Commit | Line | Data |
f6066bac |
1 | package Text::Tradition::Parser::TEI; |
2 | |
3 | use strict; |
4 | use warnings; |
00311328 |
5 | use Text::Tradition::Error; |
910a0a6d |
6 | use Text::Tradition::Parser::Util qw( collate_variants ); |
f6066bac |
7 | use XML::LibXML; |
8 | use XML::LibXML::XPathContext; |
9 | |
10 | =head1 NAME |
11 | |
12 | Text::Tradition::Parser::TEI |
13 | |
3b853983 |
14 | =head1 SYNOPSIS |
15 | |
16 | use Text::Tradition; |
17 | |
18 | my $t_from_file = Text::Tradition->new( |
19 | 'name' => 'my text', |
20 | 'input' => 'TEI', |
21 | 'file' => '/path/to/parallel_seg_file.xml' |
22 | ); |
23 | |
24 | my $t_from_string = Text::Tradition->new( |
25 | 'name' => 'my text', |
26 | 'input' => 'TEI', |
27 | 'string' => $parallel_seg_xml, |
28 | ); |
29 | |
30 | |
f6066bac |
31 | =head1 DESCRIPTION |
32 | |
3b853983 |
33 | Parser module for Text::Tradition, given a TEI parallel-segmentation file |
34 | that describes a text and its variants. Normally called upon |
35 | initialization of Text::Tradition. |
36 | |
37 | The witnesses for the tradition are taken from the <listWit/> element |
38 | within the TEI header; the readings are taken from any <p/> element that |
39 | appears in the text body (including <head/> elements therein.) |
f6066bac |
40 | |
41 | =head1 METHODS |
42 | |
e867486f |
43 | =head2 B<parse>( $tradition, $option_hash ) |
3b853983 |
44 | |
45 | Takes an initialized tradition and a set of options; creates the |
46 | appropriate nodes and edges on the graph, as well as the appropriate |
47 | witness objects. The $option_hash must contain either a 'file' or a |
48 | 'string' argument with the XML to be parsed. |
49 | |
50 | =begin testing |
51 | |
52 | use Text::Tradition; |
53 | binmode STDOUT, ":utf8"; |
54 | binmode STDERR, ":utf8"; |
55 | eval { no warnings; binmode $DB::OUT, ":utf8"; }; |
56 | |
57 | my $par_seg = 't/data/florilegium_tei_ps.xml'; |
58 | my $t = Text::Tradition->new( |
59 | 'name' => 'inline', |
60 | 'input' => 'TEI', |
61 | 'file' => $par_seg, |
62 | ); |
f6066bac |
63 | |
3b853983 |
64 | is( ref( $t ), 'Text::Tradition', "Parsed parallel-segmentation TEI" ); |
65 | if( $t ) { |
56eefa04 |
66 | is( scalar $t->collation->readings, 311, "Collation has all readings" ); |
67 | is( scalar $t->collation->paths, 361, "Collation has all paths" ); |
3b853983 |
68 | } |
f6066bac |
69 | |
3b853983 |
70 | =end testing |
f6066bac |
71 | |
72 | =cut |
73 | |
910a0a6d |
74 | my $text = {}; # Hash of arrays, one per eventual witness we find. |
910a0a6d |
75 | my $substitutions = {}; # Keep track of merged readings |
76 | my $app_anchors = {}; # Track apparatus references |
77 | my $app_ac = {}; # Save a.c. readings |
eca16057 |
78 | my $app_count; # Keep track of how many apps we have |
910a0a6d |
79 | |
80 | # Create the package variables for tag names. |
81 | |
82 | # Would really like to do this with varname variables, but apparently this |
83 | # is considered a bad idea. The long way round then. |
84 | my( $LISTWIT, $WITNESS, $TEXT, $W, $SEG, $APP, $RDG, $LEM ) |
85 | = ( 'listWit', 'witness', 'text', 'w', 'seg', 'app', 'rdg', 'lem' ); |
3b853983 |
86 | sub _make_tagnames { |
910a0a6d |
87 | my( $ns ) = @_; |
88 | if( $ns ) { |
89 | $LISTWIT = "$ns:$LISTWIT"; |
90 | $WITNESS = "$ns:$WITNESS"; |
91 | $TEXT = "$ns:$TEXT"; |
92 | $W = "$ns:$W"; |
93 | $SEG = "$ns:$SEG"; |
94 | $APP = "$ns:$APP"; |
95 | $RDG = "$ns:$RDG"; |
96 | $LEM = "$ns:$LEM"; |
97 | } |
98 | } |
99 | |
100 | # Parse the TEI file. |
f6066bac |
101 | sub parse { |
dfc37e38 |
102 | my( $tradition, $opts ) = @_; |
f6066bac |
103 | |
104 | # First, parse the XML. |
105 | my $parser = XML::LibXML->new(); |
dfc37e38 |
106 | my $doc; |
107 | if( exists $opts->{'string'} ) { |
108 | $doc = $parser->parse_string( $opts->{'string'} ); |
109 | } elsif ( exists $opts->{'file'} ) { |
110 | $doc = $parser->parse_file( $opts->{'file'} ); |
111 | } else { |
112 | warn "Could not find string or file option to parse"; |
113 | return; |
114 | } |
f6066bac |
115 | my $tei = $doc->documentElement(); |
00311328 |
116 | unless( $tei->nodeName =~ /^tei(corpus)?$/i ) { |
117 | throw( "Parsed document has non-TEI root element " . $tei->nodeName ); |
118 | } |
f2b9605f |
119 | my $xpc = XML::LibXML::XPathContext->new( $tei ); |
910a0a6d |
120 | my $ns; |
121 | if( $tei->namespaceURI ) { |
122 | $ns = 'tei'; |
123 | $xpc->registerNs( $ns, $tei->namespaceURI ); |
124 | } |
3b853983 |
125 | _make_tagnames( $ns ); |
910a0a6d |
126 | |
f6066bac |
127 | # Then get the witnesses and create the witness objects. |
910a0a6d |
128 | foreach my $wit_el ( $xpc->findnodes( "//$LISTWIT/$WITNESS" ) ) { |
129 | my $sig = $wit_el->getAttribute( 'xml:id' ); |
130 | my $source = $wit_el->toString(); |
82fa4d57 |
131 | $tradition->add_witness( sigil => $sig, sourcetype => 'collation' ); |
f6066bac |
132 | } |
910a0a6d |
133 | map { $text->{$_->sigil} = [] } $tradition->witnesses; |
eca16057 |
134 | |
910a0a6d |
135 | # Look for all word/seg node IDs and note their pre-existence. |
a7fb3133 |
136 | my @attrs = $xpc->findnodes( "//$W/attribute::xml:id" ); |
3b853983 |
137 | _save_preexisting_nodeids( @attrs ); |
910a0a6d |
138 | |
eca16057 |
139 | # Count up how many apps we have. |
140 | my @apps = $xpc->findnodes( "//$APP" ); |
141 | $app_count = scalar( @apps ); |
142 | |
910a0a6d |
143 | # Now go through the children of the text element and pull out the |
144 | # actual text. |
145 | foreach my $xml_el ( $xpc->findnodes( "//$TEXT" ) ) { |
146 | foreach my $xn ( $xml_el->childNodes ) { |
147 | _get_readings( $tradition, $xn ); |
148 | } |
149 | } |
150 | # Our $text global now has lists of readings, one per witness. |
151 | # Join them up. |
152 | my $c = $tradition->collation; |
153 | foreach my $sig ( keys %$text ) { |
910a0a6d |
154 | # Determine the list of readings for |
155 | my $sequence = $text->{$sig}; |
156 | my @real_sequence = ( $c->start ); |
157 | push( @$sequence, $c->end ); |
342c1111 |
158 | foreach( _clean_sequence( $sig, $sequence, 1 ) ) { |
159 | push( @real_sequence, _return_rdg( $_ ) ); |
910a0a6d |
160 | } |
910a0a6d |
161 | # See if we need to make an a.c. version of the witness. |
162 | if( exists $app_ac->{$sig} ) { |
163 | my @uncorrected; |
164 | push( @uncorrected, @real_sequence ); |
342c1111 |
165 | # Get rid of any remaining placeholders. |
166 | @real_sequence = _clean_sequence( $sig, \@uncorrected ); |
167 | # Do the uncorrections |
910a0a6d |
168 | foreach my $app ( keys %{$app_ac->{$sig}} ) { |
169 | my $start = _return_rdg( $app_anchors->{$app}->{$sig}->{'start'} ); |
170 | my $end = _return_rdg( $app_anchors->{$app}->{$sig}->{'end'} ); |
171 | my @new = map { _return_rdg( $_ ) } @{$app_ac->{$sig}->{$app}}; |
172 | _replace_sequence( \@uncorrected, $start, $end, @new ); |
173 | } |
342c1111 |
174 | # and record the results. |
175 | $tradition->witness( $sig )->uncorrected_path( \@uncorrected ); |
1f7aa795 |
176 | $tradition->witness( $sig )->is_layered( 1 ); |
910a0a6d |
177 | } |
342c1111 |
178 | $tradition->witness( $sig )->path( \@real_sequence ); |
910a0a6d |
179 | } |
342c1111 |
180 | # Now make our witness paths. |
181 | $tradition->collation->make_witness_paths(); |
3b853983 |
182 | |
183 | # Calculate the ranks for the nodes. |
a7fb3133 |
184 | $tradition->collation->calculate_ranks(); |
3bc0cd18 |
185 | |
186 | # Now that we have ranks, see if we have distinct nodes with identical |
187 | # text and identical rank that can be merged. |
0e476982 |
188 | $tradition->collation->flatten_ranks(); |
861c3e27 |
189 | |
15db7774 |
190 | # And now that we've done that, calculate the common nodes. |
191 | $tradition->collation->calculate_common_readings(); |
192 | |
861c3e27 |
193 | # Save the text for each witness so that we can ensure consistency |
194 | # later on |
195 | $tradition->collation->text_from_paths(); |
910a0a6d |
196 | } |
197 | |
198 | sub _clean_sequence { |
342c1111 |
199 | my( $wit, $sequence, $keep_ac ) = @_; |
910a0a6d |
200 | my @clean_sequence; |
201 | foreach my $rdg ( @$sequence ) { |
202 | if( $rdg =~ /^PH-(.*)$/ ) { |
342c1111 |
203 | # It is a placeholder. Keep it only if we need it for a later |
204 | # a.c. run. |
910a0a6d |
205 | my $app_id = $1; |
342c1111 |
206 | if( $keep_ac && exists $app_ac->{$wit} && |
78fab1cf |
207 | exists $app_ac->{$wit}->{$app_id} ) { |
342c1111 |
208 | # print STDERR "Retaining empty placeholder for $app_id\n"; |
209 | push( @clean_sequence, $rdg ); |
910a0a6d |
210 | } |
211 | } else { |
212 | push( @clean_sequence, $rdg ); |
213 | } |
f6066bac |
214 | } |
910a0a6d |
215 | return @clean_sequence; |
216 | } |
f6066bac |
217 | |
910a0a6d |
218 | sub _replace_sequence { |
219 | my( $arr, $start, $end, @new ) = @_; |
220 | my( $start_idx, $end_idx ); |
221 | foreach my $i ( 0 .. $#{$arr} ) { |
342c1111 |
222 | # If $arr->[$i] is a placeholder, cope. |
223 | my $iid = ref( $arr->[$i] ) ? $arr->[$i]->id : $arr->[$i]; |
224 | $start_idx = $i if( $iid eq $start ); |
225 | if( $iid eq $end ) { |
910a0a6d |
226 | $end_idx = $i; |
227 | last; |
228 | } |
229 | } |
230 | unless( $start_idx && $end_idx ) { |
231 | warn "Could not find start and end"; |
232 | return; |
f2b9605f |
233 | } |
910a0a6d |
234 | my $length = $end_idx - $start_idx + 1; |
235 | splice( @$arr, $start_idx, $length, @new ); |
236 | } |
f6066bac |
237 | |
910a0a6d |
238 | sub _return_rdg { |
239 | my( $rdg ) = @_; |
240 | # If we were passed a reading name, return the name. If we were |
241 | # passed a reading object, return the object. |
242 | my $wantobj = ref( $rdg ) eq 'Text::Tradition::Collation::Reading'; |
243 | my $real = $rdg; |
e4b0f464 |
244 | if( exists $substitutions->{ $wantobj ? $rdg->id : $rdg } ) { |
245 | $real = $substitutions->{ $wantobj ? $rdg->id : $rdg }; |
246 | $real = $real->id unless $wantobj; |
910a0a6d |
247 | } |
248 | return $real; |
f6066bac |
249 | } |
250 | |
3b853983 |
251 | ## TODO test specific sorts of nodes of the parallel-seg XML. |
252 | |
910a0a6d |
253 | ## Recursive helper function to help us navigate through nested XML, |
254 | ## picking out the text. $tradition is the tradition, needed for |
255 | ## making readings; $xn is the XML node currently being looked at, |
256 | ## $in_var is a flag to say that we are inside a variant, $ac is a |
257 | ## flag to say that we are inside an ante-correctionem reading, and |
258 | ## @cur_wits is the list of witnesses to which this XML node applies. |
259 | ## Returns the list of readings, if any, created on the run. |
260 | |
261 | { |
3bc0cd18 |
262 | my %active_wits; |
910a0a6d |
263 | my $current_app; |
eca16057 |
264 | my $seen_apps; |
910a0a6d |
265 | |
266 | sub _get_readings { |
267 | my( $tradition, $xn, $in_var, $ac, @cur_wits ) = @_; |
3bc0cd18 |
268 | @cur_wits = grep { $active_wits{$_} } keys %active_wits unless $in_var; |
910a0a6d |
269 | |
270 | my @new_readings; |
271 | if( $xn->nodeType == XML_TEXT_NODE ) { |
272 | # Some words, thus make some readings. |
273 | my $str = $xn->data; |
274 | return unless $str =~ /\S/; # skip whitespace-only text nodes |
275 | #print STDERR "Handling text node " . $str . "\n"; |
276 | # Check that all the witnesses we have are active. |
277 | foreach my $c ( @cur_wits ) { |
3bc0cd18 |
278 | warn "$c is not among active wits" unless $active_wits{$c}; |
910a0a6d |
279 | } |
280 | $str =~ s/^\s+//; |
281 | my $final = $str =~ s/\s+$//; |
282 | foreach my $w ( split( /\s+/, $str ) ) { |
283 | # For now, skip punctuation. |
284 | next if $w !~ /[[:alnum:]]/; |
3b853983 |
285 | my $rdg = _make_reading( $tradition->collation, $w ); |
910a0a6d |
286 | push( @new_readings, $rdg ); |
910a0a6d |
287 | foreach ( @cur_wits ) { |
288 | warn "Empty wit!" unless $_; |
289 | warn "Empty reading!" unless $rdg; |
290 | push( @{$text->{$_}}, $rdg ) unless $ac; |
291 | } |
292 | } |
293 | } elsif( $xn->nodeName eq 'w' ) { |
294 | # Everything in this tag is one word. Also save any original XML ID. |
295 | #print STDERR "Handling word " . $xn->toString . "\n"; |
296 | # Check that all the witnesses we have are active. |
297 | foreach my $c ( @cur_wits ) { |
3bc0cd18 |
298 | warn "$c is not among active wits" unless $active_wits{$c}; |
910a0a6d |
299 | } |
300 | my $xml_id = $xn->getAttribute( 'xml:id' ); |
3b853983 |
301 | my $rdg = _make_reading( $tradition->collation, $xn->textContent, $xml_id ); |
910a0a6d |
302 | push( @new_readings, $rdg ); |
910a0a6d |
303 | foreach( @cur_wits ) { |
304 | warn "Empty wit!" unless $_; |
305 | warn "Empty reading!" unless $rdg; |
306 | push( @{$text->{$_}}, $rdg ) unless $ac; |
307 | } |
308 | } elsif ( $xn->nodeName eq 'app' ) { |
eca16057 |
309 | $seen_apps++; |
910a0a6d |
310 | $current_app = $xn->getAttribute( 'xml:id' ); |
311 | # print STDERR "Handling app $current_app\n"; |
312 | # Keep the reading sets in this app. |
313 | my @sets; |
314 | # Recurse through all children (i.e. rdgs) for sets of words. |
315 | foreach ( $xn->childNodes ) { |
316 | my @rdg_set = _get_readings( $tradition, $_, $in_var, $ac, @cur_wits ); |
317 | push( @sets, \@rdg_set ) if @rdg_set; |
318 | } |
319 | # Now collate these sets if we have more than one. |
320 | my $subs = collate_variants( $tradition->collation, @sets ) if @sets > 1; |
321 | map { $substitutions->{$_} = $subs->{$_} } keys %$subs; |
910a0a6d |
322 | # Return the entire set of unique readings. |
323 | my %unique; |
324 | foreach my $s ( @sets ) { |
e4b0f464 |
325 | map { $unique{$_->id} = $_ } @$s; |
910a0a6d |
326 | } |
327 | push( @new_readings, values( %unique ) ); |
328 | # Exit the current app. |
329 | $current_app = ''; |
330 | } elsif ( $xn->nodeName eq 'lem' || $xn->nodeName eq 'rdg' ) { |
331 | # Alter the current witnesses and recurse. |
332 | #print STDERR "Handling reading for " . $xn->getAttribute( 'wit' ) . "\n"; |
3bc0cd18 |
333 | # TODO handle p.c. and s.l. designations too |
910a0a6d |
334 | $ac = $xn->getAttribute( 'type' ) && $xn->getAttribute( 'type' ) eq 'a.c.'; |
3b853983 |
335 | my @rdg_wits = _get_sigla( $xn ); |
a7fb3133 |
336 | return unless @rdg_wits; # Skip readings that appear in no witnesses |
910a0a6d |
337 | my @words; |
338 | foreach ( $xn->childNodes ) { |
339 | my @rdg_set = _get_readings( $tradition, $_, 1, $ac, @rdg_wits ); |
340 | push( @words, @rdg_set ) if @rdg_set; |
341 | } |
342 | # If we have more than one word in a reading, it should become a segment. |
343 | # $tradition->collation->add_segment( @words ) if @words > 1; |
344 | |
345 | if( $ac ) { |
346 | # Add the reading set to the a.c. readings. |
347 | foreach ( @rdg_wits ) { |
348 | $app_ac->{$_}->{$current_app} = \@words; |
349 | } |
350 | } else { |
351 | # Add the reading set to the app anchors for each witness |
352 | # or put in placeholders for empty p.c. readings |
353 | foreach ( @rdg_wits ) { |
e4b0f464 |
354 | my $start = @words ? $words[0]->id : "PH-$current_app"; |
355 | my $end = @words ? $words[-1]->id : "PH-$current_app"; |
910a0a6d |
356 | $app_anchors->{$current_app}->{$_}->{'start'} = $start; |
357 | $app_anchors->{$current_app}->{$_}->{'end'} = $end; |
358 | push( @{$text->{$_}}, $start ) unless @words; |
359 | } |
360 | } |
361 | push( @new_readings, @words ); |
362 | } elsif( $xn->nodeName eq 'witStart' ) { |
363 | # Add the relevant wit(s) to the active list. |
364 | #print STDERR "Handling witStart\n"; |
3bc0cd18 |
365 | map { $active_wits{$_} = 1 } @cur_wits; |
366 | # Record a lacuna in all non-active witnesses if this is |
367 | # the first app. Get the full list from $text. |
368 | if( $seen_apps == 1 ) { |
369 | my $i = 0; |
370 | foreach my $sig ( keys %$text ) { |
371 | next if $active_wits{$sig}; |
e4b0f464 |
372 | my $l = $tradition->collation->add_reading( { |
e4b0f464 |
373 | 'id' => $current_app . "_$i", |
374 | 'is_lacuna' => 1 } ); |
3bc0cd18 |
375 | $i++; |
376 | push( @{$text->{$sig}}, $l ); |
377 | } |
378 | } |
910a0a6d |
379 | } elsif( $xn->nodeName eq 'witEnd' ) { |
380 | # Take the relevant wit(s) out of the list. |
381 | #print STDERR "Handling witEnd\n"; |
3bc0cd18 |
382 | map { $active_wits{$_} = undef } @cur_wits; |
eca16057 |
383 | # Record a lacuna, unless this is the last app. |
384 | unless( $seen_apps == $app_count ) { |
385 | foreach my $i ( 0 .. $#cur_wits ) { |
386 | my $w = $cur_wits[$i]; |
e4b0f464 |
387 | my $l = $tradition->collation->add_reading( { |
e4b0f464 |
388 | 'id' => $current_app . "_$i", |
389 | 'is_lacuna' => 1 } ); |
eca16057 |
390 | push( @{$text->{$w}}, $l ); |
391 | } |
392 | } |
a7fb3133 |
393 | } elsif( $xn->nodeName eq 'witDetail' |
394 | || $xn->nodeName eq 'note' ) { |
910a0a6d |
395 | # Ignore these for now. |
396 | return; |
397 | } else { |
398 | # Recurse as if this tag weren't there. |
399 | #print STDERR "Recursing on tag " . $xn->nodeName . "\n"; |
400 | foreach( $xn->childNodes ) { |
401 | push( @new_readings, _get_readings( $tradition, $_, $in_var, $ac, @cur_wits ) ); |
402 | } |
403 | } |
404 | return @new_readings; |
405 | } |
406 | |
407 | } |
408 | |
3b853983 |
409 | =begin testing |
410 | |
411 | use XML::LibXML; |
412 | use XML::LibXML::XPathContext; |
413 | use Text::Tradition::Parser::TEI; |
414 | |
415 | my $xml_str = '<tei><rdg wit="#A #B #C #D">some text</rdg></tei>'; |
416 | my $el = XML::LibXML->new()->parse_string( $xml_str )->documentElement; |
417 | my $xpc = XML::LibXML::XPathContext->new( $el ); |
418 | my $obj = $xpc->find( '//rdg' ); |
419 | |
420 | my @wits = Text::Tradition::Parser::TEI::_get_sigla( $obj ); |
421 | is( join( ' ', @wits) , "A B C D", "correctly parsed reading wit string" ); |
422 | |
423 | =end testing |
424 | |
425 | =cut |
426 | |
910a0a6d |
427 | # Helper to extract a list of witness sigla from a reading element. |
3b853983 |
428 | sub _get_sigla { |
f6066bac |
429 | my( $rdg ) = @_; |
430 | # Cope if we have been handed a NodeList. There is only |
431 | # one reading here. |
432 | if( ref( $rdg ) eq 'XML::LibXML::NodeList' ) { |
910a0a6d |
433 | $rdg = $rdg->shift; |
f6066bac |
434 | } |
435 | |
436 | my @wits; |
437 | if( ref( $rdg ) eq 'XML::LibXML::Element' ) { |
910a0a6d |
438 | my $witstr = $rdg->getAttribute( 'wit' ); |
a7fb3133 |
439 | return () unless $witstr; |
910a0a6d |
440 | $witstr =~ s/^\s+//; |
441 | $witstr =~ s/\s+$//; |
442 | @wits = split( /\s+/, $witstr ); |
443 | map { $_ =~ s/^\#// } @wits; |
f6066bac |
444 | } |
445 | return @wits; |
446 | } |
447 | |
910a0a6d |
448 | # Helper with its counters to actually make the readings. |
f2b9605f |
449 | { |
450 | my $word_ctr = 0; |
451 | my %used_nodeids; |
452 | |
3b853983 |
453 | sub _save_preexisting_nodeids { |
910a0a6d |
454 | foreach( @_ ) { |
455 | $used_nodeids{$_->getValue()} = 1; |
456 | } |
457 | } |
458 | |
3b853983 |
459 | sub _make_reading { |
910a0a6d |
460 | my( $graph, $word, $xml_id ) = @_; |
461 | if( $xml_id ) { |
462 | if( exists $used_nodeids{$xml_id} ) { |
463 | if( $used_nodeids{$xml_id} != 1 ) { |
464 | warn "Already used assigned XML ID somewhere else!"; |
465 | $xml_id = undef; |
466 | } |
467 | } else { |
468 | warn "Undetected pre-existing XML ID"; |
469 | } |
470 | } |
471 | if( !$xml_id ) { |
472 | until( $xml_id ) { |
473 | my $try_id = 'w'.$word_ctr++; |
474 | next if exists $used_nodeids{$try_id}; |
475 | $xml_id = $try_id; |
476 | } |
477 | } |
e4b0f464 |
478 | my $rdg = $graph->add_reading( |
49d4f2ac |
479 | { 'id' => $xml_id, |
e4b0f464 |
480 | 'text' => $word } |
481 | ); |
910a0a6d |
482 | $used_nodeids{$xml_id} = $rdg; |
483 | return $rdg; |
f2b9605f |
484 | } |
485 | } |
486 | |
f6066bac |
487 | 1; |
3b853983 |
488 | |
00311328 |
489 | sub throw { |
490 | Text::Tradition::Error->throw( |
491 | 'ident' => 'Parser::TEI error', |
492 | 'message' => $_[0], |
493 | ); |
494 | } |
495 | |
3b853983 |
496 | =head1 BUGS / TODO |
497 | |
498 | =over |
499 | |
500 | =item * More unit testing |
501 | |
e867486f |
502 | =item * Handle special designations apart from a.c. |
503 | |
504 | =item * Mark common nodes within collated variants |
505 | |
3b853983 |
506 | =back |
507 | |
508 | =head1 LICENSE |
509 | |
510 | This package is free software and is provided "as is" without express |
511 | or implied warranty. You can redistribute it and/or modify it under |
512 | the same terms as Perl itself. |
513 | |
514 | =head1 AUTHOR |
515 | |
516 | Tara L Andrews E<lt>aurum@cpan.orgE<gt> |