1 package Text::Tradition;
3 use JSON qw / from_json /;
6 use Text::Tradition::Collation;
7 use Text::Tradition::Stemma;
8 use Text::Tradition::Witness;
9 use Text::Tradition::User;
11 use vars qw( $VERSION );
16 isa => 'Text::Tradition::Collation',
17 writer => '_save_collation',
20 has 'witness_hash' => (
22 isa => 'HashRef[Text::Tradition::Witness]',
26 del_witness => 'delete',
27 has_witness => 'exists',
28 witnesses => 'values',
30 default => sub { {} },
36 default => 'Tradition',
42 predicate => 'has_language',
47 isa => 'ArrayRef[Text::Tradition::Stemma]',
49 stemmata => 'elements',
50 _add_stemma => 'push',
52 stemma_count => 'count',
53 clear_stemmata => 'clear',
55 default => sub { [] },
60 isa => 'Text::Tradition::User',
62 predicate => 'has_user',
63 clearer => 'clear_user',
70 default => sub { 0; },
73 # Create the witness before trying to add it
74 around 'add_witness' => sub {
77 # TODO allow add of a Witness object?
78 my %args = @_ == 1 ? %{$_[0]} : @_;
79 $args{'tradition'} = $self;
80 $args{'language'} = $self->language
81 if( $self->language && !exists $args{'language'} );
82 my $new_wit = Text::Tradition::Witness->new( %args );
83 $self->$orig( $new_wit->sigil => $new_wit );
87 # Allow deletion of witness by object as well as by sigil
88 around 'del_witness' => sub {
92 foreach my $arg ( @_ ) {
94 ref( $arg ) eq 'Text::Tradition::Witness' ? $arg->sigil : $arg );
96 return $self->$orig( @key_args );
99 # Don't allow an empty hash value
100 around 'witness' => sub {
101 my( $orig, $self, $arg ) = @_;
102 return unless $self->has_witness( $arg );
103 return $self->$orig( $arg );
108 Text::Tradition - a software model for a set of collated texts
113 my $t = Text::Tradition->new(
114 'name' => 'this is a text',
116 'file' => '/path/to/tei_parallel_seg_file.xml' );
118 my @text_wits = $t->witnesses();
119 my $manuscript_a = $t->witness( 'A' );
121 $t = Text::Tradition->new();
122 $t->add_witness( 'sourcetype' => 'xmldesc',
123 'file' => '/path/to/teitranscription.xml' );
124 $t->add_witness( 'sourcetype => 'plaintext', 'sigil' => 'Q',
125 'string' => 'The quick brown fox jumped over the lazy dogs' );
129 my $text_path_svg = $t->collation->as_svg();
130 ## See Text::Tradition::Collation for more on text collation itself
134 Text::Tradition is a library for representation and analysis of collated
135 texts, particularly medieval ones. A 'tradition' refers to the aggregation
136 of surviving versions of a text, generally preserved in multiple
137 manuscripts (or 'witnesses'). A Tradition object thus has one more more
138 Witnesses, as well as a Collation that represents the unity of all versions
145 Creates and returns a new text tradition object. The following options are
152 =item B<name> - The name of the text.
156 Initialization based on a collation file:
160 =item B<input> - The input format of the collation file. Can be one of the
165 =item * Self - a GraphML format produced by this module
167 =item * CollateX - a GraphML format produced by CollateX
169 =item * CTE - a TEI XML format produced by Classical Text Editor
171 =item * JSON - an alignment table in JSON format, as produced by CollateX and other tools
173 =item * KUL - a specific CSV format for variants, not documented here
175 =item * TEI - a TEI parallel segmentation format file
177 =item * Tabular - a comma- or tab-separated collation. Takes an additional
178 option, 'sep_char', which defaults to the tab character.
182 =item B<file> - The name of the file that contains the data. One of 'file'
183 or 'string' should be specified.
185 =item B<string> - A text string that contains the data. One of 'file' or
186 'string' should be specified.
188 =item B<base> - The name of a text file that contains the base text, to be
189 used with input formats that require it (currently only KUL).
193 Initialization based on a list of witnesses [NOT YET IMPLEMENTED]:
197 =item B<witnesses> - A reference to an array of Text::Tradition::Witness
198 objects that carry the text to be collated.
200 =item B<collator> - A reference to a collation program that will accept
207 Return the Text::Tradition::Witness objects associated with this tradition,
210 =head2 B<witness>( $sigil )
212 Returns the Text::Tradition::Witness object whose sigil is $sigil, or undef
213 if there is no such object within the tradition.
215 =head2 B<add_witness>( %opts )
217 Instantiate a new witness with the given options (see documentation for
218 Text::Tradition::Witness) and add it to the tradition.
220 =head2 B<del_witness>( $sigil )
222 Delete the witness with the given sigil from the tradition. Returns the
223 witness object for the deleted witness.
227 use_ok( 'Text::Tradition', "can use module" );
229 my $t = Text::Tradition->new( 'name' => 'empty' );
230 is( ref( $t ), 'Text::Tradition', "initialized an empty Tradition object" );
231 is( $t->name, 'empty', "object has the right name" );
232 is( scalar $t->witnesses, 0, "object has no witnesses" );
234 my $simple = 't/data/simple.txt';
235 my $s = Text::Tradition->new(
237 'input' => 'Tabular',
240 is( ref( $s ), 'Text::Tradition', "initialized a Tradition object" );
241 is( $s->name, 'inline', "object has the right name" );
242 is( scalar $s->witnesses, 3, "object has three witnesses" );
244 my $wit_a = $s->witness('A');
245 is( ref( $wit_a ), 'Text::Tradition::Witness', "Found a witness A" );
247 is( $wit_a->sigil, 'A', "Witness A has the right sigil" );
249 is( $s->witness('X'), undef, "There is no witness X" );
250 ok( !exists $s->{'witnesses'}->{'X'}, "Witness key X not created" );
252 my $wit_d = $s->add_witness( 'sigil' => 'D', 'sourcetype' => 'collation' );
253 is( ref( $wit_d ), 'Text::Tradition::Witness', "new witness created" );
254 is( $wit_d->sigil, 'D', "witness has correct sigil" );
255 is( scalar $s->witnesses, 4, "object now has four witnesses" );
257 my $del = $s->del_witness( 'D' );
258 is( $del, $wit_d, "Deleted correct witness" );
259 is( scalar $s->witnesses, 3, "object has three witnesses again" );
261 # TODO test initialization by witness list when we have it
269 my( $self, $init_args ) = @_;
271 # First, make a collation object. This will use only those arguments in
272 # init_args that apply to the collation.
273 my $collation = Text::Tradition::Collation->new( %$init_args,
274 'tradition' => $self );
275 $self->_save_collation( $collation );
277 if( exists $init_args->{'input'} ) {
278 # Call the appropriate parser on the given data
279 my @format_standalone = qw/ Self CollateText CollateX CTE JSON TEI Tabular /;
280 my @format_basetext = qw/ KUL /;
282 my $format = $init_args->{'input'};
283 if( $format && !( grep { $_ eq $format } @format_standalone )
284 && !( grep { $_ eq $format } @format_basetext ) ) {
285 warn "Unrecognized input format $format; not parsing";
288 if( $format && grep { $_ eq $format } @format_basetext ) {
290 if( !exists $init_args->{'base'} ) {
291 warn "Cannot make a collation from $format without a base text";
296 # Now do the parsing.
299 $format = 'BaseText'; # Use the BaseText module for parsing,
300 # but retain the original input arg.
302 my $mod = "Text::Tradition::Parser::$format";
304 $mod->can('parse')->( $self, $init_args );
310 =head2 add_json_witnesses( $jsonstring, $options )
312 Adds a set of witnesses from a JSON array specification. This is a wrapper
313 to parse the JSON and call add_witness (with the specified $options) for
314 each element therein.
318 sub add_json_witnesses {
319 my( $self, $jsonstr, $extraopts ) = @_;
320 my $witarray = from_json( $jsonstr );
321 foreach my $witspec ( @{$witarray->{witnesses}} ) {
322 my $opts = $extraopts || {};
323 $opts->{'sourcetype'} = 'json';
324 $opts->{'object'} = $witspec;
325 $self->add_witness( $opts );
329 =head2 add_stemma( $dotfile )
331 Initializes a Text::Tradition::Stemma object from the given dotfile,
332 and associates it with the tradition.
338 my $t = Text::Tradition->new(
339 'name' => 'simple test',
340 'input' => 'Tabular',
341 'file' => 't/data/simple.txt',
344 is( $t->stemma_count, 0, "No stemmas added yet" );
346 ok( $s = $t->add_stemma( dotfile => 't/data/simple.dot' ), "Added a simple stemma" );
347 is( ref( $s ), 'Text::Tradition::Stemma', "Got a stemma object returned" );
348 is( $t->stemma_count, 1, "Tradition claims to have a stemma" );
349 is( $t->stemma(0), $s, "Tradition hands back the right stemma" );
359 if( $opts{'dotfile'} ) {
360 open $stemma_fh, '<', $opts{'dotfile'}
361 or warn "Could not open file " . $opts{'dotfile'};
362 } elsif( $opts{'dot'} ) {
363 my $str = $opts{'dot'};
364 open $stemma_fh, '<', \$str;
367 binmode $stemma_fh, ':utf8';
368 my $stemma = Text::Tradition::Stemma->new(
369 'collation' => $self->collation,
370 'dot' => $stemma_fh );
371 $self->_add_stemma( $stemma ) if $stemma;
377 unless( $self->has_language ) {
378 warn "Please set a language to lemmatize a tradition";
381 my $mod = "Text::Tradition::Language::" . $self->language;
383 $mod->can( 'lemmatize' )->( $self );
387 __PACKAGE__->meta->make_immutable;
394 =item * Allow tradition to be initialized via passing to a collator.
400 This package is free software and is provided "as is" without express
401 or implied warranty. You can redistribute it and/or modify it under
402 the same terms as Perl itself.
406 Tara L Andrews E<lt>aurum@cpan.orgE<gt>