| 1 | ###########################################################################
|
|---|
| 2 | #
|
|---|
| 3 | # HBPlug.pm --
|
|---|
| 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 | # plugin which processes an HTML book directory
|
|---|
| 27 |
|
|---|
| 28 | # This plugin is used by the Humanity Library collections and does not handle
|
|---|
| 29 | # input encodings other than ascii or extended ascii
|
|---|
| 30 |
|
|---|
| 31 | # this code is kind of ugly and could no doubt be made to run faster, by leaving
|
|---|
| 32 | # it in this state I hope to encourage people to make their collections use
|
|---|
| 33 | # HBSPlug instead ;-)
|
|---|
| 34 |
|
|---|
| 35 | # Use HBSPlug if creating a new collection and marking up files like the
|
|---|
| 36 | # Humanity Library collections. HBSPlug accepts all input encodings but
|
|---|
| 37 | # expects the marked up files to be cleaner than those used by the
|
|---|
| 38 | # Humanity Library collections
|
|---|
| 39 |
|
|---|
| 40 | package HBPlug;
|
|---|
| 41 |
|
|---|
| 42 | use ghtml;
|
|---|
| 43 | use BasPlug;
|
|---|
| 44 | use unicode;
|
|---|
| 45 | use util;
|
|---|
| 46 | use doc;
|
|---|
| 47 |
|
|---|
| 48 | use strict;
|
|---|
| 49 | no strict 'refs'; # allow filehandles to be variables and viceversa
|
|---|
| 50 |
|
|---|
| 51 | sub BEGIN {
|
|---|
| 52 | @HBPlug::ISA = ('BasPlug');
|
|---|
| 53 | }
|
|---|
| 54 |
|
|---|
| 55 | my $arguments =
|
|---|
| 56 | [ { 'name' => "process_exp",
|
|---|
| 57 | 'desc' => "{BasPlug.process_exp}",
|
|---|
| 58 | 'type' => "regexp",
|
|---|
| 59 | 'reqd' => "no",
|
|---|
| 60 | 'deft' => &get_default_process_exp() }
|
|---|
| 61 | ];
|
|---|
| 62 |
|
|---|
| 63 | my $options = { 'name' => "HBPlug",
|
|---|
| 64 | 'desc' => "{HBPlug.desc}",
|
|---|
| 65 | 'abstract' => "no",
|
|---|
| 66 | 'inherits' => "yes",
|
|---|
| 67 | 'args' => $arguments };
|
|---|
| 68 |
|
|---|
| 69 | sub new {
|
|---|
| 70 | my ($class) = shift (@_);
|
|---|
| 71 | my ($pluginlist,$inputargs,$hashArgOptLists) = @_;
|
|---|
| 72 | push(@$pluginlist, $class);
|
|---|
| 73 |
|
|---|
| 74 | if(defined $arguments){ push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});}
|
|---|
| 75 | if(defined $options) { push(@{$hashArgOptLists->{"OptList"}},$options)};
|
|---|
| 76 |
|
|---|
| 77 | my $self = (defined $hashArgOptLists)? new BasPlug($pluginlist,$inputargs,$hashArgOptLists): new BasPlug($pluginlist,$inputargs);
|
|---|
| 78 |
|
|---|
| 79 | return bless $self, $class;
|
|---|
| 80 | }
|
|---|
| 81 |
|
|---|
| 82 | sub init {
|
|---|
| 83 | my $self = shift (@_);
|
|---|
| 84 | my ($verbosity, $outhandle) = @_;
|
|---|
| 85 |
|
|---|
| 86 | $self->BasPlug::init($verbosity, $outhandle);
|
|---|
| 87 | $self->{'input_encoding'} = "iso_8859_1";
|
|---|
| 88 |
|
|---|
| 89 | # this plugin only handles ascii encodings
|
|---|
| 90 | if ($self->{'input_encoding'} !~ /^(iso_8859_1|ascii)$/) {
|
|---|
| 91 | die "ERROR: HBPlug can handle only iso_8859_1 or ascii encodings.\n" .
|
|---|
| 92 | $self->{'input_encoding'} . " is not an acceptable input_encoding value\n";
|
|---|
| 93 | }
|
|---|
| 94 | }
|
|---|
| 95 |
|
|---|
| 96 | # this is included only to prevent warnings being printed out
|
|---|
| 97 | # from BasPlug::init. The process_exp is not used by this plugin
|
|---|
| 98 | sub get_default_process_exp {
|
|---|
| 99 | my $self = shift (@_);
|
|---|
| 100 |
|
|---|
| 101 | return "This plugin does not use a process_exp\n";
|
|---|
| 102 | }
|
|---|
| 103 |
|
|---|
| 104 |
|
|---|
| 105 | sub HB_read_html_file {
|
|---|
| 106 | my $self = shift (@_);
|
|---|
| 107 | my ($htmlfile, $text) = @_;
|
|---|
| 108 |
|
|---|
| 109 | # load in the file
|
|---|
| 110 | if (!open (FILE, $htmlfile)) {
|
|---|
| 111 | my $outhandle = $self->{'outhandle'};
|
|---|
| 112 | print $outhandle "ERROR - could not open $htmlfile\n";
|
|---|
| 113 | return;
|
|---|
| 114 | }
|
|---|
| 115 |
|
|---|
| 116 | my $foundbody = 0;
|
|---|
| 117 | $self->HB_gettext (\$foundbody, $text, "FILE");
|
|---|
| 118 | close FILE;
|
|---|
| 119 |
|
|---|
| 120 | # just in case there was no <body> tag
|
|---|
| 121 | if (!$foundbody) {
|
|---|
| 122 | $foundbody = 1;
|
|---|
| 123 | open (FILE, $htmlfile) || return;
|
|---|
| 124 | $self->HB_gettext (\$foundbody, $text, "FILE");
|
|---|
| 125 | close FILE;
|
|---|
| 126 | }
|
|---|
| 127 | # text is in utf8
|
|---|
| 128 | }
|
|---|
| 129 |
|
|---|
| 130 | # converts the text to utf8, as ghtml does that for é etc.
|
|---|
| 131 | sub HB_gettext {
|
|---|
| 132 | my $self = shift (@_);
|
|---|
| 133 | my ($foundbody, $text, $handle) = @_;
|
|---|
| 134 | my $outhandle = $self->{'outhandle'};
|
|---|
| 135 |
|
|---|
| 136 | my $line = "";
|
|---|
| 137 | while (defined ($line = <$handle>)) {
|
|---|
| 138 | # look for body tag
|
|---|
| 139 | if (!$$foundbody) {
|
|---|
| 140 | if ($line =~ s/^.*<body[^>]*>//i) {
|
|---|
| 141 | $$foundbody = 1;
|
|---|
| 142 | } else {
|
|---|
| 143 | next;
|
|---|
| 144 | }
|
|---|
| 145 | }
|
|---|
| 146 |
|
|---|
| 147 | # check for symbol fonts
|
|---|
| 148 | if ($line =~ /<font [^>]*?face\s*=\s*\"?(\w+)\"?/i) {
|
|---|
| 149 | my $font = $1;
|
|---|
| 150 | print $outhandle "HBPlug::HB_gettext - warning removed font $font\n"
|
|---|
| 151 | if ($font !~ /^arial$/i);
|
|---|
| 152 | }
|
|---|
| 153 |
|
|---|
| 154 | $line =~ s/<\/p>//ig; # remove </p> tags
|
|---|
| 155 | $line =~ s/<\/?(body|html|font)\b[^>]*>//ig; # remove any unwanted tags
|
|---|
| 156 |
|
|---|
| 157 | $$text .= $line;
|
|---|
| 158 | }
|
|---|
| 159 | #
|
|---|
| 160 | if ($self->{'input_encoding'} eq "iso_8859_1") {
|
|---|
| 161 | # convert to utf-8
|
|---|
| 162 | $$text=&unicode::unicode2utf8(&unicode::convert2unicode("iso_8859_1", $text));
|
|---|
| 163 | }
|
|---|
| 164 | # convert any alphanumeric character entities to their utf-8
|
|---|
| 165 | # equivalent for indexing purposes
|
|---|
| 166 | &ghtml::convertcharentities ($$text);
|
|---|
| 167 |
|
|---|
| 168 | $$text =~ s/\s+/ /g; # remove \n's
|
|---|
| 169 | }
|
|---|
| 170 |
|
|---|
| 171 | sub HB_clean_section {
|
|---|
| 172 | my $self = shift (@_);
|
|---|
| 173 | my ($section) = @_;
|
|---|
| 174 |
|
|---|
| 175 | # remove tags without a starting tag from the section
|
|---|
| 176 | my ($tag, $tagstart);
|
|---|
| 177 | while ($section =~ /<\/([^>]{1,10})>/) {
|
|---|
| 178 | $tag = $1;
|
|---|
| 179 | $tagstart = index($section, "<$tag");
|
|---|
| 180 | last if (($tagstart >= 0) && ($tagstart < index($section, "<\/$tag")));
|
|---|
| 181 | $section =~ s/<\/$tag>//;
|
|---|
| 182 | }
|
|---|
| 183 |
|
|---|
| 184 | # remove extra paragraph tags
|
|---|
| 185 | while ($section =~ s/<p\b[^>]*>\s*<p\b/<p/ig) {}
|
|---|
| 186 |
|
|---|
| 187 | # remove extra stuff at the end of the section
|
|---|
| 188 | while ($section =~ s/(<u>|<i>|<b>|<p\b[^>]*>| |\s)$//i) {}
|
|---|
| 189 |
|
|---|
| 190 | # add a newline at the beginning of each paragraph
|
|---|
| 191 | $section =~ s/(.)\s*<p\b/$1\n\n<p/gi;
|
|---|
| 192 |
|
|---|
| 193 | # add a newline every 80 characters at a word boundary
|
|---|
| 194 | # Note: this regular expression puts a line feed before
|
|---|
| 195 | # the last word in each section, even when it is not
|
|---|
| 196 | # needed.
|
|---|
| 197 | $section =~ s/(.{1,80})\s/$1\n/g;
|
|---|
| 198 |
|
|---|
| 199 | # fix up the image links
|
|---|
| 200 | $section =~ s/<img[^>]*?src=\"?([^\">]+)\"?[^>]*>/
|
|---|
| 201 | <center><img src=\"_httpdocimg_\/$1\"><\/center><br>/ig;
|
|---|
| 202 | $section =~ s/<<I>>\s*([^\.]+\.(png|jpg|gif))/
|
|---|
| 203 | <center><img src=\"_httpdocimg_\/$1\"><\/center><br>/ig;
|
|---|
| 204 |
|
|---|
| 205 | return $section;
|
|---|
| 206 | }
|
|---|
| 207 |
|
|---|
| 208 |
|
|---|
| 209 | sub shorten {
|
|---|
| 210 | my $self = shift (@_);
|
|---|
| 211 | my ($text) = @_;
|
|---|
| 212 |
|
|---|
| 213 | return "\"$text\"" if (length($text) < 100);
|
|---|
| 214 |
|
|---|
| 215 | return "\"" . substr ($text, 0, 50) . "\" ... \"" .
|
|---|
| 216 | substr ($text, length($text)-50) . "\"";
|
|---|
| 217 | }
|
|---|
| 218 |
|
|---|
| 219 | # if input_encoding is ascii we can call add_utf8_metadata
|
|---|
| 220 | # directly but if it's iso_8859_1 (the default) we need to call
|
|---|
| 221 | # add_metadata so that the ascii2utf8 conversion is done first
|
|---|
| 222 | # this should speed things up a little if processing an ascii only
|
|---|
| 223 | # document with input_encoding set to ascii
|
|---|
| 224 | sub HB_add_metadata {
|
|---|
| 225 | my $self = shift (@_);
|
|---|
| 226 | my ($doc_obj, $cursection, $field, $value) = @_;
|
|---|
| 227 |
|
|---|
| 228 | # All text should now be in utf-8
|
|---|
| 229 | # if ($self->{'input_encoding'} eq "ascii") {
|
|---|
| 230 | $doc_obj->add_utf8_metadata ($cursection, $field, $value);
|
|---|
| 231 | # } else {
|
|---|
| 232 | # $doc_obj->add_metadata ($cursection, $field, $value);
|
|---|
| 233 | # }
|
|---|
| 234 | }
|
|---|
| 235 |
|
|---|
| 236 | # return number of files processed, undef if can't process
|
|---|
| 237 | # Note that $base_dir might be "" and that $file might
|
|---|
| 238 | # include directories
|
|---|
| 239 | sub read {
|
|---|
| 240 | my $self = shift (@_);
|
|---|
| 241 | my ($pluginfo, $base_dir, $file, $metadata, $processor, $maxdocs, $total_count, $gli) = @_;
|
|---|
| 242 | my $outhandle = $self->{'outhandle'};
|
|---|
| 243 |
|
|---|
| 244 | # get the html filename and see if this is an HTML Book...
|
|---|
| 245 | my $jobnumber = $file;
|
|---|
| 246 | if ($file =~ /[\\\/]/) {
|
|---|
| 247 | ($jobnumber) = $file =~ /[\\\/]([^\\\/]+)$/;
|
|---|
| 248 | }
|
|---|
| 249 | return undef unless defined $jobnumber;
|
|---|
| 250 | my $htmlfile = &util::filename_cat($base_dir, $file, "$jobnumber.htm");
|
|---|
| 251 | return undef unless -e $htmlfile;
|
|---|
| 252 |
|
|---|
| 253 | print STDERR "<Processing n='$file' p='HBPlug'>\n" if ($gli);
|
|---|
| 254 | print $outhandle "HBPlug: processing $file\n";
|
|---|
| 255 |
|
|---|
| 256 | # read in the file and do basic html cleaning (removing header etc)
|
|---|
| 257 | my $html = "";
|
|---|
| 258 | $self->HB_read_html_file ($htmlfile, \$html);
|
|---|
| 259 | # html is in utf8
|
|---|
| 260 |
|
|---|
| 261 | # create a new document
|
|---|
| 262 | my $doc_obj = new doc ($file, "indexed_doc");
|
|---|
| 263 | $doc_obj->set_OIDtype ($processor->{'OIDtype'});
|
|---|
| 264 |
|
|---|
| 265 | # copy the book cover if it exists
|
|---|
| 266 | my $bookcover = &util::filename_cat($base_dir, $file, "$jobnumber.jpg");
|
|---|
| 267 | $doc_obj->associate_file($bookcover, "cover.jpg", "image/jpeg");
|
|---|
| 268 | $doc_obj->add_utf8_metadata($doc_obj->get_top_section(), "Plugin", "$self->{'plugin_type'}");
|
|---|
| 269 | $doc_obj->add_utf8_metadata($doc_obj->get_top_section(), "FileFormat", "HB");
|
|---|
| 270 | $doc_obj->add_utf8_metadata($doc_obj->get_top_section(), "FileSize", (-s $htmlfile));
|
|---|
| 271 |
|
|---|
| 272 | my $cursection = $doc_obj->get_top_section();
|
|---|
| 273 |
|
|---|
| 274 | # add metadata for top level of document
|
|---|
| 275 | foreach my $field (keys(%$metadata)) {
|
|---|
| 276 | # $metadata->{$field} may be an array reference
|
|---|
| 277 | if (ref ($metadata->{$field}) eq "ARRAY") {
|
|---|
| 278 | map {
|
|---|
| 279 | $self->HB_add_metadata ($doc_obj, $cursection, $field, $_);
|
|---|
| 280 | } @{$metadata->{$field}};
|
|---|
| 281 | } else {
|
|---|
| 282 | $self->HB_add_metadata ($doc_obj, $cursection, $field, $metadata->{$field});
|
|---|
| 283 | }
|
|---|
| 284 | }
|
|---|
| 285 |
|
|---|
| 286 | # process the file one section at a time
|
|---|
| 287 | my $curtoclevel = 1;
|
|---|
| 288 | my $firstsection = 1;
|
|---|
| 289 | while (length ($html) > 0) {
|
|---|
| 290 | if ($html =~ s/^.*?(?:<p\b[^>]*>)?((<b>|<i>|<u>|\s)*)<<TOC(\d+)>>\s*(.*?)<p\b/<p/i) {
|
|---|
| 291 | my $toclevel = $3;
|
|---|
| 292 | my $title = $4;
|
|---|
| 293 | my $sectiontext = "";
|
|---|
| 294 | if ($html =~ s/^(.*?)((?:<p\b[^>]*>)?((<b>|<i>|<u>|\s)*)<<TOC\d+>>)/$2/i) {
|
|---|
| 295 | $sectiontext = $1;
|
|---|
| 296 | } else {
|
|---|
| 297 | $sectiontext = $html;
|
|---|
| 298 | $html = "";
|
|---|
| 299 | }
|
|---|
| 300 |
|
|---|
| 301 | # remove tags and extra spaces from the title
|
|---|
| 302 | $title =~ s/<\/?[^>]+>//g;
|
|---|
| 303 | $title =~ s/^\s+|\s+$//g;
|
|---|
| 304 |
|
|---|
| 305 | # close any sections below the current level and
|
|---|
| 306 | # create a new section (special case for the firstsection)
|
|---|
| 307 | while (($curtoclevel > $toclevel) ||
|
|---|
| 308 | (!$firstsection && $curtoclevel == $toclevel)) {
|
|---|
| 309 | $cursection = $doc_obj->get_parent_section ($cursection);
|
|---|
| 310 | $curtoclevel--;
|
|---|
| 311 | }
|
|---|
| 312 | if ($curtoclevel+1 < $toclevel) {
|
|---|
| 313 | print $outhandle "WARNING - jump in toc levels in $htmlfile " .
|
|---|
| 314 | "from $curtoclevel to $toclevel\n";
|
|---|
| 315 | }
|
|---|
| 316 | while ($curtoclevel < $toclevel) {
|
|---|
| 317 | $curtoclevel++;
|
|---|
| 318 | $cursection =
|
|---|
| 319 | $doc_obj->insert_section($doc_obj->get_end_child($cursection));
|
|---|
| 320 | }
|
|---|
| 321 |
|
|---|
| 322 | # add the metadata to this section
|
|---|
| 323 | $self->HB_add_metadata ($doc_obj, $cursection, "Title", $title);
|
|---|
| 324 |
|
|---|
| 325 | # clean up the section html
|
|---|
| 326 | $sectiontext = $self->HB_clean_section($sectiontext);
|
|---|
| 327 |
|
|---|
| 328 | # associate any files
|
|---|
| 329 | map { $doc_obj->associate_file(&util::filename_cat ($base_dir, $file, $1), $1)
|
|---|
| 330 | if /_httpdocimg_\/([^\"]+)\"/; 0; }
|
|---|
| 331 | split (/(_httpdocimg_\/[^\"]+\")/, $sectiontext);
|
|---|
| 332 |
|
|---|
| 333 | # add the text for this section
|
|---|
| 334 | # All read text should now be in utf-8
|
|---|
| 335 | # if ($self->{'input_encoding'} eq "ascii") {
|
|---|
| 336 | $doc_obj->add_utf8_text ($cursection, $sectiontext);
|
|---|
| 337 | # } else {
|
|---|
| 338 | # $doc_obj->add_text ($cursection, $sectiontext);
|
|---|
| 339 | # }
|
|---|
| 340 | } else {
|
|---|
| 341 | print $outhandle "WARNING - leftover text\n" , $self->shorten($html),
|
|---|
| 342 | "\nin $htmlfile\n";
|
|---|
| 343 | last;
|
|---|
| 344 | }
|
|---|
| 345 | $firstsection = 0;
|
|---|
| 346 | }
|
|---|
| 347 |
|
|---|
| 348 | # add a OID
|
|---|
| 349 | $doc_obj->set_OID ();
|
|---|
| 350 |
|
|---|
| 351 | # process the document
|
|---|
| 352 | $processor->process($doc_obj, &util::filename_cat($file, "$jobnumber.htm"));
|
|---|
| 353 |
|
|---|
| 354 | return 1; # processed the file
|
|---|
| 355 | }
|
|---|
| 356 |
|
|---|
| 357 |
|
|---|
| 358 | 1;
|
|---|
