+// Copyright 2011 Google Inc. All Rights Reserved.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
package com.hughes.android.dictionary.parser;
+import java.io.BufferedInputStream;
+import java.io.DataInputStream;
+import java.io.EOFException;
import java.io.File;
+import java.io.FileInputStream;
import java.io.IOException;
import java.util.ArrayList;
import java.util.Arrays;
+import java.util.Collection;
import java.util.LinkedHashSet;
import java.util.List;
import java.util.Map;
import java.util.Set;
+import java.util.logging.Logger;
import java.util.regex.Pattern;
-import javax.xml.parsers.ParserConfigurationException;
-import javax.xml.parsers.SAXParser;
-import javax.xml.parsers.SAXParserFactory;
-
-import org.xml.sax.Attributes;
-import org.xml.sax.SAXException;
-
-import com.hughes.android.dictionary.engine.DictionaryBuilder;
+import com.hughes.android.dictionary.engine.EntryTypeName;
import com.hughes.android.dictionary.engine.IndexBuilder;
-import com.hughes.android.dictionary.parser.WikiWord.FormOf;
-import com.hughes.android.dictionary.parser.WikiWord.Translation;
-import com.hughes.util.ListUtil;
-import com.hughes.util.StringUtil;
+import com.hughes.android.dictionary.engine.IndexedEntry;
+import com.hughes.android.dictionary.engine.PairEntry;
+import com.hughes.android.dictionary.engine.PairEntry.Pair;
+
+public class EnWiktionaryXmlParser {
+
+ private static final String TRANSLITERATION_FORMAT = " (tr. %s)";
-public class EnWiktionaryXmlParser extends org.xml.sax.helpers.DefaultHandler implements WikiCallback {
+ static final Logger LOG = Logger.getLogger(EnWiktionaryXmlParser.class.getName());
+
+ // TODO: process {{ttbc}} lines
static final Pattern partOfSpeechHeader = Pattern.compile(
"Noun|Verb|Adjective|Adverb|Pronoun|Conjunction|Interjection|" +
"Preposition|Proper noun|Article|Prepositional phrase|Acronym|" +
"Abbreviation|Initialism|Contraction|Prefix|Suffix|Symbol|Letter|" +
- "Ligature|Idiom|Phrase|" +
+ "Ligature|Idiom|Phrase|\\{\\{acronym\\}\\}|\\{\\{initialism\\}\\}|" +
// These are @deprecated:
"Noun form|Verb form|Adjective form|Nominal phrase|Noun phrase|" +
"Verb phrase|Transitive verb|Intransitive verb|Reflexive verb|" +
"Determiner|Numeral|Number|Cardinal number|Ordinal number|Proverb|" +
"Particle|Interjection|Pronominal adverb" +
"Han character|Hanzi|Hanja|Kanji|Katakana character|Syllable");
-
- static final Pattern wikiMarkup = Pattern.compile("\\[\\[|\\]\\]|''+");
-
- final DictionaryBuilder dictBuilder;
- final IndexBuilder[] indexBuilders;
- final Pattern[] langPatterns;
- final int enIndexBuilder;
-
- StringBuilder titleBuilder;
- StringBuilder textBuilder;
- StringBuilder currentBuilder = null;
-
- public EnWiktionaryXmlParser(final DictionaryBuilder dictBuilder, final Pattern[] langPatterns, final int enIndexBuilder) {
- assert langPatterns.length == 2;
- this.dictBuilder = dictBuilder;
- this.indexBuilders = dictBuilder.indexBuilders.toArray(new IndexBuilder[0]);
- this.langPatterns = langPatterns;
+ final IndexBuilder enIndexBuilder;
+ final IndexBuilder otherIndexBuilder;
+ final Pattern langPattern;
+ final Pattern langCodePattern;
+ final boolean swap;
+
+ public EnWiktionaryXmlParser(final IndexBuilder enIndexBuilder, final IndexBuilder otherIndexBuilder, final Pattern langPattern, final Pattern langCodePattern, final boolean swap) {
this.enIndexBuilder = enIndexBuilder;
+ this.otherIndexBuilder = otherIndexBuilder;
+ this.langPattern = langPattern;
+ this.langCodePattern = langCodePattern;
+ this.swap = swap;
}
- @Override
- public void startElement(String uri, String localName, String qName,
- Attributes attributes) {
- currentBuilder = null;
- if ("page".equals(qName)) {
- titleBuilder = new StringBuilder();
+
+ public void parse(final File file, final int pageLimit) throws IOException {
+ int pageCount = 0;
+ final DataInputStream dis = new DataInputStream(new BufferedInputStream(new FileInputStream(file)));
+ while (true) {
+ if (pageLimit >= 0 && pageCount >= pageLimit) {
+ return;
+ }
- // Start with "\n" to better match certain strings.
- textBuilder = new StringBuilder("\n");
- } else if ("title".equals(qName)) {
- currentBuilder = titleBuilder;
- } else if ("text".equals(qName)) {
- currentBuilder = textBuilder;
- }
- }
-
- @Override
- public void characters(char[] ch, int start, int length) throws SAXException {
- if (currentBuilder != null) {
- currentBuilder.append(ch, start, length);
- }
- }
+ final String title;
+ try {
+ title = dis.readUTF();
+ } catch (EOFException e) {
+ dis.close();
+ return;
+ }
+ final String heading = dis.readUTF();
+ final int bytesLength = dis.readInt();
+ final byte[] bytes = new byte[bytesLength];
+ dis.readFully(bytes);
+ final String text = new String(bytes, "UTF8");
+
+ parseSection(title, heading, text);
- @Override
- public void endElement(String uri, String localName, String qName)
- throws SAXException {
- currentBuilder = null;
- if ("page".equals(qName)) {
- endPage();
+ ++pageCount;
+ if (pageCount % 1000 == 0) {
+ LOG.info("pageCount=" + pageCount);
+ }
}
}
-
- public void parse(final File file) throws ParserConfigurationException,
- SAXException, IOException {
- final SAXParser parser = SAXParserFactory.newInstance().newSAXParser();
- parser.parse(file, this);
- }
-
- int pageCount = 0;
- private void endPage() {
- title = titleBuilder.toString();
- ++pageCount;
- if (pageCount % 1000 == 0) {
- System.out.println("pageCount=" + pageCount);
- }
+ private void parseSection(final String title, String heading, final String text) {
if (title.startsWith("Wiktionary:") ||
title.startsWith("Template:") ||
title.startsWith("Appendix:") ||
title.startsWith("Help:")) {
return;
}
- currentDepth = 0;
- words.clear();
- currentHeading = null;
- insidePartOfSpeech = false;
-// System.err.println("Working on page: " + title);
- try {
- WikiParser.parse(textBuilder.toString(), this);
- } catch (Throwable e) {
- System.err.println("Failure on page: " + title);
- e.printStackTrace(System.err);
+
+ heading = heading.replaceAll("=", "").trim();
+ if (heading.equals("English")) {
+ doEnglishWord(title, text);
+ } else if (langPattern.matcher(heading).find()){
+ doForeignWord(heading, title, text);
}
-
- for (final WikiWord word : words) {
- word.wikiWordToQuickDic(dictBuilder, enIndexBuilder);
- } // WikiWord
-
+
} // endPage()
-
-
- // ------------------------------------------------------------------------
- // ------------------------------------------------------------------------
- // ------------------------------------------------------------------------
- // ------------------------------------------------------------------------
-
- /**
- * Two things can happen:
- *
- * We can be in a ==German== section. There we will see English definitions.
- * Each POS should get its own QuickDic entry. Pretty much everything goes
- * in.
- *
- * Or we can be in an ==English== section with English definitions
- * and maybe see translations for languages we care about.
- *
- * In either case, we need to differentiate the subsections (Noun, Verb, etc.)
- * into separate QuickDic entries, but that's tricky--how do we know when we
- * found a subsection? Just ignore anything containing pronunciation and
- * etymology?
- *
- * How do we decide when to seal the deal on an entry?
- *
- * Would be nice if the parser told us about leaving sections....
- *
- *
- */
-
- String title;
- String currentHeading;
- int currentDepth;
- final List<WikiWord> words = new ArrayList<WikiWord>();
- WikiWord currentWord;
- WikiWord.PartOfSpeech currentPartOfSpeech;
- WikiWord.TranslationSense currentTranslationSense;
- boolean insidePartOfSpeech;
- StringBuilder wikiBuilder = null;
+ // -------------------------------------------------------------------------
- @Override
- public void onWikiLink(String[] args) {
- if (wikiBuilder == null) {
- return;
+ private void doEnglishWord(String title, String text) {
+
+ String pos = null;
+ int posDepth = -1;
+
+ final WikiTokenizer wikiTokenizer = new WikiTokenizer(text);
+ while (wikiTokenizer.nextToken() != null) {
+
+ if (wikiTokenizer.isHeading()) {
+ final String headerName = wikiTokenizer.headingWikiText();
+
+ if (wikiTokenizer.headingDepth() <= posDepth) {
+ pos = null;
+ posDepth = -1;
+ }
+
+ if (partOfSpeechHeader.matcher(headerName).matches()) {
+ posDepth = wikiTokenizer.headingDepth();
+ pos = wikiTokenizer.headingWikiText();
+ // TODO: if we're inside the POS section, we should handle the first title line...
+
+ } else if (headerName.equals("Translations")) {
+ if (pos == null) {
+ LOG.warning("Translations without POS: " + title);
+ }
+ doTranslations(title, wikiTokenizer, pos);
+ } else if (headerName.equals("Pronunciation")) {
+ //doPronunciation(wikiLineReader);
+ }
+ }
}
- wikiBuilder.append(args[args.length - 1]);
}
+
+
+ private static Set<String> encodings = new LinkedHashSet<String>(Arrays.asList("zh-ts",
+ "sd-Arab", "ku-Arab", "Arab", "unicode", "Laoo", "ur-Arab", "Thai",
+ "fa-Arab", "Khmr", "zh-tsp", "Cyrl", "IPAchar", "ug-Arab", "ko-inline",
+ "Jpan", "Kore", "Hebr", "rfscript", "Beng", "Mong", "Knda", "Cyrs",
+ "yue-tsj", "Mlym", "Tfng", "Grek", "yue-yue-j"));
- // ttbc: translations to be checked.
- static final Set<String> useRemainingArgTemplates = new LinkedHashSet<String>(Arrays.asList(
- "Arab", "Cyrl", "fa-Arab", "italbrac", "Khmr", "ku-Arab", "IPAchar", "Laoo",
- "sd-Arab", "Thai", "ttbc", "unicode", "ur-Arab", "yue-yue-j", "zh-ts",
- "zh-tsp", "zh-zh-p"));
- static final Set<String> ignoreTemplates = new LinkedHashSet<String>(Arrays.asList(""));
- static final Set<String> grammarTemplates = new LinkedHashSet<String>(Arrays.asList("impf", "pf"));
- static final Set<String> passThroughTemplates = new LinkedHashSet<String>(Arrays.asList("zzzzzzzzzzzzzzz"));
-
- @Override
- public void onTemplate(final List<String> positionalArgs, final Map<String,String> namedArgs) {
- if (positionalArgs.isEmpty()) {
- // This happens very rarely with special templates.
- return;
+ private void doTranslations(final String title, final WikiTokenizer wikiTokenizer, final String pos) {
+ if (title.equals("absolutely")) {
+ //System.out.println();
}
- final String name = positionalArgs.get(0);
- namedArgs.remove("lang");
- namedArgs.remove("nocat");
- namedArgs.remove("nocap");
- namedArgs.remove("sc");
-
- // Pronunciation
- if (currentWord != null) {
- if (name.equals("a")) {
- // accent tag
- currentWord.currentPronunciation = new StringBuilder();
- currentWord.accentToPronunciation.put(positionalArgs.get(1), currentWord.currentPronunciation);
+ String topLevelLang = null;
+ String sense = null;
+ boolean done = false;
+ while (wikiTokenizer.nextToken() != null) {
+ if (wikiTokenizer.isHeading()) {
+ wikiTokenizer.returnToLineStart();
return;
}
+ if (done) {
+ continue;
+ }
- if (name.equals("IPA") || name.equals("SAMPA") || name.equals("X-SAMPA") || name.equals("enPR")) {
- namedArgs.remove("lang");
- for (int i = 0; i < 100 && !namedArgs.isEmpty(); ++i) {
- final String pron = namedArgs.remove("" + i);
- if (pron != null) {
- positionalArgs.add(pron);
- } else {
- if (i > 10) {
- break;
- }
+ // Check whether we care about this line:
+
+ if (wikiTokenizer.isFunction()) {
+ final String functionName = wikiTokenizer.functionName();
+ final List<String> positionArgs = wikiTokenizer.functionPositionArgs();
+
+ if (functionName.equals("trans-top")) {
+ sense = null;
+ if (wikiTokenizer.functionPositionArgs().size() >= 1) {
+ sense = positionArgs.get(0);
+ // TODO: could emphasize words in [[brackets]] inside sense.
+ sense = WikiTokenizer.toPlainText(sense);
+ //LOG.info("Sense: " + sense);
}
+ } else if (functionName.equals("trans-bottom")) {
+ sense = null;
+ } else if (functionName.equals("trans-mid")) {
+ } else if (functionName.equals("trans-see")) {
+ // TODO: would also be nice...
+ } else if (functionName.startsWith("picdic")) {
+ } else if (functionName.startsWith("checktrans")) {
+ done = true;
+ } else if (functionName.startsWith("ttbc")) {
+ wikiTokenizer.nextLine();
+ // TODO: would be great to handle ttbc
+ // TODO: Check this: done = true;
+ } else {
+ LOG.warning("Unexpected translation wikifunction: " + wikiTokenizer.token() + ", title=" + title);
}
- if (!(positionalArgs.size() >= 2 && namedArgs.isEmpty())) {
- System.err.println("Invalid pronunciation: " + positionalArgs.toString() + namedArgs.toString());
- }
- if (currentWord.currentPronunciation == null) {
- currentWord.currentPronunciation = new StringBuilder();
- currentWord.accentToPronunciation.put("", currentWord.currentPronunciation);
+ } else if (wikiTokenizer.isListItem()) {
+ final String line = wikiTokenizer.listItemWikiText();
+ // This line could produce an output...
+
+ if (line.contains("ich hoan dich gear")) {
+ //System.out.println();
}
- if (currentWord.currentPronunciation.length() > 0) {
- currentWord.currentPronunciation.append("; ");
+
+ // First strip the language and check whether it matches.
+ // And hold onto it for sub-lines.
+ final int colonIndex = line.indexOf(":");
+ if (colonIndex == -1) {
+ continue;
}
- for (int i = 1; i < positionalArgs.size(); ++i) {
- if (i > 1) {
- currentWord.currentPronunciation.append(",");
+
+ final String lang = trim(WikiTokenizer.toPlainText(line.substring(0, colonIndex)));
+ final boolean appendLang;
+ if (wikiTokenizer.listItemPrefix().length() == 1) {
+ topLevelLang = lang;
+ final boolean thisFind = langPattern.matcher(lang).find();
+ if (!thisFind) {
+ continue;
+ }
+ appendLang = !langPattern.matcher(lang).matches();
+ } else if (topLevelLang == null) {
+ continue;
+ } else {
+ // Two-level -- the only way we won't append is if this second level matches exactly.
+ if (!langPattern.matcher(lang).matches() && !langPattern.matcher(topLevelLang).find()) {
+ continue;
}
- final String pron = wikiMarkup.matcher(positionalArgs.get(1)).replaceAll("");
- currentWord.currentPronunciation.append(pron).append("");
+ appendLang = !langPattern.matcher(lang).matches();
}
- currentWord.currentPronunciation.append(" (").append(name).append(")");
- return;
- }
-
- if (name.equals("qualifier")) {
- //assert positionalArgs.size() == 2 && namedArgs.isEmpty() : positionalArgs.toString() + namedArgs.toString();
- if (wikiBuilder == null) {
- return;
+
+ String rest = line.substring(colonIndex + 1).trim();
+ if (rest.length() > 0) {
+ doTranslationLine(line, appendLang ? lang : null, title, pos, sense, rest);
+ }
+
+ } else if (wikiTokenizer.remainderStartsWith("''See''")) {
+ wikiTokenizer.nextLine();
+ LOG.fine("Skipping line: " + wikiTokenizer.token());
+ } else if (wikiTokenizer.isWikiLink()) {
+ final String wikiLink = wikiTokenizer.wikiLinkText();
+ if (wikiLink.contains(":") && wikiLink.contains(title)) {
+ } else if (wikiLink.contains("Category:")) {
+ } else {
+ LOG.warning("Unexpected wikiLink: " + wikiTokenizer.token() + ", title=" + title);
+ }
+ } else if (wikiTokenizer.isNewline() || wikiTokenizer.isMarkup() || wikiTokenizer.isComment()) {
+ } else {
+ final String token = wikiTokenizer.token();
+ if (token.equals("----")) {
+ } else {
+ LOG.warning("Unexpected translation token: " + wikiTokenizer.token() + ", title=" + title);
}
- wikiBuilder.append(" (").append(positionalArgs.get(1)).append(")");
- return;
- }
-
- if (name.equals("...")) {
- // Skipping any elided text for brevity.
- wikiBuilder.append("...");
- return;
- }
-
- if (passThroughTemplates.contains(name)) {
- assert positionalArgs.size() == 1 && namedArgs.isEmpty() : positionalArgs.toString() + namedArgs;
- wikiBuilder.append(name);
- return;
- }
-
- if (name.equals("audio") || name.equals("rhymes") || name.equals("hyphenation")) {
- return;
}
- if ("Pronunciation".equals(currentHeading)) {
- System.err.println("Unhandled pronunciation template: " + positionalArgs + namedArgs);
- return;
- }
- } // Pronunciation
+ }
+ }
+
+ private static <T> T get(final List<T> list, final int index, final T defaultValue) {
+ return index < list.size() ? list.get(index) : defaultValue;
+ }
+
+ private static <T> T get(final List<T> list, final int index) {
+ return get(list, index, null);
+ }
+
+ private static <T> T remove(final List<T> list, final int index, final T defaultValue) {
+ return index < list.size() ? list.remove(index) : defaultValue;
+ }
+
+ private void doTranslationLine(final String line, final String lang, final String title, final String pos, final String sense, final String rest) {
+ // Good chance we'll actually file this one...
+ final PairEntry pairEntry = new PairEntry();
+ final IndexedEntry indexedEntry = new IndexedEntry(pairEntry);
- // Part of speech
- if (insidePartOfSpeech) {
+ final StringBuilder otherText = new StringBuilder();
+ final WikiTokenizer wikiTokenizer = new WikiTokenizer(rest, false);
+ while (wikiTokenizer.nextToken() != null) {
- // form of
- if (name.equals("form of")) {
- namedArgs.remove("sc");
- if (positionalArgs.size() < 3 || positionalArgs.size() > 4) {
- System.err.println("Invalid form of.");
- }
- final String token = positionalArgs.get(positionalArgs.size() == 3 ? 2 : 3);
- final String grammarForm = WikiParser.simpleParse(positionalArgs.get(1));
- currentPartOfSpeech.formOfs.add(new FormOf(grammarForm, token));
- return;
- }
-
- // The fallback plan: append the template!
- if (wikiBuilder != null) {
- wikiBuilder.append("{");
- boolean first = true;
- for (final String arg : positionalArgs) {
- if (!first) {
- wikiBuilder.append(", ");
+ if (wikiTokenizer.isPlainText()) {
+ final String plainText = wikiTokenizer.token();
+ otherText.append("").append(plainText);
+ otherIndexBuilder.addEntryWithString(indexedEntry, plainText, EntryTypeName.WIKTIONARY_TRANSLATION_OTHER_TEXT);
+
+ } else if (wikiTokenizer.isWikiLink()) {
+ final String plainText = wikiTokenizer.wikiLinkText();
+ otherText.append("").append(plainText);
+ otherIndexBuilder.addEntryWithString(indexedEntry, plainText, EntryTypeName.WIKTIONARY_TRANSLATION_WIKI_TEXT);
+
+ } else if (wikiTokenizer.isFunction()) {
+ final String functionName = wikiTokenizer.functionName();
+ final List<String> args = wikiTokenizer.functionPositionArgs();
+ final Map<String,String> namedArgs = wikiTokenizer.functionNamedArgs();
+
+ if (functionName.equals("t") || functionName.equals("t+") || functionName.equals("t-") || functionName.equals("tø") || functionName.equals("apdx-t")) {
+ if (args.size() < 2) {
+ LOG.warning("{{t}} with too few args: " + line + ", title=" + title);
+ continue;
}
- first = false;
- wikiBuilder.append(arg);
- }
- // This one isn't so useful.
- for (final Map.Entry<String, String> entry : namedArgs.entrySet()) {
- if (!first) {
- wikiBuilder.append(", ");
+ final String langCode = get(args, 0);
+ //if (this.langCodePattern.matcher(langCode).matches()) {
+ final String word = get(args, 1);
+ final String gender = get(args, 2);
+ final String transliteration = namedArgs.get("tr");
+ if (otherText.length() > 0) {
+ otherText.append("");
+ }
+ otherText.append(word);
+ otherIndexBuilder.addEntryWithString(indexedEntry, word, EntryTypeName.WIKTIONARY_TITLE_SINGLE, EntryTypeName.WIKTIONARY_TITLE_MULTI);
+ if (gender != null) {
+ otherText.append(String.format(" {%s}", gender));
+ }
+ if (transliteration != null) {
+ otherText.append(String.format(TRANSLITERATION_FORMAT, transliteration));
+ otherIndexBuilder.addEntryWithString(indexedEntry, transliteration, EntryTypeName.WIKTIONARY_TRANSLITERATION);
+ }
+ //}
+ } else if (functionName.equals("qualifier")) {
+ if (args.size() == 0) {
+ otherText.append(wikiTokenizer.token());
+ } else {
+ String qualifier = args.get(0);
+ if (!namedArgs.isEmpty() || args.size() > 1) {
+ LOG.warning("weird qualifier: " + line);
+ }
+ // Unindexed!
+ otherText.append("(").append(qualifier).append(")");
}
- first = false;
- wikiBuilder.append(entry.getKey()).append("=").append(entry.getValue());
+ } else if (encodings.contains(functionName)) {
+ otherText.append("").append(args.get(0));
+ otherIndexBuilder.addEntryWithString(indexedEntry, args.get(0), EntryTypeName.WIKTIONARY_TRANSLATION_OTHER_TEXT);
+ } else if (isGender(functionName)) {
+ appendGender(otherText, functionName, args);
+ } else if (functionName.equals("g")) {
+ otherText.append("{g}");
+ } else if (functionName.equals("l")) {
+ // encodes text in various langs.
+ // lang is arg 0.
+ otherText.append("").append(args.get(1));
+ otherIndexBuilder.addEntryWithString(indexedEntry, args.get(1), EntryTypeName.WIKTIONARY_TRANSLATION_OTHER_TEXT);
+ // TODO: transliteration
+ } else if (functionName.equals("term")) {
+ // cross-reference to another dictionary
+ otherText.append("").append(args.get(0));
+ otherIndexBuilder.addEntryWithString(indexedEntry, args.get(0), EntryTypeName.WIKTIONARY_TRANSLATION_OTHER_TEXT);
+ // TODO: transliteration
+ } else if (functionName.equals("italbrac") || functionName.equals("gloss")) {
+ // TODO: put this text aside to use it.
+ otherText.append("[").append(args.get(0)).append("]");
+ otherIndexBuilder.addEntryWithString(indexedEntry, args.get(0), EntryTypeName.WIKTIONARY_TRANSLATION_OTHER_TEXT);
+ } else if (functionName.equals("ttbc")) {
+ LOG.warning("Unexpected {{ttbc}}");
+ } else if (functionName.equals("trreq")) {
+ } else if (functionName.equals("not used")) {
+ otherText.append("(not used)");
+ } else if (functionName.equals("t-image")) {
+ // American sign language
+ } else {
+ // Unindexed!
+ namedArgs.keySet().removeAll(USELESS_WIKI_ARGS);
+ WikiTokenizer.appendFunction(otherText.append("{{"), functionName, args, namedArgs).append("}}");
}
- wikiBuilder.append("}");
+
+ } else if (wikiTokenizer.isNewline()) {
+ assert false;
+ } else if (wikiTokenizer.isComment()) {
+ } else if (wikiTokenizer.isMarkup()) {
+ } else {
+ LOG.warning("Bad translation token: " + wikiTokenizer.token());
}
-
- //System.err.println("Unhandled part of speech template: " + positionalArgs + namedArgs);
+ }
+ if (otherText.length() == 0) {
+ LOG.warning("Empty otherText: " + line);
return;
- } // Part of speech
-
+ }
- // Translations
- if (name.equals("trans-top")) {
- assert positionalArgs.size() >= 1 && namedArgs.isEmpty() : positionalArgs.toString() + namedArgs + title;
-
- if (currentPartOfSpeech == null) {
- assert currentWord != null && !currentWord.partsOfSpeech.isEmpty() : title;
- System.err.println("Assuming last part of speech for non-nested translation section: " + title);
- currentPartOfSpeech = ListUtil.getLast(currentWord.partsOfSpeech);
- }
-
- currentTranslationSense = new WikiWord.TranslationSense();
- currentPartOfSpeech.translationSenses.add(currentTranslationSense);
- if (positionalArgs.size() > 1) {
- currentTranslationSense.sense = positionalArgs.get(1);
- }
- return;
- } // Translations
+ if (lang != null) {
+ otherText.insert(0, String.format("(%s) ", lang));
+ }
+
+ StringBuilder englishText = new StringBuilder();
+
+ englishText.append(title);
+ if (sense != null) {
+ englishText.append(" (").append(sense).append(")");
+ enIndexBuilder.addEntryWithString(indexedEntry, sense, EntryTypeName.WIKTIONARY_TRANSLATION_SENSE, EntryTypeName.WIKTIONARY_TRANSLATION_SENSE);
+ }
+ if (pos != null) {
+ englishText.append(" (").append(pos.toLowerCase()).append(")");
+ }
+ enIndexBuilder.addEntryWithString(indexedEntry, title, EntryTypeName.WIKTIONARY_TITLE_SINGLE, EntryTypeName.WIKTIONARY_TITLE_MULTI);
+
+ final Pair pair = new Pair(trim(englishText.toString()), trim(otherText.toString()), swap);
+ pairEntry.pairs.add(pair);
+ if (!pairsAdded.add(pair.toString())) {
+ LOG.warning("Duplicate pair: " + pair.toString());
+ }
+ if (pair.toString().equals("libero {m} :: free (adjective)")) {
+ System.out.println();
+ }
- if (wikiBuilder == null) {
- return;
- }
- if (name.equals("m") || name.equals("f") || name.equals("n") || name.equals("c")) {
- assert positionalArgs.size() >= 1 && namedArgs.isEmpty() : positionalArgs.toString() + namedArgs.toString();
- wikiBuilder.append("{");
- for (int i = 1; i < positionalArgs.size(); ++i) {
- wikiBuilder.append(i > 1 ? "," : "");
- wikiBuilder.append(positionalArgs.get(i));
- }
- wikiBuilder.append(name).append("}");
-
- } else if (name.equals("p")) {
- assert positionalArgs.size() == 1 && namedArgs.isEmpty();
- wikiBuilder.append("pl.");
+ }
- } else if (name.equals("s")) {
- assert positionalArgs.size() == 1 && namedArgs.isEmpty() || title.equals("dobra");
- wikiBuilder.append("sg.");
-
- } else if (grammarTemplates.contains(name)) {
- assert positionalArgs.size() == 1 && namedArgs.isEmpty();
- wikiBuilder.append(name).append(".");
- } else if (name.equals("l")) {
- // This template is designed to generate a link to a specific language-section on the target page.
- wikiBuilder.append(positionalArgs.size() >= 4 ? positionalArgs.get(3) : positionalArgs.get(2));
-
- } else if (name.equals("t") || name.equals("t+") || name.equals("t-") || name.equals("tø")) {
- if (positionalArgs.size() > 2) {
- wikiBuilder.append(positionalArgs.get(2));
- }
- for (int i = 3; i < positionalArgs.size(); ++i) {
- wikiBuilder.append(i == 3 ? " {" : ",");
- wikiBuilder.append(positionalArgs.get(i));
- wikiBuilder.append(i == positionalArgs.size() - 1 ? "}" : "");
- }
- final String transliteration = namedArgs.remove("tr");
- if (transliteration != null) {
- wikiBuilder.append(" (").append(transliteration).append(")");
- }
-
- } else if (name.equals("trreq")) {
- wikiBuilder.append("{{trreq}}");
-
- } else if (name.equals("qualifier")) {
- //assert positionalArgs.size() == 2 && namedArgs.isEmpty() : positionalArgs.toString() + namedArgs.toString();
- wikiBuilder.append(" (").append(positionalArgs.get(1)).append(")");
-
- } else if (useRemainingArgTemplates.contains(name)) {
- for (int i = 1; i < positionalArgs.size(); ++i) {
- if (i != 1) {
- wikiBuilder.append(", ");
+ private void appendGender(final StringBuilder otherText,
+ final String functionName, final List<String> args) {
+ otherText.append("{");
+ otherText.append(functionName);
+ for (int i = 0; i < args.size(); ++i) {
+ otherText.append("|").append(args.get(i));
+ }
+ otherText.append("}");
+ }
+
+
+ private boolean isGender(final String functionName) {
+ return functionName.equals("m") || functionName.equals("f") || functionName.equals("n") || functionName.equals("p");
+ }
+
+ Set<String> pairsAdded = new LinkedHashSet<String>();
+
+ // -------------------------------------------------------------------------
+
+ private void doForeignWord(final String lang, final String title, final String text) {
+ final WikiTokenizer wikiTokenizer = new WikiTokenizer(text);
+ while (wikiTokenizer.nextToken() != null) {
+ if (wikiTokenizer.isHeading()) {
+ final String headingName = wikiTokenizer.headingWikiText();
+ if (headingName.equals("Translations")) {
+ LOG.warning("Translations not in English section: " + title);
+ } else if (headingName.equals("Pronunciation")) {
+ //doPronunciation(wikiLineReader);
+ } else if (partOfSpeechHeader.matcher(headingName).matches()) {
+ doForeignPartOfSpeech(lang, title, headingName, wikiTokenizer.headingDepth(), wikiTokenizer);
}
- wikiBuilder.append(positionalArgs.get(i));
- }
- } else if (ignoreTemplates.contains(name)) {
- // Do nothing.
-
- } else if (name.equals("initialism")) {
- assert positionalArgs.size() <= 2 && namedArgs.isEmpty() : positionalArgs.toString() + namedArgs;
- wikiBuilder.append("Initialism");
- } else if (name.equals("abbreviation")) {
- assert positionalArgs.size() <= 2 && namedArgs.isEmpty() : positionalArgs.toString() + namedArgs;
- wikiBuilder.append("Abbreviation");
- } else if (name.equals("acronym")) {
- assert positionalArgs.size() <= 2 && namedArgs.isEmpty() : positionalArgs.toString() + namedArgs;
- wikiBuilder.append("Acronym");
- } else {
- if (currentTranslationSense != null) {
- System.err.println("Unhandled template: " + positionalArgs.toString() + namedArgs);
+ } else {
}
}
}
+
+ static final class ListSection {
+ final String firstPrefix;
+ final String firstLine;
+ final List<String> nextPrefixes = new ArrayList<String>();
+ final List<String> nextLines = new ArrayList<String>();
+
+ public ListSection(String firstPrefix, String firstLine) {
+ this.firstPrefix = firstPrefix;
+ this.firstLine = firstLine;
+ }
- @Override
- public void onText(String text) {
- if (wikiBuilder != null) {
- wikiBuilder.append(text);
- return;
+ @Override
+ public String toString() {
+ return firstPrefix + firstLine + "{ " + nextPrefixes + "}";
}
}
- @Override
- public void onHeadingStart(int depth) {
- wikiBuilder = new StringBuilder();
- currentDepth = depth;
- if (currentPartOfSpeech != null && depth <= currentPartOfSpeech.depth) {
- currentPartOfSpeech = null;
- insidePartOfSpeech = false;
+
+ int foreignCount = 0;
+ private void doForeignPartOfSpeech(final String lang, String title, final String posHeading, final int posDepth, WikiTokenizer wikiTokenizer) {
+ if (++foreignCount % 1000 == 0) {
+ LOG.info("***" + lang + ", " + title + ", pos=" + posHeading + ", foreignCount=" + foreignCount);
}
- if (currentWord != null && depth <= currentWord.depth) {
- currentWord = null;
+ if (title.equals("moro")) {
+ System.out.println();
}
- currentHeading = null;
- }
-
- @Override
- public void onHeadingEnd(int depth) {
- final String name = wikiBuilder.toString().trim();
- wikiBuilder = null;
- currentTranslationSense = null;
- currentHeading = name;
+ boolean titleAppended = false;
+ final StringBuilder foreignBuilder = new StringBuilder();
+ final Collection<String> wordForms = new ArrayList<String>();
+ final List<ListSection> listSections = new ArrayList<ListSection>();
- final boolean lang0 = langPatterns[0].matcher(name).matches();
- final boolean lang1 = langPatterns[1].matcher(name).matches();
- if (name.equalsIgnoreCase("English") || lang0 || lang1 || name.equalsIgnoreCase("Translingual")) {
- currentWord = new WikiWord(title, depth);
- if (lang0 && lang1) {
- System.err.println("Word is indexed in both index1 and index2: " + title);
- }
- currentWord.language = name;
- currentWord.index = lang0 ? 0 : (lang1 ? 1 : -1);
- words.add(currentWord);
- return;
- }
+ try {
- if (currentWord == null) {
- return;
- }
+ ListSection lastListSection = null;
- if (currentPartOfSpeech != null && depth <= currentPartOfSpeech.depth) {
- currentPartOfSpeech = null;
- }
+ int currentHeadingDepth = posDepth;
+ while (wikiTokenizer.nextToken() != null) {
+ if (wikiTokenizer.isHeading()) {
+ currentHeadingDepth = wikiTokenizer.headingDepth();
+
+ if (currentHeadingDepth <= posDepth) {
+ wikiTokenizer.returnToLineStart();
+ return;
+ }
+ }
+
+ if (currentHeadingDepth > posDepth) {
+ // TODO: deal with other neat info sections
+ continue;
+ }
+
+ if (wikiTokenizer.isFunction()) {
+ final String name = wikiTokenizer.functionName();
+ final List<String> args = wikiTokenizer.functionPositionArgs();
+ final Map<String,String> namedArgs = wikiTokenizer.functionNamedArgs();
+ // First line is generally a repeat of the title with some extra information.
+ // We need to build up the left side (foreign text, tokens) separately from the
+ // right side (English). The left-side may get paired with multiple right sides.
+ // The left side should get filed under every form of the word in question (singular, plural).
+
+ // For verbs, the conjugation comes later on in a deeper section.
+ // Ideally, we'd want to file every English entry with the verb
+ // under every verb form coming from the conjugation.
+ // Ie. under "fa": see: "make :: fare" and "do :: fare"
+ // But then where should we put the conjugation table?
+ // 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!)
+ // for the conjugation table from "fa".
+ // Would like to be able to link to a lang#token.
+ if (isGender(name)) {
+ appendGender(foreignBuilder, name, args);
+ } else if (name.equals("wikipedia")) {
+ namedArgs.remove("lang");
+ if (args.size() > 1 || !namedArgs.isEmpty()) {
+ // Unindexed!
+ foreignBuilder.append(wikiTokenizer.token());
+ } else if (args.size() == 1) {
+ foreignBuilder.append(wikiTokenizer.token());
+ } else {
+ //foreignBuilder.append(title);
+ }
+ } else if (name.equals("attention") || name.equals("zh-attention")) {
+ // See: http://en.wiktionary.org/wiki/Template:attention
+ // Ignore these.
+ } else if (name.equals("infl")) {
+ // See: http://en.wiktionary.org/wiki/Template:infl
+ final String langCode = get(args, 0);
+ String head = namedArgs.remove("head");
+ if (head == null) {
+ head = namedArgs.remove("title"); // Bug
+ }
+ if (head == null) {
+ head = title;
+ } else {
+ head = WikiTokenizer.toPlainText(head);
+ }
+ titleAppended = true;
+
+ namedArgs.keySet().removeAll(USELESS_WIKI_ARGS);
+
+ final String tr = namedArgs.remove("tr");
+ String g = namedArgs.remove("g");
+ if (g == null) {
+ g = namedArgs.remove("gender");
+ }
+ final String g2 = namedArgs.remove("g2");
+ final String g3 = namedArgs.remove("g3");
+
+ foreignBuilder.append(head);
- insidePartOfSpeech = false;
- if (currentPartOfSpeech == null && partOfSpeechHeader.matcher(name).matches()) {
- currentPartOfSpeech = new WikiWord.PartOfSpeech(depth, name);
- currentWord.partsOfSpeech.add(currentPartOfSpeech);
- insidePartOfSpeech = true;
- return;
- }
+ if (g != null) {
+ foreignBuilder.append(" {").append(g);
+ if (g2 != null) {
+ foreignBuilder.append("|").append(g2);
+ }
+ if (g3 != null) {
+ foreignBuilder.append("|").append(g3);
+ }
+ foreignBuilder.append("}");
+ }
- if (name.equals("Translations")) {
- if (currentWord == null ||
- !currentWord.language.equals("English") ||
- currentPartOfSpeech == null) {
- System.err.println("Unexpected Translations section: " + title);
- return;
+ if (tr != null) {
+ foreignBuilder.append(String.format(TRANSLITERATION_FORMAT, tr));
+ wordForms.add(tr);
+ }
+
+ final String pos = get(args, 1);
+ if (pos != null) {
+ foreignBuilder.append(" (").append(pos).append(")");
+ }
+ for (int i = 2; i < args.size(); i += 2) {
+ final String inflName = get(args, i);
+ final String inflValue = get(args, i + 1);
+ foreignBuilder.append(", ").append(WikiTokenizer.toPlainText(inflName));
+ if (inflValue != null && inflValue.length() > 0) {
+ foreignBuilder.append(": ").append(WikiTokenizer.toPlainText(inflValue));
+ wordForms.add(inflValue);
+ }
+ }
+ for (final String key : namedArgs.keySet()) {
+ final String value = WikiTokenizer.toPlainText(namedArgs.get(key));
+ foreignBuilder.append(" ").append(key).append("=").append(value);
+ wordForms.add(value);
+ }
+ } else if (name.equals("it-noun")) {
+ titleAppended = true;
+ final String base = get(args, 0);
+ final String gender = get(args, 1);
+ final String singular = base + get(args, 2);
+ final String plural = base + get(args, 3);
+ foreignBuilder.append(String.format(" %s {%s}, %s {pl}", singular, gender, plural, plural));
+ wordForms.add(singular);
+ wordForms.add(plural);
+ if (!namedArgs.isEmpty() || args.size() > 4) {
+ LOG.warning("Invalid it-noun: " + wikiTokenizer.token());
+ }
+ } else if (name.equals("it-proper noun")) {
+ foreignBuilder.append(wikiTokenizer.token());
+ } else if (name.equals("it-adj")) {
+ foreignBuilder.append(wikiTokenizer.token());
+ } else if (name.startsWith("it-conj")) {
+ if (name.equals("it-conj-are")) {
+ itConjAre(args, namedArgs);
+ } else if (name.equals("it-conj-ere")) {
+ } else if (name.equals("it-conj-ire")) {
+ } else {
+ LOG.warning("Unknown conjugation: " + wikiTokenizer.token());
+ }
+ } else {
+ // Unindexed!
+ foreignBuilder.append(wikiTokenizer.token());
+ // LOG.warning("Unknown function: " + wikiTokenizer.token());
+ }
+ } else if (wikiTokenizer.isListItem()) {
+ final String prefix = wikiTokenizer.listItemPrefix();
+ if (lastListSection != null &&
+ prefix.startsWith(lastListSection.firstPrefix) &&
+ prefix.length() > lastListSection.firstPrefix.length()) {
+ lastListSection.nextPrefixes.add(prefix);
+ lastListSection.nextLines.add(wikiTokenizer.listItemWikiText());
+ } else {
+ lastListSection = new ListSection(prefix, wikiTokenizer.listItemWikiText());
+ listSections.add(lastListSection);
+ }
+ } else if (lastListSection != null) {
+ // Don't append anything after the lists, because there's crap.
+ } else if (wikiTokenizer.isWikiLink()) {
+ // Unindexed!
+ foreignBuilder.append(wikiTokenizer.wikiLinkText());
+
+ } else if (wikiTokenizer.isPlainText()) {
+ // Unindexed!
+ foreignBuilder.append(wikiTokenizer.token());
+
+ } else if (wikiTokenizer.isMarkup() || wikiTokenizer.isNewline() || wikiTokenizer.isComment()) {
+ // Do nothing.
+ } else {
+ LOG.warning("Unexpected token: " + wikiTokenizer.token());
}
- currentTranslationSense = new WikiWord.TranslationSense();
}
- }
-
- @Override
- public void onListItemStart(String header, int[] section) {
- wikiBuilder = new StringBuilder();
- if (currentWord != null) {
- currentWord.currentPronunciation = null;
+ } finally {
+ // Here's where we exit.
+ // Should we make an entry even if there are no foreign list items?
+ String foreign = foreignBuilder.toString().trim();
+ if (!titleAppended && !foreign.toLowerCase().startsWith(title.toLowerCase())) {
+ foreign = String.format("%s %s", title, foreign);
+ }
+ if (!langPattern.matcher(lang).matches()) {
+ foreign = String.format("(%s) %s", lang, foreign);
+ }
+ for (final ListSection listSection : listSections) {
+ doForeignListItem(foreign, title, wordForms, listSection);
+ }
}
}
+
+ static final Pattern UNINDEXED_WIKI_TEXT = Pattern.compile(
+ "(first|second|third)-person (singular|plural)|" +
+ "present tense|" +
+ "imperative"
+ );
- @Override
- public void onListItemEnd(String header, int[] section) {
- String item = wikiBuilder.toString().trim();
- final String oldItem = item;
- if (item.length() == 0) {
+ // Might only want to remove "lang" if it's equal to "zh", for example.
+ static final Set<String> USELESS_WIKI_ARGS = new LinkedHashSet<String>(Arrays.asList("lang", "sc", "sort", "cat"));
+
+ private void doForeignListItem(final String foreignText, String title, final Collection<String> forms, final ListSection listSection) {
+
+ final String prefix = listSection.firstPrefix;
+ if (prefix.length() > 1) {
+ // Could just get looser and say that any prefix longer than first is a sublist.
+ LOG.warning("Prefix too long: " + listSection);
return;
}
- item = WikiParser.simpleParse(item);
- wikiBuilder = null;
-
- // Part of speech
- if (insidePartOfSpeech) {
- assert currentPartOfSpeech != null : title + item;
- if (header.equals("#") ||
- header.equals("##") ||
- header.equals("###") ||
- header.equals("####") ||
- header.equals(":#") ||
- header.equals("::") ||
- header.equals(":::*")) {
- // Definition.
- // :: should append, probably.
- currentPartOfSpeech.newMeaning().meaning = item;
-
- // Source
- } else if (header.equals("#*") ||
- header.equals("##*") ||
- header.equals("###*")) {
- currentPartOfSpeech.lastMeaning().newExample().source = item;
-
- // Example
- } else if (header.equals("#:") ||
- header.equals("#*:") ||
- header.equals("#:*") ||
- header.equals("##:") ||
- header.equals("##*:") ||
- header.equals("#:*:") ||
- header.equals("#:*#") ||
- header.equals("#*:") ||
- header.equals("*:") ||
- header.equals("#:::") ||
- header.equals("#**") ||
- header.equals("#*:::") ||
- header.equals("#:#") ||
- header.equals(":::") ||
- header.equals("##:*") ||
- header.equals("###*:")) {
- StringUtil.appendLine(currentPartOfSpeech.lastMeaning().newExample().example, item);
-
- // Example in English
- } else if (header.equals("#::") ||
- header.equals("#*::") ||
- header.equals("#:**") ||
- header.equals("#*#") ||
- header.equals("##*::")) {
- StringUtil.appendLine(currentPartOfSpeech.lastMeaning().lastExample().exampleInEnglish, item);
+
+ final PairEntry pairEntry = new PairEntry();
+ final IndexedEntry indexedEntry = new IndexedEntry(pairEntry);
+
+ final StringBuilder englishBuilder = new StringBuilder();
+
+ final String mainLine = listSection.firstLine;
+ final WikiTokenizer englishTokenizer = new WikiTokenizer(mainLine, false);
+ while (englishTokenizer.nextToken() != null) {
+ // TODO handle form of....
+ if (englishTokenizer.isPlainText()) {
+ englishBuilder.append(englishTokenizer.token());
+ enIndexBuilder.addEntryWithString(indexedEntry, englishTokenizer.token(), EntryTypeName.WIKTIONARY_ENGLISH_DEF);
+ } else if (englishTokenizer.isWikiLink()) {
+ final String text = englishTokenizer.wikiLinkText();
+ final String link = englishTokenizer.wikiLinkDest();
+ if (link != null) {
+ if (link.contains("#English")) {
+ englishBuilder.append(text);
+ enIndexBuilder.addEntryWithString(indexedEntry, text, EntryTypeName.WIKTIONARY_ENGLISH_DEF_WIKI_LINK);
+ } else if (link.contains("#") && this.langPattern.matcher(link).find()) {
+ englishBuilder.append(text);
+ otherIndexBuilder.addEntryWithString(indexedEntry, text, EntryTypeName.WIKTIONARY_ENGLISH_DEF_OTHER_LANG);
+ } else if (link.equals("plural")) {
+ englishBuilder.append(text);
+ } else {
+ //LOG.warning("Special link: " + englishTokenizer.token());
+ enIndexBuilder.addEntryWithString(indexedEntry, text, EntryTypeName.WIKTIONARY_ENGLISH_DEF_WIKI_LINK);
+ englishBuilder.append(text);
+ }
+ } else {
+ // link == null
+ englishBuilder.append(text);
+ if (!UNINDEXED_WIKI_TEXT.matcher(text).find()) {
+ enIndexBuilder.addEntryWithString(indexedEntry, text, EntryTypeName.WIKTIONARY_ENGLISH_DEF_WIKI_LINK);
+ }
+ }
+ } else if (englishTokenizer.isFunction()) {
+ final String name = englishTokenizer.functionName();
+ final List<String> args = englishTokenizer.functionPositionArgs();
+ final Map<String,String> namedArgs = englishTokenizer.functionNamedArgs();
- // Skip
- } else if (header.equals("*") ||
- header.equals("**") ||
- header.equals("***") ||
- header.equals("*#") ||
- header.equals(":") ||
- header.equals("::*") ||
- header.equals("#**") ||
- header.equals(":*") ||
- header.equals("#*:*") ||
- header.equals("#*:**") ||
- header.equals("#*:#") ||
- header.equals("#*:*:") ||
- header.equals("#*:*") ||
- header.equals(";")) {
- // might have: * {{seeCites}}
- // * [[w:Arabic numerals|Arabic numerals]]: 2
- //assert item.trim().length() == 0;
- System.err.println("Skipping meaning: " + header + " " + item);
+ if (
+ name.equals("form of") ||
+ name.contains("conjugation of") ||
+ name.contains("participle of") ||
+ name.contains("gerund of") ||
+ name.contains("feminine of") ||
+ name.contains("plural of")) {
+ String formName = name;
+ if (name.equals("form of")) {
+ formName = args.remove(0);
+ }
+ String baseForm = get(args, 1, "");
+ if ("".equals(baseForm)) {
+ baseForm = get(args, 0, null);
+ remove(args, 1, "");
+ } else {
+ remove(args, 0, null);
+ }
+ namedArgs.keySet().removeAll(USELESS_WIKI_ARGS);
+ WikiTokenizer.appendFunction(englishBuilder.append("{"), formName, args, namedArgs).append("}");
+ otherIndexBuilder.addEntryWithString(indexedEntry, baseForm, EntryTypeName.WIKTIONARY_BASE_FORM_SINGLE, EntryTypeName.WIKTIONARY_BASE_FORM_MULTI);
+
+ } else {
+ namedArgs.keySet().removeAll(USELESS_WIKI_ARGS);
+ WikiTokenizer.appendFunction(englishBuilder.append("{{"), name, args, namedArgs).append("}}");
+// LOG.warning("Unexpected function: " + englishTokenizer.token());
+ }
} else {
- if (title.equals("Yellowknife")) {
- return;
+ if (englishTokenizer.isComment() || englishTokenizer.isMarkup()) {
+ } else {
+ LOG.warning("Unexpected definition type: " + englishTokenizer.token());
}
- System.err.println("Busted heading: " + title + " "+ header + " " + item);
}
- return;
}
- // Part of speech
+
+ final String english = trim(englishBuilder.toString());
+ if (english.length() > 0) {
+ final Pair pair = new Pair(english, trim(foreignText), this.swap);
+ pairEntry.pairs.add(pair);
+ otherIndexBuilder.addEntryWithString(indexedEntry, title, EntryTypeName.WIKTIONARY_TITLE_SINGLE, EntryTypeName.WIKTIONARY_TITLE_MULTI);
+ for (final String form : forms) {
+ otherIndexBuilder.addEntryWithString(indexedEntry, form, EntryTypeName.WIKTIONARY_INFLECTD_FORM_SINGLE, EntryTypeName.WIKTIONARY_INFLECTED_FORM_MULTI);
+ }
+ }
- // Translation
- if (currentTranslationSense != null) {
- if (item.indexOf("{{[trreq]{}}}") != -1) {
- return;
+ // Do examples.
+ String lastForeign = null;
+ for (int i = 0; i < listSection.nextPrefixes.size(); ++i) {
+ final String nextPrefix = listSection.nextPrefixes.get(i);
+ final String nextLine = listSection.nextLines.get(i);
+ int dash = nextLine.indexOf("—");
+ int mdashLen = 7;
+ if (dash == -1) {
+ dash = nextLine.indexOf("—");
+ mdashLen = 1;
}
-
- if (currentPartOfSpeech.translationSenses.isEmpty()) {
- currentPartOfSpeech.translationSenses.add(currentTranslationSense);
+ if (dash == -1) {
+ dash = nextLine.indexOf(" - ");
+ mdashLen = 3;
}
-
- final int colonPos = item.indexOf(':');
- if (colonPos == -1) {
- System.err.println("Invalid translation: title=" + title + ", item=" + item);
- return;
+
+ if ((nextPrefix.equals("#:") || nextPrefix.equals("##:")) && dash != -1) {
+ final String foreignEx = nextLine.substring(0, dash);
+ final String englishEx = nextLine.substring(dash + mdashLen);
+ final Pair pair = new Pair(formatAndIndexExampleString(englishEx, enIndexBuilder, indexedEntry), formatAndIndexExampleString(foreignEx, otherIndexBuilder, indexedEntry), swap);
+ if (pair.lang1 != "--" && pair.lang1 != "--") {
+ pairEntry.pairs.add(pair);
+ }
+ lastForeign = null;
+ } else if (nextPrefix.equals("#:") || nextPrefix.equals("##:")){
+ final Pair pair = new Pair("--", formatAndIndexExampleString(nextLine, null, indexedEntry), swap);
+ lastForeign = nextLine;
+ if (pair.lang1 != "--" && pair.lang1 != "--") {
+ pairEntry.pairs.add(pair);
+ }
+ } else if (nextPrefix.equals("#::") || nextPrefix.equals("#**")) {
+ if (lastForeign != null && pairEntry.pairs.size() > 0) {
+ pairEntry.pairs.remove(pairEntry.pairs.size() - 1);
+ final Pair pair = new Pair(formatAndIndexExampleString(nextLine, enIndexBuilder, indexedEntry), formatAndIndexExampleString(lastForeign, otherIndexBuilder, indexedEntry), swap);
+ if (pair.lang1 != "--" || pair.lang2 != "--") {
+ pairEntry.pairs.add(pair);
+ }
+ lastForeign = null;
+ } else {
+ LOG.warning("TODO: English example with no foreign: " + title + ", " + nextLine);
+ final Pair pair = new Pair("--", formatAndIndexExampleString(nextLine, null, indexedEntry), swap);
+ if (pair.lang1 != "--" || pair.lang2 != "--") {
+ pairEntry.pairs.add(pair);
+ }
+ }
+ } else if (nextPrefix.equals("#*")) {
+ // Can't really index these.
+ final Pair pair = new Pair("--", formatAndIndexExampleString(nextLine, null, indexedEntry), swap);
+ lastForeign = nextLine;
+ if (pair.lang1 != "--" || pair.lang2 != "--") {
+ pairEntry.pairs.add(pair);
+ }
+ } else if (nextPrefix.equals("#::*") || nextPrefix.equals("##") || nextPrefix.equals("#*:") || nextPrefix.equals("#:*") || true) {
+ final Pair pair = new Pair("--", formatAndIndexExampleString(nextLine, null, indexedEntry), swap);
+ if (pair.lang1 != "--" || pair.lang2 != "--") {
+ pairEntry.pairs.add(pair);
+ }
+// } else {
+// assert false;
}
- final String lang = item.substring(0, colonPos);
- final String trans = item.substring(colonPos + 1).trim();
- for (int i = 0; i < 2; ++i) {
- if (langPatterns[i].matcher(lang).find()) {
- currentTranslationSense.translations.get(i).add(new Translation(lang, trans));
+ }
+ }
+
+ private String formatAndIndexExampleString(final String example, final IndexBuilder indexBuilder, final IndexedEntry indexedEntry) {
+ final WikiTokenizer wikiTokenizer = new WikiTokenizer(example, false);
+ final StringBuilder builder = new StringBuilder();
+ boolean insideTripleQuotes = false;
+ while (wikiTokenizer.nextToken() != null) {
+ if (wikiTokenizer.isPlainText()) {
+ builder.append(wikiTokenizer.token());
+ if (indexBuilder != null) {
+ indexBuilder.addEntryWithString(indexedEntry, wikiTokenizer.token(), EntryTypeName.WIKTIONARY_EXAMPLE);
+ }
+ } else if (wikiTokenizer.isWikiLink()) {
+ final String text = wikiTokenizer.wikiLinkText().replaceAll("'", "");
+ builder.append(text);
+ if (indexBuilder != null) {
+ indexBuilder.addEntryWithString(indexedEntry, text, EntryTypeName.WIKTIONARY_EXAMPLE);
}
+ } else if (wikiTokenizer.isFunction()) {
+ builder.append(wikiTokenizer.token());
+ } else if (wikiTokenizer.isMarkup()) {
+ if (wikiTokenizer.token().equals("'''")) {
+ insideTripleQuotes = !insideTripleQuotes;
+ }
+ } else if (wikiTokenizer.isComment() || wikiTokenizer.isNewline()) {
+ // Do nothing.
+ } else {
+ LOG.warning("unexpected token: " + wikiTokenizer.token());
}
- } // Translation
+ }
+ final String result = trim(builder.toString());
+ return result.length() > 0 ? result : "--";
}
- @Override
- public void onNewLine() {
- }
- @Override
- public void onNewParagraph() {
+ private void itConjAre(List<String> args, Map<String, String> namedArgs) {
+ final String base = args.get(0);
+ final String aux = args.get(1);
+
+ putIfMissing(namedArgs, "inf", base + "are");
+ putIfMissing(namedArgs, "aux", aux);
+ putIfMissing(namedArgs, "ger", base + "ando");
+ putIfMissing(namedArgs, "presp", base + "ante");
+ putIfMissing(namedArgs, "pastp", base + "ato");
+ // Present
+ putIfMissing(namedArgs, "pres1s", base + "o");
+ putIfMissing(namedArgs, "pres2s", base + "i");
+ putIfMissing(namedArgs, "pres3s", base + "a");
+ putIfMissing(namedArgs, "pres1p", base + "iamo");
+ putIfMissing(namedArgs, "pres2p", base + "ate");
+ putIfMissing(namedArgs, "pres3p", base + "ano");
+ // Imperfect
+ putIfMissing(namedArgs, "imperf1s", base + "avo");
+ putIfMissing(namedArgs, "imperf2s", base + "avi");
+ putIfMissing(namedArgs, "imperf3s", base + "ava");
+ putIfMissing(namedArgs, "imperf1p", base + "avamo");
+ putIfMissing(namedArgs, "imperf2p", base + "avate");
+ putIfMissing(namedArgs, "imperf3p", base + "avano");
+ // Passato remoto
+ putIfMissing(namedArgs, "prem1s", base + "ai");
+ putIfMissing(namedArgs, "prem2s", base + "asti");
+ putIfMissing(namedArgs, "prem3s", base + "ò");
+ putIfMissing(namedArgs, "prem1p", base + "ammo");
+ putIfMissing(namedArgs, "prem2p", base + "aste");
+ putIfMissing(namedArgs, "prem3p", base + "arono");
+ // Future
+ putIfMissing(namedArgs, "fut1s", base + "erò");
+ putIfMissing(namedArgs, "fut2s", base + "erai");
+ putIfMissing(namedArgs, "fut3s", base + "erà");
+ putIfMissing(namedArgs, "fut1p", base + "eremo");
+ putIfMissing(namedArgs, "fut2p", base + "erete");
+ putIfMissing(namedArgs, "fut3p", base + "eranno");
+ // Conditional
+ putIfMissing(namedArgs, "cond1s", base + "erei");
+ putIfMissing(namedArgs, "cond2s", base + "eresti");
+ putIfMissing(namedArgs, "cond3s", base + "erebbe");
+ putIfMissing(namedArgs, "cond1p", base + "eremmo");
+ putIfMissing(namedArgs, "cond2p", base + "ereste");
+ putIfMissing(namedArgs, "cond3p", base + "erebbero");
+ // Subjunctive / congiuntivo
+ putIfMissing(namedArgs, "sub123s", base + "i");
+ putIfMissing(namedArgs, "sub1p", base + "iamo");
+ putIfMissing(namedArgs, "sub2p", base + "iate");
+ putIfMissing(namedArgs, "sub3p", base + "ino");
+ // Imperfect subjunctive
+ putIfMissing(namedArgs, "impsub12s", base + "assi");
+ putIfMissing(namedArgs, "impsub3s", base + "asse");
+ putIfMissing(namedArgs, "impsub1p", base + "assimo");
+ putIfMissing(namedArgs, "impsub2p", base + "aste");
+ putIfMissing(namedArgs, "impsub3p", base + "assero");
+ // Imperative
+ putIfMissing(namedArgs, "imp2s", base + "a");
+ putIfMissing(namedArgs, "imp3s", base + "i");
+ putIfMissing(namedArgs, "imp1p", base + "iamo");
+ putIfMissing(namedArgs, "imp2p", base + "ate");
+ putIfMissing(namedArgs, "imp3p", base + "ino");
+
+
+ itConj(args, namedArgs);
}
- // ----------------------------------------------------------------------
-
- @Override
- public void onComment(String text) {
- }
- @Override
- public void onFormatBold(boolean boldOn) {
+ private void itConj(List<String> args, Map<String, String> namedArgs) {
+ // TODO Auto-generated method stub
+
}
- @Override
- public void onFormatItalic(boolean italicOn) {
- }
- @Override
- public void onUnterminated(String start, String rest) {
- throw new RuntimeException(start + rest);
+ private static void putIfMissing(final Map<String, String> namedArgs, final String key,
+ final String value) {
+ final String oldValue = namedArgs.get(key);
+ if (oldValue == null || oldValue.length() == 0) {
+ namedArgs.put(key, value);
+ }
}
- @Override
- public void onInvalidHeaderEnd(String rest) {
- throw new RuntimeException(rest);
+
+ // TODO: check how ='' and =| are manifested....
+ // TODO: get this right in -are
+ private static void putOrNullify(final Map<String, String> namedArgs, final String key,
+ final String value) {
+ final String oldValue = namedArgs.get(key);
+ if (oldValue == null/* || oldValue.length() == 0*/) {
+ namedArgs.put(key, value);
+ } else {
+ if (oldValue.equals("''")) {
+ namedArgs.put(key, "");
+ }
+ }
}
+ static final Pattern whitespace = Pattern.compile("\\s+");
+ static String trim(final String s) {
+ return whitespace.matcher(s).replaceAll(" ").trim();
+ }
+
+
}
projects / DictionaryPC.git / blobdiff