1 // Copyright 2011 Google Inc. All Rights Reserved.
3 // Licensed under the Apache License, Version 2.0 (the "License");
4 // you may not use this file except in compliance with the License.
5 // You may obtain a copy of the License at
7 // http://www.apache.org/licenses/LICENSE-2.0
9 // Unless required by applicable law or agreed to in writing, software
10 // distributed under the License is distributed on an "AS IS" BASIS,
11 // WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
12 // See the License for the specific language governing permissions and
13 // limitations under the License.
15 package com.hughes.android.dictionary.parser;
17 import java.io.BufferedInputStream;
18 import java.io.DataInputStream;
19 import java.io.EOFException;
21 import java.io.FileInputStream;
22 import java.io.IOException;
23 import java.util.ArrayList;
24 import java.util.Arrays;
25 import java.util.Collection;
26 import java.util.Collections;
27 import java.util.LinkedHashSet;
28 import java.util.List;
31 import java.util.regex.Pattern;
33 import com.hughes.android.dictionary.engine.EntryTypeName;
34 import com.hughes.android.dictionary.engine.IndexBuilder;
35 import com.hughes.android.dictionary.engine.IndexedEntry;
36 import com.hughes.android.dictionary.engine.PairEntry;
37 import com.hughes.android.dictionary.engine.PairEntry.Pair;
39 public class EnWiktionaryXmlParser {
41 // TODO: look for {{ and [[ and <adf> <!-- in output.
42 // TODO: process {{ttbc}} lines
44 static final Pattern partOfSpeechHeader = Pattern.compile(
45 "Noun|Verb|Adjective|Adverb|Pronoun|Conjunction|Interjection|" +
46 "Preposition|Proper noun|Article|Prepositional phrase|Acronym|" +
47 "Abbreviation|Initialism|Contraction|Prefix|Suffix|Symbol|Letter|" +
48 "Ligature|Idiom|Phrase|" +
49 // These are @deprecated:
50 "Noun form|Verb form|Adjective form|Nominal phrase|Noun phrase|" +
51 "Verb phrase|Transitive verb|Intransitive verb|Reflexive verb|" +
52 // These are extras I found:
53 "Determiner|Numeral|Number|Cardinal number|Ordinal number|Proverb|" +
54 "Particle|Interjection|Pronominal adverb" +
55 "Han character|Hanzi|Hanja|Kanji|Katakana character|Syllable");
57 final IndexBuilder enIndexBuilder;
58 final IndexBuilder otherIndexBuilder;
59 final Pattern langPattern;
60 final Pattern langCodePattern;
63 public EnWiktionaryXmlParser(final IndexBuilder enIndexBuilder, final IndexBuilder otherIndexBuilder, final Pattern langPattern, final Pattern langCodePattern, final boolean swap) {
64 this.enIndexBuilder = enIndexBuilder;
65 this.otherIndexBuilder = otherIndexBuilder;
66 this.langPattern = langPattern;
67 this.langCodePattern = langCodePattern;
72 public void parse(final File file, final int pageLimit) throws IOException {
74 final DataInputStream dis = new DataInputStream(new BufferedInputStream(new FileInputStream(file)));
76 if (pageLimit >= 0 && pageCount >= pageLimit) {
82 title = dis.readUTF();
83 } catch (EOFException e) {
87 final String heading = dis.readUTF();
88 final int bytesLength = dis.readInt();
89 final byte[] bytes = new byte[bytesLength];
91 final String text = new String(bytes, "UTF8");
93 parseSection(title, heading, text);
96 if (pageCount % 1000 == 0) {
97 System.out.println("pageCount=" + pageCount);
102 private void parseSection(final String title, String heading, final String text) {
103 if (title.startsWith("Wiktionary:") ||
104 title.startsWith("Template:") ||
105 title.startsWith("Appendix:") ||
106 title.startsWith("Category:") ||
107 title.startsWith("Index:") ||
108 title.startsWith("MediaWiki:") ||
109 title.startsWith("TransWiki:") ||
110 title.startsWith("Citations:") ||
111 title.startsWith("Concordance:") ||
112 title.startsWith("Help:")) {
116 heading = heading.replaceAll("=", "").trim();
117 if (heading.equals("English")) {
118 doEnglishWord(title, text);
119 } else if (langPattern.matcher(heading).matches()){
120 doForeignWord(title, text);
125 // -------------------------------------------------------------------------
130 private void doEnglishWord(String title, String text) {
131 final WikiTokenizer wikiTokenizer = new WikiTokenizer(text);
132 while (wikiTokenizer.nextToken() != null) {
134 if (wikiTokenizer.isHeading()) {
135 final String headerName = wikiTokenizer.headingWikiText();
137 if (wikiTokenizer.headingDepth() <= posDepth) {
142 if (partOfSpeechHeader.matcher(headerName).matches()) {
143 posDepth = wikiTokenizer.headingDepth();
144 pos = wikiTokenizer.headingWikiText();
145 } else if (headerName.equals("Translations")) {
146 doTranslations(title, wikiTokenizer);
147 } else if (headerName.equals("Pronunciation")) {
148 //doPronunciation(wikiLineReader);
155 private static Set<String> encodings = new LinkedHashSet<String>(Arrays.asList("zh-ts",
156 "sd-Arab", "ku-Arab", "Arab", "unicode", "Laoo", "ur-Arab", "Thai",
157 "fa-Arab", "Khmr", "zh-tsp", "Cyrl", "IPAchar", "ug-Arab", "ko-inline",
158 "Jpan", "Kore", "Hebr", "rfscript", "Beng", "Mong", "Knda", "Cyrs",
159 "yue-tsj", "Mlym", "Tfng", "Grek", "yue-yue-j"));
161 private void doTranslations(final String title, final WikiTokenizer wikiTokenizer) {
163 boolean done = false;
164 while (wikiTokenizer.nextToken() != null) {
165 if (wikiTokenizer.isHeading()) {
166 wikiTokenizer.returnToLineStart();
173 // Check whether we care about this line:
175 //line = WikiLineReader.removeSquareBrackets(line);
177 if (wikiTokenizer.isFunction()) {
178 final String functionName = wikiTokenizer.functionName();
179 final List<String> positionArgs = wikiTokenizer.functionPositionArgs();
181 if (functionName.equals("trans-top")) {
183 if (wikiTokenizer.functionPositionArgs().size() >= 1) {
184 sense = positionArgs.get(0);
185 // TODO: could emphasize words in [[brackets]] inside sense.
186 sense = WikiTokenizer.toPlainText(sense);
187 //System.out.println("Sense: " + sense);
189 } else if (functionName.equals("trans-bottom")) {
191 } else if (functionName.equals("trans-mid")) {
192 } else if (functionName.equals("trans-see")) {
193 } else if (functionName.startsWith("checktrans")) {
194 //TODO: Check this: done = true;
196 System.err.println("Unexpected translation wikifunction: " + wikiTokenizer.token() + ", title=" + title);
198 } else if (wikiTokenizer.isListItem() && wikiTokenizer.listItemPrefix().startsWith("*")) {
199 final String line = wikiTokenizer.listItemWikiText();
200 // This line could produce an output...
202 // First strip the language and check whether it matches.
203 // And hold onto it for sub-lines.
204 final int colonIndex = line.indexOf(":");
205 if (colonIndex == -1) {
209 final String lang = line.substring(0, colonIndex);
210 if (!this.langPattern.matcher(lang).find()) {
214 String rest = line.substring(colonIndex + 1).trim();
215 doTranslationLine(line, title, sense, rest);
217 } else if (wikiTokenizer.remainderStartsWith("''See''")) {
218 wikiTokenizer.nextLine();
219 System.out.println("Skipping line: " + wikiTokenizer.token());
220 } else if (wikiTokenizer.isWikiLink()) {
221 final String wikiLink = wikiTokenizer.wikiLinkText();
222 if (wikiLink.contains(":") && wikiLink.contains(title)) {
223 } else if (wikiLink.contains("Category:")) {
225 System.err.println("Unexpected wikiLink: " + wikiTokenizer.token() + ", title=" + title);
227 } else if (wikiTokenizer.isNewline() || wikiTokenizer.isMarkup() || wikiTokenizer.isComment()) {
229 final String token = wikiTokenizer.token();
230 if (token.equals("----")) {
232 System.err.println("Unexpected translation token: " + wikiTokenizer.token() + ", title=" + title);
239 private static <T> T get(final List<T> list, final int index) {
240 return index < list.size() ? list.get(index) : null;
243 private void doTranslationLine(final String line, final String title, final String sense, final String rest) {
244 // Good chance we'll actually file this one...
245 final PairEntry pairEntry = new PairEntry();
246 final IndexedEntry indexedEntry = new IndexedEntry(pairEntry);
248 final StringBuilder otherText = new StringBuilder();
249 final WikiTokenizer wikiTokenizer = new WikiTokenizer(rest);
250 while (wikiTokenizer.nextToken() != null) {
252 if (wikiTokenizer.isPlainText()) {
253 final String plainText = wikiTokenizer.token();
254 otherText.append("").append(plainText);
255 otherIndexBuilder.addEntryWithString(indexedEntry, plainText, EntryTypeName.WIKTIONARY_TRANSLATION_OTHER_TEXT);
257 } else if (wikiTokenizer.isWikiLink()) {
258 final String plainText = wikiTokenizer.wikiLinkText();
259 otherText.append("").append(plainText);
260 otherIndexBuilder.addEntryWithString(indexedEntry, plainText, EntryTypeName.WIKTIONARY_TRANSLATION_WIKI_TEXT);
262 } else if (wikiTokenizer.isFunction()) {
263 final String functionName = wikiTokenizer.functionName();
264 final List<String> args = wikiTokenizer.functionPositionArgs();
265 final Map<String,String> namedArgs = wikiTokenizer.functionNamedArgs();
267 if (functionName.equals("t") || functionName.equals("t+") || functionName.equals("t-") || functionName.equals("tø")) {
268 if (args.size() < 2) {
269 System.err.println("{{t}} with too few args: " + line + ", title=" + title);
272 final String langCode = get(args, 0);
273 if (this.langCodePattern.matcher(langCode).matches()) {
274 final String word = get(args, 1);
275 final String gender = get(args, 2);
276 final String transliteration = namedArgs.get("tr");
277 if (otherText.length() > 0) {
278 otherText.append("");
280 otherText.append(word);
281 otherIndexBuilder.addEntryWithString(indexedEntry, word, EntryTypeName.WIKTIONARY_TITLE_SINGLE, EntryTypeName.WIKTIONARY_TITLE_MULTI);
282 if (gender != null) {
283 otherText.append(String.format(" {%s}", gender));
285 if (transliteration != null) {
286 otherText.append(String.format(" (tr. %s)", transliteration));
287 otherIndexBuilder.addEntryWithString(indexedEntry, transliteration, EntryTypeName.WIKTIONARY_TRANSLITERATION);
290 } else if (functionName.equals("qualifier")) {
291 String qualifier = args.get(0);
292 if (!namedArgs.isEmpty() || args.size() > 1) {
293 System.err.println("weird qualifier: " + line);
295 otherText.append("(").append(qualifier).append(")");
296 } else if (encodings.contains(functionName)) {
297 otherText.append("").append(args.get(0));
298 otherIndexBuilder.addEntryWithString(indexedEntry, args.get(0), EntryTypeName.WIKTIONARY_TRANSLATION_OTHER_TEXT);
299 } else if (functionName.equals("m") || functionName.equals("f") || functionName.equals("n") || functionName.equals("p")) {
300 otherText.append("{");
301 otherText.append(functionName);
302 for (int i = 0; i < args.size(); ++i) {
303 otherText.append("|").append(args.get(i));
305 otherText.append("}");
306 } else if (functionName.equals("g")) {
307 otherText.append("{g}");
308 } else if (functionName.equals("l")) {
309 // encodes text in various langs.
311 otherText.append("").append(args.get(1));
312 otherIndexBuilder.addEntryWithString(indexedEntry, args.get(1), EntryTypeName.WIKTIONARY_TRANSLATION_OTHER_TEXT);
313 // TODO: transliteration
314 } else if (functionName.equals("term")) {
315 // cross-reference to another dictionary
316 otherText.append("").append(args.get(0));
317 otherIndexBuilder.addEntryWithString(indexedEntry, args.get(0), EntryTypeName.WIKTIONARY_TRANSLATION_OTHER_TEXT);
318 // TODO: transliteration
319 } else if (functionName.equals("italbrac") || functionName.equals("gloss")) {
320 // TODO: put this text aside to use it.
321 otherText.append("[").append(args.get(0)).append("]");
322 otherIndexBuilder.addEntryWithString(indexedEntry, args.get(0), EntryTypeName.WIKTIONARY_TRANSLATION_OTHER_TEXT);
323 } else if (functionName.equals("ttbc")) {
324 } else if (functionName.equals("trreq")) {
325 } else if (functionName.equals("not used")) {
326 otherText.append("(not used)");
327 } else if (functionName.equals("t-image")) {
328 // American sign language
329 } else if (args.isEmpty() && namedArgs.isEmpty()) {
330 otherText.append("{UNK. FUNC.: ").append(functionName).append("}");
332 System.err.println("Unexpected t+- wikifunction: " + line + ", title=" + title);
335 } else if (wikiTokenizer.isNewline()) {
337 } else if (wikiTokenizer.isComment()) {
338 } else if (wikiTokenizer.isMarkup()) {
340 System.err.println("Bad translation token: " + wikiTokenizer.token());
345 StringBuilder englishText = new StringBuilder();
347 englishText.append(title);
349 englishText.append(" (").append(sense).append(")");
350 enIndexBuilder.addEntryWithString(indexedEntry, sense, EntryTypeName.WIKTIONARY_TRANSLATION_SENSE, EntryTypeName.WIKTIONARY_TRANSLATION_SENSE);
353 englishText.append(" (").append(pos.toLowerCase()).append(")");
355 enIndexBuilder.addEntryWithString(indexedEntry, title, EntryTypeName.WIKTIONARY_TITLE_SINGLE, EntryTypeName.WIKTIONARY_TITLE_MULTI);
357 final Pair pair = new Pair(trim(englishText.toString()), trim(otherText.toString()), swap);
358 pairEntry.pairs.add(pair);
359 assert (pairsAdded.add(pair.toString()));
360 if (pair.toString().equals("libero {m} :: free (adjective)")) {
361 System.out.println();
366 static final Pattern whitespace = Pattern.compile("\\s+");
368 static String trim(final String s) {
369 return whitespace.matcher(s).replaceAll(" ").trim();
372 Set<String> pairsAdded = new LinkedHashSet<String>();
374 // -------------------------------------------------------------------------
376 private void doForeignWord(final String title, final String text) {
377 final WikiTokenizer wikiTokenizer = new WikiTokenizer(text);
378 while (wikiTokenizer.nextToken() != null) {
379 if (wikiTokenizer.isHeading()) {
380 final String headingName = wikiTokenizer.headingWikiText();
381 if (headingName.equals("Translations")) {
382 System.err.println("Translations not in English section: " + title);
383 } else if (headingName.equals("Pronunciation")) {
384 //doPronunciation(wikiLineReader);
385 } else if (partOfSpeechHeader.matcher(headingName).matches()) {
386 doPartOfSpeech(title, headingName, wikiTokenizer.headingDepth(), wikiTokenizer);
394 private void doPartOfSpeech(String title, final String posHeading, final int posDepth, WikiTokenizer wikiTokenizer) {
395 System.out.println("***" + title);
396 System.out.println(posHeading);
397 //final StringBuilder foreignBuilder = new StringBuilder();
400 Collection<String> forms = Collections.emptyList();
402 int currentHeadingDepth = posDepth;
403 while (wikiTokenizer.nextToken() != null) {
404 if (wikiTokenizer.isHeading()) {
405 currentHeadingDepth = wikiTokenizer.headingDepth();
407 if (currentHeadingDepth <= posDepth) {
408 wikiTokenizer.returnToLineStart();
413 if (currentHeadingDepth > posDepth) {
418 if (wikiTokenizer.isFunction()) {
419 final String name = wikiTokenizer.functionName();
420 final List<String> args = wikiTokenizer.functionPositionArgs();
421 final Map<String,String> namedArgs = wikiTokenizer.functionNamedArgs();
422 // First line is generally a repeat of the title with some extra information.
423 // We need to build up the left side (foreign text, tokens) separately from the
424 // right side (English). The left-side may get paired with multiple right sides.
425 // The left side should get filed under every form of the word in question (singular, plural).
427 // For verbs, the conjugation comes later on in a deeper section.
428 // Ideally, we'd want to file every English entry with the verb
429 // under every verb form coming from the conjugation.
430 // Ie. under "fa": see: "make :: fare" and "do :: fare"
431 // But then where should we put the conjugation table?
432 // I think just under fare. But then we need a way to link to the entry (actually the row, since entries doesn't show up!)
433 // for the conjugation table from "fa".
434 // Would like to be able to link to a lang#token.
435 if (name.equals("it-noun")) {
436 assert forms.isEmpty();
437 final String base = get(args, 0);
438 final String gender = get(args, 1);
439 final String singular = base + get(args, 2);
440 final String plural = base + get(args, 3);
441 side = String.format("%s {%s}, %s {pl}", singular, gender, plural, plural);
442 forms = Arrays.asList(singular, plural);
443 } else if (name.equals("it-proper noun")) {
445 } else if (name.equals("it-adj")) {
447 } else if (name.startsWith("it-conj")) {
448 if (name.equals("it-conj-are")) {
449 itConjAre(args, namedArgs);
450 } else if (name.equals("it-conj-ere")) {
451 } else if (name.equals("it-conj-ire")) {
453 System.err.println("Unknown conjugation: " + wikiTokenizer.token());
457 System.err.println("Unknown function: " + wikiTokenizer.token());
460 } else if (wikiTokenizer.isListItem()) {
461 handleForeignListItem(side != null ? side : title, title, forms, wikiTokenizer);
463 } else if (wikiTokenizer.isWikiLink()) {
471 private void itConjAre(List<String> args, Map<String, String> namedArgs) {
472 final String base = args.get(0);
473 final String aux = args.get(1);
475 putIfMissing(namedArgs, "inf", base + "are");
476 putIfMissing(namedArgs, "aux", aux);
477 putIfMissing(namedArgs, "ger", base + "ando");
478 putIfMissing(namedArgs, "presp", base + "ante");
479 putIfMissing(namedArgs, "pastp", base + "ato");
481 putIfMissing(namedArgs, "pres1s", base + "o");
482 putIfMissing(namedArgs, "pres2s", base + "i");
483 putIfMissing(namedArgs, "pres3s", base + "a");
484 putIfMissing(namedArgs, "pres1p", base + "iamo");
485 putIfMissing(namedArgs, "pres2p", base + "ate");
486 putIfMissing(namedArgs, "pres3p", base + "ano");
488 putIfMissing(namedArgs, "imperf1s", base + "avo");
489 putIfMissing(namedArgs, "imperf2s", base + "avi");
490 putIfMissing(namedArgs, "imperf3s", base + "ava");
491 putIfMissing(namedArgs, "imperf1p", base + "avamo");
492 putIfMissing(namedArgs, "imperf2p", base + "avate");
493 putIfMissing(namedArgs, "imperf3p", base + "avano");
495 putIfMissing(namedArgs, "prem1s", base + "ai");
496 putIfMissing(namedArgs, "prem2s", base + "asti");
497 putIfMissing(namedArgs, "prem3s", base + "ò");
498 putIfMissing(namedArgs, "prem1p", base + "ammo");
499 putIfMissing(namedArgs, "prem2p", base + "aste");
500 putIfMissing(namedArgs, "prem3p", base + "arono");
502 putIfMissing(namedArgs, "fut1s", base + "erò");
503 putIfMissing(namedArgs, "fut2s", base + "erai");
504 putIfMissing(namedArgs, "fut3s", base + "erà");
505 putIfMissing(namedArgs, "fut1p", base + "eremo");
506 putIfMissing(namedArgs, "fut2p", base + "erete");
507 putIfMissing(namedArgs, "fut3p", base + "eranno");
509 putIfMissing(namedArgs, "cond1s", base + "erei");
510 putIfMissing(namedArgs, "cond2s", base + "eresti");
511 putIfMissing(namedArgs, "cond3s", base + "erebbe");
512 putIfMissing(namedArgs, "cond1p", base + "eremmo");
513 putIfMissing(namedArgs, "cond2p", base + "ereste");
514 putIfMissing(namedArgs, "cond3p", base + "erebbero");
515 // Subjunctive / congiuntivo
516 putIfMissing(namedArgs, "sub123s", base + "i");
517 putIfMissing(namedArgs, "sub1p", base + "iamo");
518 putIfMissing(namedArgs, "sub2p", base + "iate");
519 putIfMissing(namedArgs, "sub3p", base + "ino");
520 // Imperfect subjunctive
521 putIfMissing(namedArgs, "impsub12s", base + "assi");
522 putIfMissing(namedArgs, "impsub3s", base + "asse");
523 putIfMissing(namedArgs, "impsub1p", base + "assimo");
524 putIfMissing(namedArgs, "impsub2p", base + "aste");
525 putIfMissing(namedArgs, "impsub3p", base + "assero");
527 putIfMissing(namedArgs, "imp2s", base + "a");
528 putIfMissing(namedArgs, "imp3s", base + "i");
529 putIfMissing(namedArgs, "imp1p", base + "iamo");
530 putIfMissing(namedArgs, "imp2p", base + "ate");
531 putIfMissing(namedArgs, "imp3p", base + "ino");
534 itConj(args, namedArgs);
538 private void itConj(List<String> args, Map<String, String> namedArgs) {
539 // TODO Auto-generated method stub
544 private static void putIfMissing(final Map<String, String> namedArgs, final String key,
545 final String value) {
546 final String oldValue = namedArgs.get(key);
547 if (oldValue == null || oldValue.length() == 0) {
548 namedArgs.put(key, value);
552 // TODO: check how ='' and =| are manifested....
553 // TODO: get this right in -are
554 private static void putOrNullify(final Map<String, String> namedArgs, final String key,
555 final String value) {
556 final String oldValue = namedArgs.get(key);
557 if (oldValue == null/* || oldValue.length() == 0*/) {
558 namedArgs.put(key, value);
560 if (oldValue.equals("''")) {
561 namedArgs.put(key, "");
566 final List<String> listPrefixes = new ArrayList<String>();
567 final List<String> listLines = new ArrayList<String>();
569 static final Pattern UNINDEXED_WIKI_TEXT = Pattern.compile(
570 "(first|second|third)-person (singular|plural)|" +
575 private void handleForeignListItem(final String foreignText, String title, final Collection<String> forms, final WikiTokenizer wikiTokenizer) {
577 final String prefix = wikiTokenizer.listItemPrefix();
578 if (prefix.length() > 1) {
579 System.err.println("Prefix too long: " + wikiTokenizer.token());
583 listPrefixes.clear();
585 listPrefixes.add(prefix);
586 listLines.add(wikiTokenizer.listItemWikiText());
587 while(wikiTokenizer.nextToken() != null &&
588 wikiTokenizer.isNewline() ||
589 wikiTokenizer.isComment() ||
590 (wikiTokenizer.isListItem() &&
591 wikiTokenizer.listItemPrefix().length() > prefix.length() &&
592 wikiTokenizer.listItemPrefix().startsWith(prefix))) {
593 if (wikiTokenizer.isListItem()) {
594 listPrefixes.add(wikiTokenizer.listItemPrefix());
595 listLines.add(wikiTokenizer.listItemWikiText());
598 if (wikiTokenizer.nextToken() != null) {
599 wikiTokenizer.returnToLineStart();
601 System.out.println("list lines: " + listLines);
602 System.out.println("list prefixes: " + listPrefixes);
604 final PairEntry pairEntry = new PairEntry();
605 final IndexedEntry indexedEntry = new IndexedEntry(pairEntry);
607 final String foreign = trim(title);
609 final StringBuilder englishBuilder = new StringBuilder();
611 final String mainLine = listLines.get(0);
613 final WikiTokenizer englishTokenizer = new WikiTokenizer(mainLine);
614 while (englishTokenizer.nextToken() != null) {
615 // TODO handle form of....
616 if (englishTokenizer.isPlainText()) {
617 englishBuilder.append(englishTokenizer.token());
618 enIndexBuilder.addEntryWithString(indexedEntry, englishTokenizer.token(), EntryTypeName.WIKTIONARY_ENGLISH_DEF);
619 } else if (englishTokenizer.isWikiLink()) {
620 final String text = englishTokenizer.wikiLinkText();
621 final String link = englishTokenizer.wikiLinkDest();
623 if (link.contains("#English")) {
624 englishBuilder.append(text);
625 enIndexBuilder.addEntryWithString(indexedEntry, text, EntryTypeName.WIKTIONARY_ENGLISH_DEF_WIKI_LINK);
626 } else if (link.contains("#") && this.langPattern.matcher(link).find()) {
627 englishBuilder.append(text);
628 otherIndexBuilder.addEntryWithString(indexedEntry, text, EntryTypeName.WIKTIONARY_ENGLISH_DEF_OTHER_LANG);
630 System.err.println("Special link: " + englishTokenizer.token());
631 // TODO: something here...
635 englishBuilder.append(text);
636 if (!UNINDEXED_WIKI_TEXT.matcher(text).find()) {
637 enIndexBuilder.addEntryWithString(indexedEntry, text, EntryTypeName.WIKTIONARY_ENGLISH_DEF_WIKI_LINK);
640 } else if (englishTokenizer.isFunction()) {
641 final String name = englishTokenizer.functionName();
642 if (name.contains(" conjugation of ") ||
643 name.contains(" form of ") ||
644 name.contains(" feminine of ") ||
645 name.contains(" plural of ")) {
646 // Ignore these in the index, they're really annoying....
647 englishBuilder.append(englishTokenizer.token());
649 System.err.println("Unexpected function: " + englishTokenizer.token());
652 if (englishTokenizer.isComment() || englishTokenizer.isMarkup()) {
654 System.err.println("Unexpected definition text: " + englishTokenizer.token());
658 final String english = trim(englishBuilder.toString());
659 if (english.length() > 0) {
660 final Pair pair = new Pair(english, trim(foreignText), this.swap);
661 pairEntry.pairs.add(pair);
662 otherIndexBuilder.addEntryWithString(indexedEntry, title, EntryTypeName.WIKTIONARY_TITLE_SINGLE, EntryTypeName.WIKTIONARY_TITLE_MULTI);
663 for (final String form : forms) {
664 otherIndexBuilder.addEntryWithString(indexedEntry, form, EntryTypeName.WIKTIONARY_FORM_SINGLE, EntryTypeName.WIKTIONARY_FORM_MULTI);