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.enwiktionary;
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.LinkedHashSet;
27 import java.util.List;
30 import java.util.logging.Logger;
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;
38 import com.hughes.android.dictionary.parser.WikiTokenizer;
39 import com.hughes.util.ListUtil;
41 public class EnWiktionaryXmlParser {
43 private static final String TRANSLITERATION_FORMAT = " (tr. %s)";
45 static final Logger LOG = Logger.getLogger(EnWiktionaryXmlParser.class.getName());
47 // TODO: process {{ttbc}} lines
49 static final Pattern partOfSpeechHeader = Pattern.compile(
50 "Noun|Verb|Adjective|Adverb|Pronoun|Conjunction|Interjection|" +
51 "Preposition|Proper noun|Article|Prepositional phrase|Acronym|" +
52 "Abbreviation|Initialism|Contraction|Prefix|Suffix|Symbol|Letter|" +
53 "Ligature|Idiom|Phrase|\\{\\{acronym\\}\\}|\\{\\{initialism\\}\\}|" +
54 // These are @deprecated:
55 "Noun form|Verb form|Adjective form|Nominal phrase|Noun phrase|" +
56 "Verb phrase|Transitive verb|Intransitive verb|Reflexive verb|" +
57 // These are extras I found:
58 "Determiner|Numeral|Number|Cardinal number|Ordinal number|Proverb|" +
59 "Particle|Interjection|Pronominal adverb" +
60 "Han character|Hanzi|Hanja|Kanji|Katakana character|Syllable");
62 final IndexBuilder enIndexBuilder;
63 final IndexBuilder foreignIndexBuilder;
64 final Pattern langPattern;
65 final Pattern langCodePattern;
68 // State used while parsing.
71 ENGLISH_DEF_OF_FOREIGN,
78 public EnWiktionaryXmlParser(final IndexBuilder enIndexBuilder, final IndexBuilder otherIndexBuilder, final Pattern langPattern, final Pattern langCodePattern, final boolean swap) {
79 this.enIndexBuilder = enIndexBuilder;
80 this.foreignIndexBuilder = otherIndexBuilder;
81 this.langPattern = langPattern;
82 this.langCodePattern = langCodePattern;
87 public void parse(final File file, final int pageLimit) throws IOException {
89 final DataInputStream dis = new DataInputStream(new BufferedInputStream(new FileInputStream(file)));
91 if (pageLimit >= 0 && pageCount >= pageLimit) {
96 title = dis.readUTF();
97 } catch (EOFException e) {
98 LOG.warning("Error reading split!");
102 final String heading = dis.readUTF();
103 final int bytesLength = dis.readInt();
104 final byte[] bytes = new byte[bytesLength];
105 dis.readFully(bytes);
106 final String text = new String(bytes, "UTF8");
108 parseSection(heading, text);
111 if (pageCount % 1000 == 0) {
112 LOG.info("pageCount=" + pageCount);
117 private void parseSection(String heading, final String text) {
118 if (title.startsWith("Wiktionary:") ||
119 title.startsWith("Template:") ||
120 title.startsWith("Appendix:") ||
121 title.startsWith("Category:") ||
122 title.startsWith("Index:") ||
123 title.startsWith("MediaWiki:") ||
124 title.startsWith("TransWiki:") ||
125 title.startsWith("Citations:") ||
126 title.startsWith("Concordance:") ||
127 title.startsWith("Help:")) {
131 heading = heading.replaceAll("=", "").trim();
132 if (heading.equals("English")) {
134 } else if (langPattern.matcher(heading).find()){
135 doForeignWord(heading, text);
140 // -------------------------------------------------------------------------
142 private void doEnglishWord(String text) {
147 final WikiTokenizer wikiTokenizer = new WikiTokenizer(text);
148 while (wikiTokenizer.nextToken() != null) {
150 if (wikiTokenizer.isHeading()) {
151 final String headerName = wikiTokenizer.headingWikiText();
153 if (wikiTokenizer.headingDepth() <= posDepth) {
158 if (partOfSpeechHeader.matcher(headerName).matches()) {
159 posDepth = wikiTokenizer.headingDepth();
160 pos = wikiTokenizer.headingWikiText();
161 // TODO: if we're inside the POS section, we should handle the first title line...
163 } else if (headerName.equals("Translations")) {
165 LOG.warning("Translations without POS: " + title);
167 doTranslations(wikiTokenizer, pos);
168 } else if (headerName.equals("Pronunciation")) {
169 //doPronunciation(wikiLineReader);
171 } else if (wikiTokenizer.isFunction()) {
172 final String name = wikiTokenizer.functionName();
173 if (name.equals("head")) {
174 LOG.warning("{{head}} without POS: " + title);
180 final AppendAndIndexWikiCallback appendAndIndexWikiCallback = new AppendAndIndexWikiCallback(this);
182 private void doTranslations(final WikiTokenizer wikiTokenizer, final String pos) {
183 if (title.equals("absolutely")) {
184 //System.out.println();
187 String topLevelLang = null;
189 boolean done = false;
190 while (wikiTokenizer.nextToken() != null) {
191 if (wikiTokenizer.isHeading()) {
192 wikiTokenizer.returnToLineStart();
199 // Check whether we care about this line:
201 if (wikiTokenizer.isFunction()) {
202 final String functionName = wikiTokenizer.functionName();
203 final List<String> positionArgs = wikiTokenizer.functionPositionArgs();
205 if (functionName.equals("trans-top")) {
207 if (wikiTokenizer.functionPositionArgs().size() >= 1) {
208 sense = positionArgs.get(0);
209 // TODO: could emphasize words in [[brackets]] inside sense.
210 sense = WikiTokenizer.toPlainText(sense);
211 //LOG.info("Sense: " + sense);
213 } else if (functionName.equals("trans-bottom")) {
215 } else if (functionName.equals("trans-mid")) {
216 } else if (functionName.equals("trans-see")) {
217 // TODO: would also be nice...
218 } else if (functionName.startsWith("picdic")) {
219 } else if (functionName.startsWith("checktrans")) {
221 } else if (functionName.startsWith("ttbc")) {
222 wikiTokenizer.nextLine();
223 // TODO: would be great to handle ttbc
224 // TODO: Check this: done = true;
226 LOG.warning("Unexpected translation wikifunction: " + wikiTokenizer.token() + ", title=" + title);
228 } else if (wikiTokenizer.isListItem()) {
229 final String line = wikiTokenizer.listItemWikiText();
230 // This line could produce an output...
232 if (line.contains("ich hoan dich gear")) {
233 //System.out.println();
236 // First strip the language and check whether it matches.
237 // And hold onto it for sub-lines.
238 final int colonIndex = line.indexOf(":");
239 if (colonIndex == -1) {
243 final String lang = trim(WikiTokenizer.toPlainText(line.substring(0, colonIndex)));
244 final boolean appendLang;
245 if (wikiTokenizer.listItemPrefix().length() == 1) {
247 final boolean thisFind = langPattern.matcher(lang).find();
251 appendLang = !langPattern.matcher(lang).matches();
252 } else if (topLevelLang == null) {
255 // Two-level -- the only way we won't append is if this second level matches exactly.
256 if (!langPattern.matcher(lang).matches() && !langPattern.matcher(topLevelLang).find()) {
259 appendLang = !langPattern.matcher(lang).matches();
262 String rest = line.substring(colonIndex + 1).trim();
263 if (rest.length() > 0) {
264 doTranslationLine(line, appendLang ? lang : null, pos, sense, rest);
267 } else if (wikiTokenizer.remainderStartsWith("''See''")) {
268 wikiTokenizer.nextLine();
269 LOG.fine("Skipping See line: " + wikiTokenizer.token());
270 } else if (wikiTokenizer.isWikiLink()) {
271 final String wikiLink = wikiTokenizer.wikiLinkText();
272 if (wikiLink.contains(":") && wikiLink.contains(title)) {
273 } else if (wikiLink.contains("Category:")) {
275 LOG.warning("Unexpected wikiLink: " + wikiTokenizer.token() + ", title=" + title);
277 } else if (wikiTokenizer.isNewline() || wikiTokenizer.isMarkup() || wikiTokenizer.isComment()) {
279 final String token = wikiTokenizer.token();
280 if (token.equals("----")) {
282 LOG.warning("Unexpected translation token: " + wikiTokenizer.token() + ", title=" + title);
289 private void doTranslationLine(final String line, final String lang, final String pos, final String sense, final String rest) {
290 state = State.TRANSLATION_LINE;
291 // Good chance we'll actually file this one...
292 final PairEntry pairEntry = new PairEntry();
293 final IndexedEntry indexedEntry = new IndexedEntry(pairEntry);
295 final StringBuilder foreignText = new StringBuilder();
296 appendAndIndexWikiCallback.reset(foreignText, indexedEntry);
297 appendAndIndexWikiCallback.functionCallbacks.clear();
298 appendAndIndexWikiCallback.functionCallbacks.putAll(FunctionCallbacksDefault.DEFAULT);
299 appendAndIndexWikiCallback.dispatch(rest, foreignIndexBuilder, EntryTypeName.WIKTIONARY_TRANSLATION_OTHER_TEXT);
301 if (foreignText.length() == 0) {
302 LOG.warning("Empty foreignText: " + line);
307 foreignText.insert(0, String.format("(%s) ", lang));
310 StringBuilder englishText = new StringBuilder();
312 englishText.append(title);
314 englishText.append(" (").append(sense).append(")");
315 enIndexBuilder.addEntryWithString(indexedEntry, sense, EntryTypeName.WIKTIONARY_TRANSLATION_SENSE);
318 englishText.append(" (").append(pos.toLowerCase()).append(")");
320 enIndexBuilder.addEntryWithString(indexedEntry, title, EntryTypeName.WIKTIONARY_TITLE_MULTI);
322 final Pair pair = new Pair(trim(englishText.toString()), trim(foreignText.toString()), swap);
323 pairEntry.pairs.add(pair);
324 if (!pairsAdded.add(pair.toString())) {
325 LOG.warning("Duplicate pair: " + pair.toString());
327 if (pair.toString().equals("libero {m} :: free (adjective)")) {
328 System.out.println();
334 Set<String> pairsAdded = new LinkedHashSet<String>();
336 // -------------------------------------------------------------------------
338 private void doForeignWord(final String lang, final String text) {
339 final WikiTokenizer wikiTokenizer = new WikiTokenizer(text);
340 while (wikiTokenizer.nextToken() != null) {
341 if (wikiTokenizer.isHeading()) {
342 final String headingName = wikiTokenizer.headingWikiText();
343 if (headingName.equals("Translations")) {
344 LOG.warning("Translations not in English section: " + title);
345 } else if (headingName.equals("Pronunciation")) {
346 //doPronunciation(wikiLineReader);
347 } else if (partOfSpeechHeader.matcher(headingName).matches()) {
348 doForeignPartOfSpeech(lang, headingName, wikiTokenizer.headingDepth(), wikiTokenizer);
355 static final class ListSection {
356 final String firstPrefix;
357 final String firstLine;
358 final List<String> nextPrefixes = new ArrayList<String>();
359 final List<String> nextLines = new ArrayList<String>();
361 public ListSection(String firstPrefix, String firstLine) {
362 this.firstPrefix = firstPrefix;
363 this.firstLine = firstLine;
367 public String toString() {
368 return firstPrefix + firstLine + "{ " + nextPrefixes + "}";
373 int foreignCount = 0;
374 private void doForeignPartOfSpeech(final String lang, String posHeading, final int posDepth, WikiTokenizer wikiTokenizer) {
375 if (++foreignCount % 1000 == 0) {
376 LOG.info("***" + lang + ", " + title + ", pos=" + posHeading + ", foreignCount=" + foreignCount);
378 if (title.equals("moro")) {
379 System.out.println();
382 boolean titleAppended = false;
383 final StringBuilder foreignBuilder = new StringBuilder();
384 final Collection<String> wordForms = new ArrayList<String>();
385 final List<ListSection> listSections = new ArrayList<ListSection>();
389 ListSection lastListSection = null;
391 int currentHeadingDepth = posDepth;
392 while (wikiTokenizer.nextToken() != null) {
393 if (wikiTokenizer.isHeading()) {
394 currentHeadingDepth = wikiTokenizer.headingDepth();
396 if (currentHeadingDepth <= posDepth) {
397 wikiTokenizer.returnToLineStart();
402 if (currentHeadingDepth > posDepth) {
403 // TODO: deal with other neat info sections
407 if (wikiTokenizer.isFunction()) {
408 final String name = wikiTokenizer.functionName();
409 final List<String> args = wikiTokenizer.functionPositionArgs();
410 final Map<String,String> namedArgs = wikiTokenizer.functionNamedArgs();
411 // First line is generally a repeat of the title with some extra information.
412 // We need to build up the left side (foreign text, tokens) separately from the
413 // right side (English). The left-side may get paired with multiple right sides.
414 // The left side should get filed under every form of the word in question (singular, plural).
416 // For verbs, the conjugation comes later on in a deeper section.
417 // Ideally, we'd want to file every English entry with the verb
418 // under every verb form coming from the conjugation.
419 // Ie. under "fa": see: "make :: fare" and "do :: fare"
420 // But then where should we put the conjugation table?
421 // 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!)
422 // for the conjugation table from "fa".
423 // Would like to be able to link to a lang#token.
424 if (FunctionCallbacksDefault.DEFAULT.get(name) instanceof FunctionCallbacksDefault.Gender) {
426 appendAndIndexWikiCallback.reset(foreignBuilder, null);
427 FunctionCallbacksDefault.DEFAULT.get(name).onWikiFunction(wikiTokenizer, name, args, namedArgs, this, appendAndIndexWikiCallback);
428 } else if (name.equals("wikipedia")) {
429 namedArgs.remove("lang");
430 if (args.size() > 1 || !namedArgs.isEmpty()) {
432 foreignBuilder.append(wikiTokenizer.token());
433 } else if (args.size() == 1) {
434 foreignBuilder.append(wikiTokenizer.token());
436 //foreignBuilder.append(title);
438 } else if (name.equals("attention") || name.equals("zh-attention")) {
439 // See: http://en.wiktionary.org/wiki/Template:attention
441 } else if (name.equals("infl") || name.equals("head")) {
442 // See: http://en.wiktionary.org/wiki/Template:infl
443 final String langCode = ListUtil.get(args, 0);
444 String head = namedArgs.remove("head");
446 head = namedArgs.remove("title"); // Bug
451 head = WikiTokenizer.toPlainText(head);
453 titleAppended = true;
455 namedArgs.keySet().removeAll(USELESS_WIKI_ARGS);
457 final String tr = namedArgs.remove("tr");
458 String g = namedArgs.remove("g");
460 g = namedArgs.remove("gender");
462 final String g2 = namedArgs.remove("g2");
463 final String g3 = namedArgs.remove("g3");
465 foreignBuilder.append(head);
468 foreignBuilder.append(" {").append(g);
470 foreignBuilder.append("|").append(g2);
473 foreignBuilder.append("|").append(g3);
475 foreignBuilder.append("}");
479 foreignBuilder.append(String.format(TRANSLITERATION_FORMAT, tr));
483 final String pos = ListUtil.get(args, 1);
485 foreignBuilder.append(" (").append(pos).append(")");
487 for (int i = 2; i < args.size(); i += 2) {
488 final String inflName = ListUtil.get(args, i);
489 final String inflValue = ListUtil.get(args, i + 1);
490 foreignBuilder.append(", ").append(WikiTokenizer.toPlainText(inflName));
491 if (inflValue != null && inflValue.length() > 0) {
492 foreignBuilder.append(": ").append(WikiTokenizer.toPlainText(inflValue));
493 wordForms.add(inflValue);
496 for (final String key : namedArgs.keySet()) {
497 final String value = WikiTokenizer.toPlainText(namedArgs.get(key));
498 foreignBuilder.append(" ").append(key).append("=").append(value);
499 wordForms.add(value);
501 } else if (name.equals("it-noun")) {
502 titleAppended = true;
503 final String base = ListUtil.get(args, 0);
504 final String gender = ListUtil.get(args, 1);
505 final String singular = base + ListUtil.get(args, 2, null);
506 final String plural = base + ListUtil.get(args, 3, null);
507 foreignBuilder.append(String.format(" %s {%s}, %s {pl}", singular, gender, plural, plural));
508 wordForms.add(singular);
509 wordForms.add(plural);
510 if (!namedArgs.isEmpty() || args.size() > 4) {
511 LOG.warning("Invalid it-noun: " + wikiTokenizer.token());
513 } else if (name.equals("it-proper noun")) {
514 foreignBuilder.append(wikiTokenizer.token());
515 } else if (name.equals("it-adj")) {
516 foreignBuilder.append(wikiTokenizer.token());
517 } else if (name.startsWith("it-conj")) {
518 if (name.equals("it-conj-are")) {
519 itConjAre(args, namedArgs);
520 } else if (name.equals("it-conj-ere")) {
521 } else if (name.equals("it-conj-ire")) {
523 LOG.warning("Unknown conjugation: " + wikiTokenizer.token());
527 foreignBuilder.append(wikiTokenizer.token());
528 // LOG.warning("Unknown function: " + wikiTokenizer.token());
530 } else if (wikiTokenizer.isListItem()) {
531 final String prefix = wikiTokenizer.listItemPrefix();
532 if (lastListSection != null &&
533 prefix.startsWith(lastListSection.firstPrefix) &&
534 prefix.length() > lastListSection.firstPrefix.length()) {
535 lastListSection.nextPrefixes.add(prefix);
536 lastListSection.nextLines.add(wikiTokenizer.listItemWikiText());
538 lastListSection = new ListSection(prefix, wikiTokenizer.listItemWikiText());
539 listSections.add(lastListSection);
541 } else if (lastListSection != null) {
542 // Don't append anything after the lists, because there's crap.
543 } else if (wikiTokenizer.isWikiLink()) {
545 foreignBuilder.append(wikiTokenizer.wikiLinkText());
547 } else if (wikiTokenizer.isPlainText()) {
549 foreignBuilder.append(wikiTokenizer.token());
551 } else if (wikiTokenizer.isMarkup() || wikiTokenizer.isNewline() || wikiTokenizer.isComment()) {
554 LOG.warning("Unexpected token: " + wikiTokenizer.token());
559 // Here's where we exit.
560 // Should we make an entry even if there are no foreign list items?
561 String foreign = foreignBuilder.toString().trim();
562 if (!titleAppended && !foreign.toLowerCase().startsWith(title.toLowerCase())) {
563 foreign = String.format("%s %s", title, foreign);
565 if (!langPattern.matcher(lang).matches()) {
566 foreign = String.format("(%s) %s", lang, foreign);
568 for (final ListSection listSection : listSections) {
569 doForeignListItem(foreign, title, wordForms, listSection);
575 // Might only want to remove "lang" if it's equal to "zh", for example.
576 static final Set<String> USELESS_WIKI_ARGS = new LinkedHashSet<String>(
584 private void doForeignListItem(final String foreignText, String title, final Collection<String> forms, final ListSection listSection) {
585 state = State.ENGLISH_DEF_OF_FOREIGN;
586 final String prefix = listSection.firstPrefix;
587 if (prefix.length() > 1) {
588 // Could just get looser and say that any prefix longer than first is a sublist.
589 LOG.warning("Prefix too long: " + listSection);
593 final PairEntry pairEntry = new PairEntry();
594 final IndexedEntry indexedEntry = new IndexedEntry(pairEntry);
596 final StringBuilder englishBuilder = new StringBuilder();
597 final String mainLine = listSection.firstLine;
599 appendAndIndexWikiCallback.reset(englishBuilder, indexedEntry);
600 appendAndIndexWikiCallback.functionCallbacks.putAll(FunctionCallbacksDefault.DEFAULT);
601 appendAndIndexWikiCallback.dispatch(mainLine, enIndexBuilder, EntryTypeName.WIKTIONARY_ENGLISH_DEF);
603 final String english = trim(englishBuilder.toString());
604 if (english.length() > 0) {
605 final Pair pair = new Pair(english, trim(foreignText), this.swap);
606 pairEntry.pairs.add(pair);
607 foreignIndexBuilder.addEntryWithString(indexedEntry, title, EntryTypeName.WIKTIONARY_TITLE_MULTI);
608 for (final String form : forms) {
609 foreignIndexBuilder.addEntryWithString(indexedEntry, form, EntryTypeName.WIKTIONARY_INFLECTED_FORM_MULTI);
614 String lastForeign = null;
615 for (int i = 0; i < listSection.nextPrefixes.size(); ++i) {
616 final String nextPrefix = listSection.nextPrefixes.get(i);
617 final String nextLine = listSection.nextLines.get(i);
618 int dash = nextLine.indexOf("—");
621 dash = nextLine.indexOf("—");
625 dash = nextLine.indexOf(" - ");
629 if ((nextPrefix.equals("#:") || nextPrefix.equals("##:")) && dash != -1) {
630 final String foreignEx = nextLine.substring(0, dash);
631 final String englishEx = nextLine.substring(dash + mdashLen);
632 final Pair pair = new Pair(formatAndIndexExampleString(englishEx, enIndexBuilder, indexedEntry), formatAndIndexExampleString(foreignEx, foreignIndexBuilder, indexedEntry), swap);
633 if (pair.lang1 != "--" && pair.lang1 != "--") {
634 pairEntry.pairs.add(pair);
637 } else if (nextPrefix.equals("#:") || nextPrefix.equals("##:")){
638 final Pair pair = new Pair("--", formatAndIndexExampleString(nextLine, null, indexedEntry), swap);
639 lastForeign = nextLine;
640 if (pair.lang1 != "--" && pair.lang1 != "--") {
641 pairEntry.pairs.add(pair);
643 } else if (nextPrefix.equals("#::") || nextPrefix.equals("#**")) {
644 if (lastForeign != null && pairEntry.pairs.size() > 0) {
645 pairEntry.pairs.remove(pairEntry.pairs.size() - 1);
646 final Pair pair = new Pair(formatAndIndexExampleString(nextLine, enIndexBuilder, indexedEntry), formatAndIndexExampleString(lastForeign, foreignIndexBuilder, indexedEntry), swap);
647 if (pair.lang1 != "--" || pair.lang2 != "--") {
648 pairEntry.pairs.add(pair);
652 LOG.warning("TODO: English example with no foreign: " + title + ", " + nextLine);
653 final Pair pair = new Pair("--", formatAndIndexExampleString(nextLine, null, indexedEntry), swap);
654 if (pair.lang1 != "--" || pair.lang2 != "--") {
655 pairEntry.pairs.add(pair);
658 } else if (nextPrefix.equals("#*")) {
659 // Can't really index these.
660 final Pair pair = new Pair("--", formatAndIndexExampleString(nextLine, null, indexedEntry), swap);
661 lastForeign = nextLine;
662 if (pair.lang1 != "--" || pair.lang2 != "--") {
663 pairEntry.pairs.add(pair);
665 } else if (nextPrefix.equals("#::*") || nextPrefix.equals("##") || nextPrefix.equals("#*:") || nextPrefix.equals("#:*") || true) {
666 final Pair pair = new Pair("--", formatAndIndexExampleString(nextLine, null, indexedEntry), swap);
667 if (pair.lang1 != "--" || pair.lang2 != "--") {
668 pairEntry.pairs.add(pair);
676 private String formatAndIndexExampleString(final String example, final IndexBuilder indexBuilder, final IndexedEntry indexedEntry) {
677 final WikiTokenizer wikiTokenizer = new WikiTokenizer(example, false);
678 final StringBuilder builder = new StringBuilder();
679 boolean insideTripleQuotes = false;
680 while (wikiTokenizer.nextToken() != null) {
681 if (wikiTokenizer.isPlainText()) {
682 builder.append(wikiTokenizer.token());
683 if (indexBuilder != null) {
684 indexBuilder.addEntryWithStringNoSingle(indexedEntry, wikiTokenizer.token(), EntryTypeName.WIKTIONARY_EXAMPLE);
686 } else if (wikiTokenizer.isWikiLink()) {
687 final String text = wikiTokenizer.wikiLinkText().replaceAll("'", "");
688 builder.append(text);
689 if (indexBuilder != null) {
690 indexBuilder.addEntryWithStringNoSingle(indexedEntry, text, EntryTypeName.WIKTIONARY_EXAMPLE);
692 } else if (wikiTokenizer.isFunction()) {
693 builder.append(wikiTokenizer.token());
694 } else if (wikiTokenizer.isMarkup()) {
695 if (wikiTokenizer.token().equals("'''")) {
696 insideTripleQuotes = !insideTripleQuotes;
698 } else if (wikiTokenizer.isComment() || wikiTokenizer.isNewline()) {
701 LOG.warning("unexpected token: " + wikiTokenizer.token());
704 final String result = trim(builder.toString());
705 return result.length() > 0 ? result : "--";
709 private void itConjAre(List<String> args, Map<String, String> namedArgs) {
710 final String base = args.get(0);
711 final String aux = args.get(1);
713 putIfMissing(namedArgs, "inf", base + "are");
714 putIfMissing(namedArgs, "aux", aux);
715 putIfMissing(namedArgs, "ger", base + "ando");
716 putIfMissing(namedArgs, "presp", base + "ante");
717 putIfMissing(namedArgs, "pastp", base + "ato");
719 putIfMissing(namedArgs, "pres1s", base + "o");
720 putIfMissing(namedArgs, "pres2s", base + "i");
721 putIfMissing(namedArgs, "pres3s", base + "a");
722 putIfMissing(namedArgs, "pres1p", base + "iamo");
723 putIfMissing(namedArgs, "pres2p", base + "ate");
724 putIfMissing(namedArgs, "pres3p", base + "ano");
726 putIfMissing(namedArgs, "imperf1s", base + "avo");
727 putIfMissing(namedArgs, "imperf2s", base + "avi");
728 putIfMissing(namedArgs, "imperf3s", base + "ava");
729 putIfMissing(namedArgs, "imperf1p", base + "avamo");
730 putIfMissing(namedArgs, "imperf2p", base + "avate");
731 putIfMissing(namedArgs, "imperf3p", base + "avano");
733 putIfMissing(namedArgs, "prem1s", base + "ai");
734 putIfMissing(namedArgs, "prem2s", base + "asti");
735 putIfMissing(namedArgs, "prem3s", base + "ò");
736 putIfMissing(namedArgs, "prem1p", base + "ammo");
737 putIfMissing(namedArgs, "prem2p", base + "aste");
738 putIfMissing(namedArgs, "prem3p", base + "arono");
740 putIfMissing(namedArgs, "fut1s", base + "erò");
741 putIfMissing(namedArgs, "fut2s", base + "erai");
742 putIfMissing(namedArgs, "fut3s", base + "erà");
743 putIfMissing(namedArgs, "fut1p", base + "eremo");
744 putIfMissing(namedArgs, "fut2p", base + "erete");
745 putIfMissing(namedArgs, "fut3p", base + "eranno");
747 putIfMissing(namedArgs, "cond1s", base + "erei");
748 putIfMissing(namedArgs, "cond2s", base + "eresti");
749 putIfMissing(namedArgs, "cond3s", base + "erebbe");
750 putIfMissing(namedArgs, "cond1p", base + "eremmo");
751 putIfMissing(namedArgs, "cond2p", base + "ereste");
752 putIfMissing(namedArgs, "cond3p", base + "erebbero");
753 // Subjunctive / congiuntivo
754 putIfMissing(namedArgs, "sub123s", base + "i");
755 putIfMissing(namedArgs, "sub1p", base + "iamo");
756 putIfMissing(namedArgs, "sub2p", base + "iate");
757 putIfMissing(namedArgs, "sub3p", base + "ino");
758 // Imperfect subjunctive
759 putIfMissing(namedArgs, "impsub12s", base + "assi");
760 putIfMissing(namedArgs, "impsub3s", base + "asse");
761 putIfMissing(namedArgs, "impsub1p", base + "assimo");
762 putIfMissing(namedArgs, "impsub2p", base + "aste");
763 putIfMissing(namedArgs, "impsub3p", base + "assero");
765 putIfMissing(namedArgs, "imp2s", base + "a");
766 putIfMissing(namedArgs, "imp3s", base + "i");
767 putIfMissing(namedArgs, "imp1p", base + "iamo");
768 putIfMissing(namedArgs, "imp2p", base + "ate");
769 putIfMissing(namedArgs, "imp3p", base + "ino");
772 itConj(args, namedArgs);
776 private void itConj(List<String> args, Map<String, String> namedArgs) {
777 // TODO Auto-generated method stub
782 private static void putIfMissing(final Map<String, String> namedArgs, final String key,
783 final String value) {
784 final String oldValue = namedArgs.get(key);
785 if (oldValue == null || oldValue.length() == 0) {
786 namedArgs.put(key, value);
790 // TODO: check how ='' and =| are manifested....
791 // TODO: get this right in -are
792 private static void putOrNullify(final Map<String, String> namedArgs, final String key,
793 final String value) {
794 final String oldValue = namedArgs.get(key);
795 if (oldValue == null/* || oldValue.length() == 0*/) {
796 namedArgs.put(key, value);
798 if (oldValue.equals("''")) {
799 namedArgs.put(key, "");
804 static final Pattern whitespace = Pattern.compile("\\s+");
805 static String trim(final String s) {
806 return whitespace.matcher(s).replaceAll(" ").trim();