pre> and </pre> have no spaces, and removes all < and > inside
#these tags
sub preprocess_text {
    my $self = shift (@_);
    my ($text, $strip_html, $para) = @_;
    my ($outtext) = "";
    if ($strip_html) { 
    while ($text =~ /<([^>]*)>/ && $text ne "") {
        
        my $tag = $1;
        $outtext .= $`." "; #add everything before the matched tag
        $text = $'; #'everything after the matched tag
        if ($para && $tag =~ /^\s*p\s/i) {
        $outtext .= $para;
        }
        elsif ($tag =~ /^pre$/) { # a pre tag
        $text =~ /<\/pre>/; # find the closing pre tag
        my $tmp_text = $`; #everything before the closing pre tag
        $text = $'; #'everything after the </pre>
        $tmp_text =~ s/[<>]//g; # remove all < and >
        $outtext.= $tmp_text . " ";
        }
    }
    
    $outtext .= $text; # add any remaining text
    return $outtext;
    } #if strip_html

    #if ($para) {
    #$text =~ s/(<p\b)/$para$1/gi;
    #return $text;
   # }
    return $text;
}
    
    

sub filter_text {
    # $self->filter_text ($field, $new_text);
    # don't want to do anything for this version, however,
    # in a particular collection you might want to override
    # this method to post-process certain fields depending on
    # the field, or whether we are outputting it for indexing
}

sub text {
    my $self = shift (@_);
    my ($doc_obj) = @_;
    my $handle = $self->{'output_handle'};
    my $outhandle = $self->{'outhandle'};

    # only output this document if it is one to be indexed
    return if ($doc_obj->get_doc_type() ne "indexed_doc");

    my $indexed_doc = $self->is_subcollection_doc($doc_obj);
    
    # this is another document
    $self->{'num_docs'} += 1;

    # get the parameters for the output
    # split on : just in case there is subcoll and lang stuff
    my ($fields) = split (/:/, $self->{'index'});

    my ($documenttag) = "";
    my($documentendtag) = "";
    if ($self->{'levels'}->{'document'}) {
    $documenttag = "\n<". $level_map{'document'} . ">\n";
    $documentendtag = "\n</". $level_map{'document'} . ">\n";
    }
    my ($sectiontag) = "";
    if ($self->{'levels'}->{'section'}) {
    $sectiontag = "\n<". $level_map{'section'} . ">\n";
    }
    my ($paratag) = "";
    
    if ($self->{'levels'}->{'paragraph'}) {
    if ($self->{'strip_html'}) {
        $paratag = "<". $level_map{'paragraph'} . ">";
    } else {
        print $outhandle "Paragraph level can not be used with no_strip_html!. Not indexing Paragraphs.\n";
    }
    }

    my $doc_section = 0; # just for this document
    
    my $text = $documenttag;
   
    # get the text for this document
    my $section = $doc_obj->get_top_section();
    while (defined $section) {
    # update a few statistics
    $doc_section++;
    $self->{'num_sections'} += 1;
    $text .= "$sectiontag";

    if ($indexed_doc) {
        if ($self->{'indexing_text'}) {
        $text .= "$paratag"; # only add para tags for indexing
        # note that we assume that metadata will not be asked for for the compressed text, so we add para tags without checking for indexing_text
        }
        $self->{'num_bytes'} += $doc_obj->get_text_length ($section);
        foreach my $field (split (/,/, $fields)) {
        # only deal with this field if it doesn't start with top or
        # this is the first section
        my $real_field = $field;
        if (!($real_field =~ s/^top//) || ($doc_section == 1)) {
            my $new_text = ""; 
            my $tmp_text = "";
            if ($real_field eq "text") {
            if ($self->{'indexing_text'}) { #tag the text with <Text>...</Text>, add the <Paragraph> tags and strip out html if needed
                $new_text .= "$paratag<TX>\n";
                $tmp_text .= $doc_obj->get_text ($section);
                $tmp_text = $self->preprocess_text($tmp_text, $self->{'strip_html'}, "</TX>$paratag<TX>");
                
                $new_text .= "$tmp_text</TX>\n";
                #if (!defined $self->{'indexfields'}->{'TextOnly'}) {
                #$self->{'indexfields'}->{'TextOnly'} = 1;   
                #}
            }
            else { # leave html stuff in, and dont add Paragraph tags - never retrieve paras at the moment
                $new_text .= $doc_obj->get_text ($section) if $self->{'store_text'};
            }
            } else { # metadata field
            if ($real_field eq "allfields") { #ignore
            }
            elsif ($real_field eq "metadata") { # insert all metadata 
                #except gsdl stuff
                my $shortname = "";
                my $metadata = $doc_obj->get_all_metadata ($section);
                foreach my $pair (@$metadata) {
                my ($mfield, $mvalue) = (@$pair);
                # check fields here, maybe others dont want - change to use dontindex!!
                if ($mfield ne "Identifier" 
                    && $mfield !~ /^gsdl/ 
                    && $mfield ne "classifytype"
                    && $mfield ne "assocfilepath"
                    && defined $mvalue && $mvalue ne "") {
                    
                    if (defined $self->{'indexfieldmap'}->{$mfield}) {
                    $shortname = $self->{'indexfieldmap'}->{$mfield};
                    }
                    else {
                    $shortname = $self->create_shortname($mfield);
                    $self->{'indexfieldmap'}->{$mfield} = $shortname;
                    $self->{'indexfieldmap'}->{$shortname} = 1;
                    }      
                    $new_text .= "$paratag<$shortname>$mvalue</$shortname>\n";
                    if (!defined $self->{'indexfields'}->{$mfield}) {
                    $self->{'indexfields'}->{$mfield} = 1;
                    }                   
                }
                }
                
            }
            else { #individual metadata specified
                my $shortname="";
                #if (!defined $self->{'indexfields'}->{$real_field}) {
                #$self->{'indexfields'}->{$real_field} = 1;
                #}
                if (defined $self->{'indexfieldmap'}->{$real_field}) {
                $shortname = $self->{'indexfieldmap'}->{$real_field};
                }
                else {
                $shortname = $self->create_shortname($real_field);
                $self->{'indexfieldmap'}->{$real_field} = $shortname;
                $self->{'indexfieldmap'}->{$shortname} = 1;
                }
                foreach my $item (@{$doc_obj->get_metadata ($section, $real_field)}) {
                $new_text .= "$paratag<$shortname>$item</$shortname>\n";
                }
            }

            }
            
            # filter the text
            $self->filter_text ($field, $new_text);

            $self->{'num_processed_bytes'} += length ($new_text);
            $text .= "$new_text";
        }
        }
    } # if (indexed_doc)
    
    $section = $doc_obj->get_next_section($section);
    } #while defined section
    print $handle "$text\n$documentendtag"; 
    
}

#chooses the first two letters or digits for the shortname
#now ignores non-letdig characters
sub create_shortname {
    my $self = shift(@_);
    
    my ($realname) = @_;
    #take the first two chars
    my $shortname;
    if ($realname =~ /^[^\w]*(\w)[^\w]*(\w)/) {
    $shortname = "$1$2";
    } else {
    # there aren't two letdig's in the field - try arbitrary combinations
    $realname = "ABCDEFGHIJKLMNOPQRSTUVWXYZ";
    $shortname = "AB";
    }
    $shortname =~ tr/a-z/A-Z/;

    #if already used, take the first and third letdigs and so on
    my $count = 1;
    while (defined $self->{'indexfieldmap'}->{$shortname}) {
    if ($realname =~ /^[^\w]*(\w)([^\w]*\w){$count}[^\w]*(\w)/) {
        $shortname = "$1$3";
        $count++;
        $shortname =~ tr/a-z/A-Z/;
    
    }
    else {
        #remove up to and incl the first letdig
        $realname =~ s/^[^\w]*\w//;
        $count = 0;
    }
    }

    return $shortname;
}

1;