| 1 | ###########################################################################
|
|---|
| 2 | #
|
|---|
| 3 | # TextPlugin.pm -- simple text 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 | # creates simple single-level document. Adds Title metadata
|
|---|
| 27 | # of first line of text (up to 100 characters long).
|
|---|
| 28 |
|
|---|
| 29 | package TextPlugin;
|
|---|
| 30 |
|
|---|
| 31 | use ReadTextFile;
|
|---|
| 32 |
|
|---|
| 33 | use strict;
|
|---|
| 34 | no strict 'refs'; # allow filehandles to be variables and viceversa
|
|---|
| 35 | no strict 'subs';
|
|---|
| 36 |
|
|---|
| 37 | sub BEGIN {
|
|---|
| 38 | @TextPlugin::ISA = ('ReadTextFile');
|
|---|
| 39 | }
|
|---|
| 40 |
|
|---|
| 41 | my $arguments =
|
|---|
| 42 | [ { 'name' => "process_exp",
|
|---|
| 43 | 'desc' => "{BasePlugin.process_exp}",
|
|---|
| 44 | 'type' => "regexp",
|
|---|
| 45 | 'deft' => &get_default_process_exp(),
|
|---|
| 46 | 'reqd' => "no" } ,
|
|---|
| 47 | { 'name' => "title_sub",
|
|---|
| 48 | 'desc' => "{TextPlugin.title_sub}",
|
|---|
| 49 | 'type' => "regexp",
|
|---|
| 50 | 'deft' => "",
|
|---|
| 51 | 'reqd' => "no" } ];
|
|---|
| 52 |
|
|---|
| 53 | my $options = { 'name' => "TextPlugin",
|
|---|
| 54 | 'desc' => "{TextPlugin.desc}",
|
|---|
| 55 | 'abstract' => "no",
|
|---|
| 56 | 'inherits' => "yes",
|
|---|
| 57 | 'srcreplaceable' => "yes", # Source docs in regular txt format can be replaced with GS-generated html
|
|---|
| 58 | 'args' => $arguments };
|
|---|
| 59 |
|
|---|
| 60 |
|
|---|
| 61 | sub new {
|
|---|
| 62 | my ($class) = shift (@_);
|
|---|
| 63 | my ($pluginlist,$inputargs,$hashArgOptLists) = @_;
|
|---|
| 64 | push(@$pluginlist, $class);
|
|---|
| 65 |
|
|---|
| 66 | push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
|
|---|
| 67 | push(@{$hashArgOptLists->{"OptList"}},$options);
|
|---|
| 68 |
|
|---|
| 69 | my $self = new ReadTextFile($pluginlist, $inputargs, $hashArgOptLists);
|
|---|
| 70 |
|
|---|
| 71 | return bless $self, $class;
|
|---|
| 72 | }
|
|---|
| 73 |
|
|---|
| 74 | sub get_default_process_exp {
|
|---|
| 75 | my $self = shift (@_);
|
|---|
| 76 |
|
|---|
| 77 | return q^(?i)\.te?xt$^;
|
|---|
| 78 | }
|
|---|
| 79 |
|
|---|
| 80 | # do plugin specific processing of doc_obj
|
|---|
| 81 | sub process {
|
|---|
| 82 | my $self = shift (@_);
|
|---|
| 83 | my ($textref, $pluginfo, $base_dir, $file, $metadata, $doc_obj, $gli) = @_;
|
|---|
| 84 | my $outhandle = $self->{'outhandle'};
|
|---|
| 85 |
|
|---|
| 86 | my $cursection = $doc_obj->get_top_section();
|
|---|
| 87 | # get title metadata
|
|---|
| 88 | # (don't need to get title if it has been passed
|
|---|
| 89 | # in from another plugin)
|
|---|
| 90 | if (!defined $metadata->{'Title'}) {
|
|---|
| 91 | my $title = $self->get_title_metadata($textref);
|
|---|
| 92 | $doc_obj->add_utf8_metadata ($cursection, "Title", $title);
|
|---|
| 93 | }
|
|---|
| 94 | # Add FileFormat metadata
|
|---|
| 95 | $doc_obj->add_metadata($cursection, "FileFormat", "Text");
|
|---|
| 96 |
|
|---|
| 97 | # insert preformat tags and add text to document object
|
|---|
| 98 | $self->text_to_html($textref); # modifies the text
|
|---|
| 99 | $doc_obj->add_utf8_text($cursection, $$textref);
|
|---|
| 100 |
|
|---|
| 101 | return 1;
|
|---|
| 102 | }
|
|---|
| 103 |
|
|---|
| 104 | sub get_title_metadata {
|
|---|
| 105 | my $self = shift (@_);
|
|---|
| 106 | my ($textref) = @_;
|
|---|
| 107 |
|
|---|
| 108 | my ($title) = $$textref;
|
|---|
| 109 | $title =~ /^\s+/s;
|
|---|
| 110 | if (defined $self->{'title_sub'} && $self->{'title_sub'}) {
|
|---|
| 111 | $title =~ s/$self->{'title_sub'}//;
|
|---|
| 112 | }
|
|---|
| 113 | $title =~ /^\s*([^\n]*)/s; $title=$1;
|
|---|
| 114 | if (length($title) > 100) {
|
|---|
| 115 | $title = substr ($title, 0, 100) . "...";
|
|---|
| 116 | }
|
|---|
| 117 | $title =~ s/\[/&\#91;/g;
|
|---|
| 118 | $title =~ s/\[/&\#93;/g;
|
|---|
| 119 | $title =~ s/\</&\#60;/g;
|
|---|
| 120 | $title =~ s/\>/&\#62;/g;
|
|---|
| 121 |
|
|---|
| 122 | return $title;
|
|---|
| 123 | }
|
|---|
| 124 |
|
|---|
| 125 | sub text_to_html {
|
|---|
| 126 | my $self = shift (@_);
|
|---|
| 127 | my ($textref) = @_;
|
|---|
| 128 |
|
|---|
| 129 | # we need to escape the escape character, or else mg will convert into
|
|---|
| 130 | # eg literal newlines, instead of leaving the text as '\n'
|
|---|
| 131 | $$textref =~ s/\\/\\\\/g; # macro language
|
|---|
| 132 | $$textref =~ s/_/\\_/g; # macro language
|
|---|
| 133 | $$textref =~ s/</</g;
|
|---|
| 134 | $$textref =~ s/>/>/g;
|
|---|
| 135 |
|
|---|
| 136 | # insert preformat tags and add text to document object
|
|---|
| 137 | $$textref = "<pre>\n$$textref\n</pre>";
|
|---|
| 138 | }
|
|---|
| 139 |
|
|---|
| 140 |
|
|---|
| 141 | # replace_srcdoc_with_html.pl requires all subroutines that support src_replaceable
|
|---|
| 142 | # to contain a method called tmp_area_convert_file - this is indeed the case with all
|
|---|
| 143 | # Perl modules that are subclasses of ConvertToPlug.pm, but as we want TextPlugin to also
|
|---|
| 144 | # be srcreplaceable and because TextPlugin does not inherit from ConvertToPlug.pm, we have
|
|---|
| 145 | # a similar subroutine with the same name here.
|
|---|
| 146 | sub tmp_area_convert_file {
|
|---|
| 147 | my $self = shift (@_);
|
|---|
| 148 | my ($output_ext, $input_filename) = @_;
|
|---|
| 149 |
|
|---|
| 150 | my $outhandle = $self->{'outhandle'};
|
|---|
| 151 | #my $failhandle = $self->{'failhandle'};
|
|---|
| 152 |
|
|---|
| 153 | # derive output filename from input filename
|
|---|
| 154 | my ($tailname, $dirname, $suffix)
|
|---|
| 155 | = &File::Basename::fileparse($input_filename, "\\.[^\\.]+\$");
|
|---|
| 156 |
|
|---|
| 157 | # Softlink to collection tmp dir
|
|---|
| 158 | # First find a temporary directory to create the output file in
|
|---|
| 159 | my $tmp_dirname = $dirname;
|
|---|
| 160 | if(defined $ENV{'GSDLCOLLECTDIR'}) {
|
|---|
| 161 | $tmp_dirname = $ENV{'GSDLCOLLECTDIR'};
|
|---|
| 162 | } elsif(defined $ENV{'GSDLHOME'}) {
|
|---|
| 163 | $tmp_dirname = $ENV{'GSDLHOME'};
|
|---|
| 164 | }
|
|---|
| 165 | $tmp_dirname = &util::filename_cat($tmp_dirname, "tmp");
|
|---|
| 166 | &util::mk_dir($tmp_dirname) if (!-e $tmp_dirname);
|
|---|
| 167 |
|
|---|
| 168 | # convert to utf-8 otherwise we have problems with the doc.xml file
|
|---|
| 169 | # later on
|
|---|
| 170 | $tailname = $self->SUPER::filepath_to_utf8($tailname) unless &unicode::check_is_utf8($tailname);
|
|---|
| 171 |
|
|---|
| 172 | $suffix = lc($suffix);
|
|---|
| 173 | my $tmp_filename = &util::filename_cat($tmp_dirname, "$tailname$suffix");
|
|---|
| 174 |
|
|---|
| 175 | # Make sure we have the absolute path to the input file
|
|---|
| 176 | # (If gsdl is remote, we're given relative path to input file, of the form import/tailname.suffix
|
|---|
| 177 | # But we can't softlink to relative paths. Therefore, we need to ensure that
|
|---|
| 178 | # the input_filename is the absolute path.
|
|---|
| 179 | my $ensure_path_absolute = 1; # true
|
|---|
| 180 |
|
|---|
| 181 | # Now make the softlink, so we don't accidentally damage the input file
|
|---|
| 182 | # softlinking creates a symbolic link to (or, if that's not possible, it makes a copy of) the original
|
|---|
| 183 | &util::soft_link($input_filename, $tmp_filename, $ensure_path_absolute);
|
|---|
| 184 |
|
|---|
| 185 | my $verbosity = $self->{'verbosity'};
|
|---|
| 186 | if ($verbosity > 0) {
|
|---|
| 187 | # need this output statement, as GShell.java's runRemote() sets status to CANCELLED
|
|---|
| 188 | # if there is no output! (Therefore, it only had this adverse affect when running GSDL remotely)
|
|---|
| 189 | print $outhandle "Converting $tailname$suffix to html\n";
|
|---|
| 190 | }
|
|---|
| 191 |
|
|---|
| 192 | #my $output_filename = $tailname$output_ext; #output_ext has to be html!
|
|---|
| 193 | my $output_filename = &util::filename_cat($tmp_dirname, $tailname.".html");
|
|---|
| 194 |
|
|---|
| 195 | # Read contents of text file line by line into an array
|
|---|
| 196 | # create an HTML file from the text file
|
|---|
| 197 | # Recreate the original file for writing the updated contents
|
|---|
| 198 | unless(open(TEXT, "<$tmp_filename")) { # open it as a new file for writing
|
|---|
| 199 | print STDERR "TextPlugin.pm: Unable to open and read from $tmp_filename for converting to html...ERROR: $!\n";
|
|---|
| 200 | return ""; # no file name
|
|---|
| 201 | }
|
|---|
| 202 |
|
|---|
| 203 | # Read the entire file at once
|
|---|
| 204 | my $text;
|
|---|
| 205 | {
|
|---|
| 206 | local $/ = undef; # Now can read the entire file at once. From http://perl.plover.com/local.html
|
|---|
| 207 | $text = <TEXT>; # File is read in as one single 'line'
|
|---|
| 208 | }
|
|---|
| 209 | close(TEXT); # close the file
|
|---|
| 210 |
|
|---|
| 211 | # Get the title before embedding the text in pre tags
|
|---|
| 212 | my $title = $self->get_title_metadata(\$text);
|
|---|
| 213 |
|
|---|
| 214 | # Now convert the text
|
|---|
| 215 | $self->text_to_html(\$text);
|
|---|
| 216 |
|
|---|
| 217 | # try creating this new file writing and try opening it for writing, else exit with error value
|
|---|
| 218 | unless(open(HTML, ">$output_filename")) { # open the new html file for writing
|
|---|
| 219 | print STDERR "TextPlugin.pm: Unable to create $output_filename for writing $tailname$suffix txt converted to html...ERROR: $!\n";
|
|---|
| 220 | return ""; # no filename
|
|---|
| 221 | }
|
|---|
| 222 | # write the html contents of the text (which is embedded in <pre> tags) out to the file with proper enclosing tags
|
|---|
| 223 | print HTML "<html>\n<head>\n<title>$title</title>\n</head>\n<body>\n";
|
|---|
| 224 | print HTML $text;
|
|---|
| 225 | print HTML "\n</body>\n</html>";
|
|---|
| 226 | close HTML;
|
|---|
| 227 |
|
|---|
| 228 | # remove the copy of the original file/remove the symbolic link to original file
|
|---|
| 229 | &util::rm($tmp_filename);
|
|---|
| 230 |
|
|---|
| 231 | return $output_filename; # return the output file path
|
|---|
| 232 | }
|
|---|
| 233 |
|
|---|
| 234 |
|
|---|
| 235 | 1;
|
|---|
