my $class = shift;
my $args = @_ == 1 ? $_[0] : { @_ };
if( exists $args->{'JSON'} ) {
- $DB::single = 1;
my @data = split( / \/\/ /, $args->{'JSON'} );
- print STDERR "Attempting to parse " . $data[2] . " into structure";
+ # print STDERR "Attempting to parse " . $data[2] . " into structure";
my $morph = Lingua::Features::Structure->from_string( $data[2] );
$args = { 'language' => $data[0], 'lemma' => $data[1],
'morphology' => $morph };
use TryCatch;
@EXPORT_OK = qw/ lemmatize_treetagger reading_lookup_treetagger treetagger_struct
- multext_struct /;
+ multext_struct lfs_morph_tags /;
=head1 NAME
return Lingua::TagSet::Multext->tag2structure( $pos );
}
+=head2 lfs_morph_tags
+
+Return a data structure describing the available parts of speech and their attributes
+from the Lingua::Features::Structure class currently defined.
+
+=cut
+
+sub lfs_morph_tags {
+ load('Lingua::Features::StructureType');
+ my $tagset = { 'structures' => [], 'features' => {} };
+ foreach my $lfs ( sort { _by_structid( $a->id, $b->id ) } Lingua::Features::StructureType->types() ) {
+ my $tsstruct = { 'id' => $lfs->id, 'use_features' => [] };
+ foreach my $ftid ( Lingua::Features::StructureType->type($lfs->id)->features ) {
+ my $ftype = $lfs->feature_type( $ftid );
+ my $tfstruct = { 'id' => $ftid, 'values' => [] };
+ foreach my $fval( $ftype->values ) {
+ push( @{$tfstruct->{'values'}},
+ { 'short' => $fval, 'long' => $ftype->value_name( $fval ) } );
+ }
+ push( @{$tsstruct->{'use_features'}}, $ftid );
+ $tagset->{'features'}->{$ftid} = $tfstruct;
+ }
+ push( @{$tagset->{'structures'}}, $tsstruct );
+ }
+ return $tagset;
+}
+
+sub _by_structid {
+ my( $a, $b ) = @_;
+ return -1 if $a eq 'cat';
+ return 1 if $b eq 'cat';
+ return $a cmp $b;
+}
+
1;
=head2 TODO
use strict;
use warnings;
-use Text::Tradition::Language::Base qw/ lemmatize_treetagger reading_lookup_treetagger /;
+use Text::Tradition::Language::Base qw/ lemmatize_treetagger reading_lookup_treetagger
+ lfs_morph_tags /;
use TryCatch;
=head1 NAME
return reading_lookup_treetagger( %opts );
}
+=head2 morphology_tags
+
+Return a data structure describing the available parts of speech and their attributes.
+
+=cut
+
+sub morphology_tags {
+ return lfs_morph_tags();
+}
+
# Utility function to turn a TreeTagger result into a WordForm
sub _parse_wordform {
my $tagresult = shift;
use strict;
use warnings;
use Module::Load qw/ load /;
-use Text::Tradition::Language::Base qw/ lemmatize_treetagger reading_lookup_treetagger treetagger_struct multext_struct /;
+use Text::Tradition::Language::Base qw/ lemmatize_treetagger reading_lookup_treetagger treetagger_struct multext_struct lfs_morph_tags /;
use TryCatch;
=head1 NAME
return reading_lookup_treetagger( %opts );
}
+=head2 morphology_tags
+
+Return a data structure describing the available parts of speech and their attributes.
+
+=cut
+
+sub morphology_tags {
+ return lfs_morph_tags();
+}
+
# Closure and utility function for the package lemmatizer
{
my $lemmatizer;
use strict;
use warnings;
use Module::Load;
-use Text::Tradition::Language::Base qw/ lemmatize_treetagger treetagger_struct /;
+use Morph::Perseus::Structure;
+use Text::Tradition::Language::Base qw/ lemmatize_treetagger treetagger_struct
+ lfs_morph_tags /;
use TryCatch;
=head1 NAME
return map { _perseus_lookup_str( $_ ) } @words;
}
+=head2 morphology_tags
+
+Return a data structure describing the available parts of speech and their attributes.
+
+=cut
+
+sub morphology_tags {
+ try {
+ load 'Morph::Perseus::Structure';
+ } catch {
+ warn "Not using Perseus Latin tags";
+ }
+ return lfs_morph_tags();
+}
+
+
{
my $morph;
}
@ret = @orig unless @ret;
- my @wordforms;
+ my %unique_wordforms;
foreach my $obj ( @ret ) {
- push( @wordforms, _wordform_from_row( $obj ) );
+ my $wf = _wordform_from_row( $obj );
+ $unique_wordforms{$wf->to_string} = $wf;
}
## TODO Use TreeTagger info - requires serious hacking of Lingua::TagSet
# Discard results that don't match the given TreeTagger POS, unless
# @ttmatch = @wordforms;
# }
# return @ttmatch;
- return @wordforms;
+ return values( %unique_wordforms );
}
sub _perseus_lookup_str {
warn "Could not create morphology structure from "
. $rowobj->code . ": $!";
}
- $DB::single = 1 unless $mpstruct;
my $lemma = $rowobj->lemma;
$lemma =~ s/^(\D+)\d*$/$1/;
my $wf = Text::Tradition::Collation::Reading::WordForm->new(
package stemmaweb::Controller::Relation;
use Moose;
+use Module::Load;
use namespace::autoclean;
use TryCatch;
$c->stash->{'template'} = 'relate.tt';
}
-=head2 help
-
- GET relation/help
-
-Returns the help window HTML.
-
-=cut
-
-sub help :Local :Args(0) {
- my( $self, $c ) = @_;
- $c->stash->{'template'} = 'relatehelp.tt';
-}
-
=head2 definitions
GET relation/definitions
$c->stash->{'startseg'} = $startseg if defined $startseg;
$c->stash->{'svg_string'} = $svg_str;
$c->stash->{'text_title'} = $tradition->name;
+ $c->stash->{'text_lang'} = $tradition->language;
$c->stash->{'template'} = 'relate.tt';
}
+=head2 help
+
+ GET relation/help/$language
+
+Returns the help window HTML.
+
+=cut
+
+sub help :Local :Args(1) {
+ my( $self, $c, $lang ) = @_;
+ # Display the morphological help for the language if it is defined.
+ if( $lang && $lang ne 'Default' ) {
+ my $mod = 'Text::Tradition::Language::' . $lang;
+ try {
+ load( $mod );
+ } catch {
+ $c->log->debug("Warning: could not load $mod");
+ }
+ my $has_mod = $mod->can('morphology_tags');
+ $DB::single = 1;
+ if( $has_mod ) {
+ my $tagset = &$has_mod;
+ $c->stash->{'tagset'} = $tagset;
+ }
+ }
+ $c->stash->{'template'} = 'relatehelp.tt';
+}
+
=head2 relationships
GET relation/$textid/relationships
<div id="topbanner">
<div id="bannerinfo">
- <a href="help" title="Relationship mapper help" class="helptag">Help / About</a>
+ <a href="help/[% text_lang %]" title="Relationship mapper help" class="helptag">Help / About</a>
</div>
<h1>Relationship mapper</h1>
<h2>[% text_title %]</h2>
<p>The premise of the tool is that, once a set of texts has been collated, there will be a need to chart the relationships between the variants—are they substantially the same word? Different words meaning the same thing? Is one an orthographic variant of the other that should be excluded from any eventual apparatus?</p>
- <h3>Instructions for use</h3>
+ <h3>Making relationships between words</h3>
<p>The tool itself is an interface for allowing these relationships to be declared. The collation is presented as a variant graph running from left to right. In a variant graph, each node is a reading, and each witness takes a single path through the readings from beginning to end. When readings appear vertically aligned with each other, it is an indication that they are variant readings, occurring at the same point in the text.
<p>The relationships are displayed as colored paths between readings; while in 'edit' mode, clicking on a relationship path will display the information associated with it, and give the user an option to delete it. Deletion of a 'global' relationship will remove that relationship throughout the graph. When you are ready to move elsewhere in the graph, click the 'hand' icon to return to select mode.<p>
- <p>Please note that this tool is known to work with recent versions of Firefox (e.g. 8, 9, 10); it is known not to work with Safari and Chrome.</p>
+[% IF language != 'NONE' %]
+ <h3>Adding [% language %] morphological information to readings</h3>
+ <p>It is also possible to add morphological information to the readings in this text (that is, lemma and morphological tagging). Double click on any reading to bring up the morphology info. The options therein are:</p>
+
+ <ul>
+ <li>Indicate if the word is a nonsense word</li>
+ <li>Indicate if the word (whether sensical or not) cannot be grammatically correct here</li>
+ <li>Indicate the "normal" form of the word for lemmatization purposes. Any readings linked to this one via 'spelling' or 'orthographic' links will appear when the entry text box is clicked.</li>
+ <li>Indicate the correct morphological form of this reading. The shorthand syntax is:<br/>
+ <LEMMA> // cat@<CATEGORY> (type@<TYPE> etc.)<br/>
+ Each form must have at least a lemma and a category, and may have a number of additional features depending on the category (e.g. type, gender, number, case.) The possible options are as follows:
+ <ul>
+[% FOREACH stype IN tagset.structures -%]
+ <li>Category [% stype.id -%]
+[% IF stype.use_features.size -%]
+ has features [% stype.use_features.join(', ') %]
+[% END -%]
+ </li>
+[% END -%]
+ </ul>
+ <ul>
+[% FOREACH ftype IN tagset.features.keys -%]
+ <li>Feature [% ftype %] can have values:
+ <ul>
+[% FOREACH fval IN tagset.features.$ftype.values -%]
+ <li>[% fval.short %] => [% fval.long %]</li>
+[% END -%]
+ </ul>
+ </li>
+[% END -%]
+ </ul>
+ </li>
+ </ul>
+
+ <p>If initial lemmatization has been performed on the text, a number of readings may appear in yellow rather than green; this means that there are multiple possible morphologies for the reading in question. Double click on the reading to select and save the correct morphology.</p>
+[% END -%]
[% PROCESS footer.tt %]
\ No newline at end of file
projects / scpubgit/stemmatology.git / commitdiff