| 1 | ###########################################################################
|
|---|
| 2 | #
|
|---|
| 3 | # PSPlug.pm -- this might look VERY similar to the PDF plugin...
|
|---|
| 4 | # A component of the Greenstone digital library software
|
|---|
| 5 | # from the New Zealand Digital Library Project at the
|
|---|
| 6 | # University of Waikato, New Zealand.
|
|---|
| 7 | #
|
|---|
| 8 | # Copyright (C) 1999 New Zealand Digital Library Project
|
|---|
| 9 | #
|
|---|
| 10 | # This program is free software; you can redistribute it and/or modify
|
|---|
| 11 | # it under the terms of the GNU General Public License as published by
|
|---|
| 12 | # the Free Software Foundation; either version 2 of the License, or
|
|---|
| 13 | # (at your option) any later version.
|
|---|
| 14 | #
|
|---|
| 15 | # This program is distributed in the hope that it will be useful,
|
|---|
| 16 | # but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|---|
| 17 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|---|
| 18 | # GNU General Public License for more details.
|
|---|
| 19 | #
|
|---|
| 20 | # You should have received a copy of the GNU General Public License
|
|---|
| 21 | # along with this program; if not, write to the Free Software
|
|---|
| 22 | # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
|---|
| 23 | #
|
|---|
| 24 | ###########################################################################
|
|---|
| 25 |
|
|---|
| 26 | # 12/05/02 Added usage datastructure - John Thompson
|
|---|
| 27 |
|
|---|
| 28 | package PSPlug;
|
|---|
| 29 |
|
|---|
| 30 | use ConvertToPlug;
|
|---|
| 31 | use sorttools;
|
|---|
| 32 |
|
|---|
| 33 | sub BEGIN {
|
|---|
| 34 | @ISA = ('ConvertToPlug');
|
|---|
| 35 | }
|
|---|
| 36 |
|
|---|
| 37 | my $arguments =
|
|---|
| 38 | [ { 'name' => "process_exp",
|
|---|
| 39 | 'desc' => "{BasPlug.process_exp}",
|
|---|
| 40 | 'type' => "string",
|
|---|
| 41 | 'deft' => &get_default_process_exp(),
|
|---|
| 42 | 'reqd' => "no" },
|
|---|
| 43 | { 'name' => "block_exp",
|
|---|
| 44 | 'desc' => "{BasPlug.block_exp}",
|
|---|
| 45 | 'type' => 'string',
|
|---|
| 46 | 'deft' => &get_default_block_exp() },
|
|---|
| 47 | { 'name' => "extract_date",
|
|---|
| 48 | 'desc' => "{PSPlug.extract_date}",
|
|---|
| 49 | 'type' => "flag" },
|
|---|
| 50 | { 'name' => "extract_pages",
|
|---|
| 51 | 'desc' => "{PSPlug.extract_pages}",
|
|---|
| 52 | 'type' => "flag" },
|
|---|
| 53 | { 'name' => "extract_title",
|
|---|
| 54 | 'desc' => "{PSPlug.extract_title}",
|
|---|
| 55 | 'type' => "flag" } ];
|
|---|
| 56 |
|
|---|
| 57 | my $options = { 'name' => "PSPlug",
|
|---|
| 58 | 'desc' => "{PSPlug.desc}",
|
|---|
| 59 | 'inherits' => "yes",
|
|---|
| 60 | 'args' => $arguments };
|
|---|
| 61 |
|
|---|
| 62 | sub new {
|
|---|
| 63 | my $class = shift (@_);
|
|---|
| 64 |
|
|---|
| 65 | # title_sub removes leading "Page 1" or "1 " from auto-extracted title.
|
|---|
| 66 |
|
|---|
| 67 | my $self = new ConvertToPlug ($class, "-convert_to", "text", @_ , "-title_sub", '^(Page\s+\d+)?(\s*1\s+)?');
|
|---|
| 68 | $self->{'plugin_type'} = "PSPlug";
|
|---|
| 69 | # 14-05-02 To allow for proper inheritance of arguments - John Thompson
|
|---|
| 70 | my $option_list = $self->{'option_list'};
|
|---|
| 71 | push( @{$option_list}, $options );
|
|---|
| 72 |
|
|---|
| 73 | if (!parsargv::parse(\@_,
|
|---|
| 74 | q^extract_date^, \$self->{'extract_date'},
|
|---|
| 75 | q^extract_pages^, \$self->{'extract_pages'},
|
|---|
| 76 | q^extract_title^, \$self->{'extract_title'},
|
|---|
| 77 | "allow_extra_options")) {
|
|---|
| 78 | print STDERR "\nIncorrect options passed to HTMLPlug, check your collect.cfg configuration file\n";
|
|---|
| 79 | $self->print_txt_usage(""); # Use default resource bundle
|
|---|
| 80 | die "\n";
|
|---|
| 81 | }
|
|---|
| 82 |
|
|---|
| 83 | return bless $self, $class;
|
|---|
| 84 | }
|
|---|
| 85 |
|
|---|
| 86 | # sub print_usage {
|
|---|
| 87 | # print STDERR "\n usage: plugin PSPlug [options]\n\n";
|
|---|
| 88 | # print STDERR " options:\n";
|
|---|
| 89 | # print STDERR " -extract_date Extract date from PS header\n";
|
|---|
| 90 | # print STDERR " -extract_pages Extract pages from PS header\n";
|
|---|
| 91 | # print STDERR " -extract_title Extract title from PS header\n";
|
|---|
| 92 | # print STDERR "\n\nNote! This is a \"poor man's\" ps to text converter. If you are serious, consider\n";
|
|---|
| 93 | # print STDERR "using the PRESCRIPT package, which is available for download at\n http://www.nzdl.org/html/software.html\n\n";
|
|---|
| 94 | # }
|
|---|
| 95 |
|
|---|
| 96 | sub get_default_block_exp {
|
|---|
| 97 | my $self = shift (@_);
|
|---|
| 98 |
|
|---|
| 99 | return q^(?i)\.(eps)$^;
|
|---|
| 100 | }
|
|---|
| 101 |
|
|---|
| 102 | sub get_default_process_exp {
|
|---|
| 103 | my $self = shift (@_);
|
|---|
| 104 |
|
|---|
| 105 | return q^(?i)\.ps$^;
|
|---|
| 106 | }
|
|---|
| 107 |
|
|---|
| 108 | sub extract_metadata_from_postscript {
|
|---|
| 109 | my $self = shift (@_);
|
|---|
| 110 | my $filename = shift (@_);
|
|---|
| 111 | my $doc = shift (@_);
|
|---|
| 112 | my $section = $doc->get_top_section();
|
|---|
| 113 |
|
|---|
| 114 | my $title_found = 0;
|
|---|
| 115 | my $pages_found = 0;
|
|---|
| 116 | my $date_found = 0;
|
|---|
| 117 |
|
|---|
| 118 | print STDERR "PSPlug: extracting PostScript metadata from \"$filename\"\n"
|
|---|
| 119 | if $self->{'verbosity'} > 1;
|
|---|
| 120 |
|
|---|
| 121 | open(INPUT, "<$filename");
|
|---|
| 122 | my $date;
|
|---|
| 123 |
|
|---|
| 124 | while(my $line =<INPUT>) {
|
|---|
| 125 | if ($self->{'extract_title'} && !$title_found) {
|
|---|
| 126 | foreach my $word ($line =~ m|Title: ([-A-Za-z0-9@/\/\(\):,. ]*)|g) {
|
|---|
| 127 | my $new_word = $word;
|
|---|
| 128 | $new_word =~ s/\(Untitled\)//i;
|
|---|
| 129 | $new_word =~ s/\(Microsoft Word\)//i;
|
|---|
| 130 | $new_word =~ s/Microsoft Word//i;
|
|---|
| 131 | $new_word =~ s/^\(//i;
|
|---|
| 132 | $new_word =~ s/\)$//i;
|
|---|
| 133 | $new_word =~ s/^ - //i;
|
|---|
| 134 | if ($new_word ne "") {
|
|---|
| 135 | $doc->add_utf8_metadata($section, "Title", $new_word );
|
|---|
| 136 | $title_found = 1;
|
|---|
| 137 | }
|
|---|
| 138 | }
|
|---|
| 139 | }
|
|---|
| 140 | if ($self->{'extract_date'} && !$date_found) {
|
|---|
| 141 | foreach my $word ($line =~ m/(Creation[-A-Za-z0-9@\/\(\):,. ]*)/g) {
|
|---|
| 142 | if ($word =~ m/ ([A-Za-z][A-Za-z][A-Za-z]) ([0-9 ][0-9]) ?[0-9: ]+ ([0-9]{4})/) {
|
|---|
| 143 | $date = &sorttools::format_date($2,$1,$3);
|
|---|
| 144 | if (defined $date) {
|
|---|
| 145 | $doc->add_utf8_metadata($section, "Date", $date );
|
|---|
| 146 | }
|
|---|
| 147 | }
|
|---|
| 148 | if ($word =~ m/D:([0-9]{4})([0-9]{2})([0-9]{2})[0-9]{6}\)/) {
|
|---|
| 149 | $date = &sorttools::format_date($3,$2,$1);
|
|---|
| 150 | if (defined $date) {
|
|---|
| 151 | $doc->add_utf8_metadata($section, "Date", $date );
|
|---|
| 152 | }
|
|---|
| 153 | }
|
|---|
| 154 | if ($word =~ m/CreationDate: ([0-9]{4}) ([A-Za-z][A-Za-z][A-Za-z]) ([0-9 ][0-9]) [0-9:]*/) {
|
|---|
| 155 | $date = &sorttools::format_date($3,$2,$1);
|
|---|
| 156 | if (defined $date) {
|
|---|
| 157 | $doc->add_utf8_metadata($section, "Date", $date );
|
|---|
| 158 | }
|
|---|
| 159 | }
|
|---|
| 160 | $date_found = 1;
|
|---|
| 161 | }
|
|---|
| 162 | }
|
|---|
| 163 | if ($self->{'extract_pages'} && !$pages_found) {
|
|---|
| 164 | foreach my $word ($line =~ m/(Pages: [0-9]*)/g) {
|
|---|
| 165 | my $digits = $word;
|
|---|
| 166 | $digits =~ s/[^0-9]//g;
|
|---|
| 167 | if ($digits ne "" && $digits ne "0") {
|
|---|
| 168 | $doc->add_utf8_metadata($section, "Pages", $digits );
|
|---|
| 169 | $pages_found = 1;
|
|---|
| 170 | }
|
|---|
| 171 | }
|
|---|
| 172 | }
|
|---|
| 173 | }
|
|---|
| 174 | }
|
|---|
| 175 |
|
|---|
| 176 | # do plugin specific processing of doc_obj for HTML type
|
|---|
| 177 | sub process {
|
|---|
| 178 | my $self = shift (@_);
|
|---|
| 179 | my ($trash, $trash2, $path, $file, $trash3, $doc) = @_;
|
|---|
| 180 |
|
|---|
| 181 | my $outhandle = $self->{'outhandle'};
|
|---|
| 182 | print $outhandle "PSPlug: passing $_[3] on to $self->{'converted_to'}Plug\n"
|
|---|
| 183 | if $self->{'verbosity'} > 1;
|
|---|
| 184 |
|
|---|
| 185 | &extract_metadata_from_postscript($self,"$path/$file", $doc);
|
|---|
| 186 | #$doc->add_utf8_metadata($doc->get_top_section(), "PSPlug", 1);
|
|---|
| 187 | return ConvertToPlug::process_type($self,"ps",@_);
|
|---|
| 188 | }
|
|---|
| 189 |
|
|---|
| 190 |
|
|---|
| 191 | 1;
|
|---|
| 192 |
|
|---|
