1 package Text::Tradition;
5 use Text::Tradition::Collation;
6 use Text::Tradition::Stemma;
7 use Text::Tradition::Witness;
9 use vars qw( $VERSION );
14 isa => 'Text::Tradition::Collation',
15 writer => '_save_collation',
18 has 'witness_hash' => (
20 isa => 'HashRef[Text::Tradition::Witness]',
24 del_witness => 'delete',
25 has_witness => 'exists',
26 witnesses => 'values',
28 default => sub { {} },
34 default => 'Tradition',
39 isa => 'ArrayRef[Text::Tradition::Stemma]',
41 stemmata => 'elements',
42 _add_stemma => 'push',
44 stemma_count => 'count',
45 clear_stemmata => 'clear',
47 default => sub { [] },
50 # Create the witness before trying to add it
51 around 'add_witness' => sub {
54 # TODO allow add of a Witness object?
55 my $new_wit = Text::Tradition::Witness->new( @_ );
56 $self->$orig( $new_wit->sigil => $new_wit );
60 # Allow deletion of witness by object as well as by sigil
61 around 'del_witness' => sub {
65 foreach my $arg ( @_ ) {
67 ref( $arg ) eq 'Text::Tradition::Witness' ? $arg->sigil : $arg );
69 return $self->$orig( @key_args );
72 # Don't allow an empty hash value
73 around 'witness' => sub {
74 my( $orig, $self, $arg ) = @_;
75 return unless $self->has_witness( $arg );
76 return $self->$orig( $arg );
81 Text::Tradition - a software model for a set of collated texts
86 my $t = Text::Tradition->new(
87 'name' => 'this is a text',
89 'file' => '/path/to/tei_parallel_seg_file.xml' );
91 my @text_wits = $t->witnesses();
92 my $manuscript_a = $t->witness( 'A' );
93 my $new_ms = $t->add_witness( 'sigil' => 'B' );
95 my $text_path_svg = $t->collation->as_svg();
96 ## See Text::Tradition::Collation for more on text collation itself
100 Text::Tradition is a library for representation and analysis of collated
101 texts, particularly medieval ones. A 'tradition' refers to the aggregation
102 of surviving versions of a text, generally preserved in multiple
103 manuscripts (or 'witnesses'). A Tradition object thus has one more more
104 Witnesses, as well as a Collation that represents the unity of all versions
111 Creates and returns a new text tradition object. The following options are
118 =item B<name> - The name of the text.
122 Initialization based on a collation file:
126 =item B<input> - The input format of the collation file. Can be one of the
131 =item * Self - a GraphML format produced by this module
133 =item * CollateX - a GraphML format produced by CollateX
135 =item * CTE - a TEI XML format produced by Classical Text Editor
137 =item * JSON - an alignment table in JSON format, as produced by CollateX and other tools
139 =item * KUL - a specific CSV format for variants, not documented here
141 =item * TEI - a TEI parallel segmentation format file
143 =item * Tabular - a comma- or tab-separated collation. Takes an additional
144 option, 'sep_char', which defaults to the tab character.
148 =item B<file> - The name of the file that contains the data. One of 'file'
149 or 'string' should be specified.
151 =item B<string> - A text string that contains the data. One of 'file' or
152 'string' should be specified.
154 =item B<base> - The name of a text file that contains the base text, to be
155 used with input formats that require it (currently only KUL).
159 Initialization based on a list of witnesses [NOT YET IMPLEMENTED]:
163 =item B<witnesses> - A reference to an array of Text::Tradition::Witness
164 objects that carry the text to be collated.
166 =item B<collator> - A reference to a collation program that will accept
173 Return the Text::Tradition::Witness objects associated with this tradition,
176 =head2 B<witness>( $sigil )
178 Returns the Text::Tradition::Witness object whose sigil is $sigil, or undef
179 if there is no such object within the tradition.
181 =head2 B<add_witness>( %opts )
183 Instantiate a new witness with the given options (see documentation for
184 Text::Tradition::Witness) and add it to the tradition.
186 =head2 B<del_witness>( $sigil )
188 Delete the witness with the given sigil from the tradition. Returns the
189 witness object for the deleted witness.
193 use_ok( 'Text::Tradition', "can use module" );
195 my $t = Text::Tradition->new( 'name' => 'empty' );
196 is( ref( $t ), 'Text::Tradition', "initialized an empty Tradition object" );
197 is( $t->name, 'empty', "object has the right name" );
198 is( scalar $t->witnesses, 0, "object has no witnesses" );
200 my $simple = 't/data/simple.txt';
201 my $s = Text::Tradition->new(
203 'input' => 'Tabular',
206 is( ref( $s ), 'Text::Tradition', "initialized a Tradition object" );
207 is( $s->name, 'inline', "object has the right name" );
208 is( scalar $s->witnesses, 3, "object has three witnesses" );
210 my $wit_a = $s->witness('A');
211 is( ref( $wit_a ), 'Text::Tradition::Witness', "Found a witness A" );
213 is( $wit_a->sigil, 'A', "Witness A has the right sigil" );
215 is( $s->witness('X'), undef, "There is no witness X" );
216 ok( !exists $s->{'witnesses'}->{'X'}, "Witness key X not created" );
218 my $wit_d = $s->add_witness( 'sigil' => 'D' );
219 is( ref( $wit_d ), 'Text::Tradition::Witness', "new witness created" );
220 is( $wit_d->sigil, 'D', "witness has correct sigil" );
221 is( scalar $s->witnesses, 4, "object now has four witnesses" );
223 my $del = $s->del_witness( 'D' );
224 is( $del, $wit_d, "Deleted correct witness" );
225 is( scalar $s->witnesses, 3, "object has three witnesses again" );
227 # TODO test initialization by witness list when we have it
235 my( $self, $init_args ) = @_;
237 if( exists $init_args->{'witnesses'} ) {
238 # We got passed an uncollated list of witnesses. Make a
239 # witness object for each witness, and then send them to the
242 foreach my $wit ( %{$init_args->{'witnesses'}} ) {
243 # Each item in the list is either a string or an arrayref.
244 # If it's a string, it is a filename; if it's an arrayref,
245 # it is a tuple of 'sigil, file'. Handle either case.
247 if( ref( $wit ) eq 'ARRAY' ) {
248 $args = { 'sigil' => $wit->[0],
249 'file' => $wit->[1] };
251 $args = { 'sigil' => chr( $autosigil+65 ),
255 $self->witnesses->add_witness( $args );
256 # TODO Now how to collate these?
259 # Else we need to parse some collation data. Make a Collation object
260 my $collation = Text::Tradition::Collation->new( %$init_args,
261 'tradition' => $self );
262 $self->_save_collation( $collation );
264 # Call the appropriate parser on the given data
265 my @format_standalone = qw/ Self CollateText CollateX CTE JSON TEI Tabular /;
266 my @format_basetext = qw/ KUL /;
268 my $format = $init_args->{'input'};
269 if( $format && !( grep { $_ eq $format } @format_standalone )
270 && !( grep { $_ eq $format } @format_basetext ) ) {
271 warn "Unrecognized input format $format; not parsing";
274 if( $format && grep { $_ eq $format } @format_basetext ) {
276 if( !exists $init_args->{'base'} ) {
277 warn "Cannot make a collation from $format without a base text";
282 # Now do the parsing.
285 $format = 'BaseText'; # Use the BaseText module for parsing,
286 # but retain the original input arg.
288 my $mod = "Text::Tradition::Parser::$format";
290 $mod->can('parse')->( $self, $init_args );
295 =head2 add_stemma( $dotfile )
297 Initializes a Text::Tradition::Stemma object from the given dotfile,
298 and associates it with the tradition.
304 my $t = Text::Tradition->new(
305 'name' => 'simple test',
306 'input' => 'Tabular',
307 'file' => 't/data/simple.txt',
310 is( $t->stemma_count, 0, "No stemmas added yet" );
312 ok( $s = $t->add_stemma( dotfile => 't/data/simple.dot' ), "Added a simple stemma" );
313 is( ref( $s ), 'Text::Tradition::Stemma', "Got a stemma object returned" );
314 is( $t->stemma_count, 1, "Tradition claims to have a stemma" );
315 is( $t->stemma(0), $s, "Tradition hands back the right stemma" );
325 if( $opts{'dotfile'} ) {
326 open $stemma_fh, '<', $opts{'dotfile'}
327 or warn "Could not open file " . $opts{'dotfile'};
328 } elsif( $opts{'dot'} ) {
329 my $str = $opts{'dot'};
330 open $stemma_fh, '<', \$str;
333 binmode $stemma_fh, ':utf8';
334 my $stemma = Text::Tradition::Stemma->new(
335 'collation' => $self->collation,
336 'dot' => $stemma_fh );
337 $self->_add_stemma( $stemma ) if $stemma;
342 __PACKAGE__->meta->make_immutable;
349 =item * Allow tradition to be initialized via passing to a collator.
355 This package is free software and is provided "as is" without express
356 or implied warranty. You can redistribute it and/or modify it under
357 the same terms as Perl itself.
361 Tara L Andrews E<lt>aurum@cpan.orgE<gt>