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.engine;
17 import com.hughes.android.dictionary.parser.wiktionary.EnTranslationToTranslationParser;
18 import com.hughes.android.dictionary.parser.wiktionary.WholeSectionToHtmlParser;
19 import com.hughes.android.dictionary.parser.wiktionary.WiktionaryLangs;
21 import junit.framework.TestCase;
24 import java.util.ArrayList;
25 import java.util.Arrays;
26 import java.util.LinkedHashMap;
27 import java.util.LinkedHashSet;
28 import java.util.List;
32 public class DictionaryBuilderMain extends TestCase {
34 static final String INPUTS = "data/inputs/";
35 static final String STOPLISTS = "data/inputs/stoplists/";
36 static final String OUTPUTS = "data/outputs/";
38 // Build the non EN ones.
39 static final String[][] nonEnPairs = new String[][] {
58 {"AR", "TR" }, // Turkish
63 {"DE", "CA" }, // Catalan
64 {"DE", "CS" }, // Czech
65 {"DE", "EO" }, // Esperanto
69 {"DE", "HU" }, // Hungarian
72 {"DE", "LA" }, // Latin
73 {"DE", "NL" }, // Dutch
74 {"DE", "PL" }, // Polish
76 {"DE", "SV" }, // Swedish
77 {"DE", "TR" }, // Turkish
79 {"DE", "TA" }, // Tamil
81 {"ES", "RU" }, // Spanish-Russian
83 {"FR", "BG" }, // Bulgarian
84 {"FR", "CS" }, // Czech
90 {"FR", "NL" }, // Dutch
92 {"FR", "TR" }, // Turkish
96 {"IT", "EL" }, // Greek
101 {"IT", "LA" }, // Latin
102 {"IT", "LV" }, // Latvian
107 {"IT", "TR" }, // Turkish
126 {"PL", "FR" }, // Polish
127 {"PL", "RU" }, // Polish
128 {"PL", "HU" }, // Polish
129 {"PL", "ES" }, // Polish
131 {"TR", "EL" }, // Turkish, Greek
133 {"FA", "HY" }, // Persian, Armenian, by request.
134 {"FA", "SV" }, // Persian, Swedish, by request.
140 static final Map<String,String> isoToDedication = new LinkedHashMap<String, String>();
142 isoToDedication.put("AF", "Afrikaans dictionary dedicated to Heiko and Mariëtte Horn.");
143 isoToDedication.put("HR", "Croatian dictionary dedicated to Ines Viskic and Miro Kresonja.");
144 isoToDedication.put("NL", "Dutch dictionary dedicated to Mike LeBeau.");
145 // German handled in file.
146 isoToDedication.put("EL", "Greek dictionary dedicated to Noah Egge.");
147 isoToDedication.put("IT", "Italian dictionary dedicated to Carolina Tropini, my favorite stardust in the whole universe! Ti amo!");
148 isoToDedication.put("KO", "Korean dictionary dedicated to Ande Elwood--fall fashion und Fernsehturms!");
149 isoToDedication.put("PT", "Portuguese dictionary dedicated to Carlos Melo, one Tough Mudder.");
150 isoToDedication.put("RO", "Romanian dictionary dedicated to Radu Teodorescu.");
151 isoToDedication.put("RU", "Russian dictionary dedicated to Maxim Aronin--best friend always!.");
152 isoToDedication.put("SR", "Serbian dictionary dedicated to Filip Crnogorac--thanks for the honey.");
153 isoToDedication.put("ES", "Spanish dictionary made especially for Carolina Tropini! <3 XoXoXXXXX!");
154 isoToDedication.put("SV", "Swedish dictionary dedicated to Kajsa Palmblad--björn kramar!");
156 private static String getDedication(String iso) {
157 return isoToDedication.containsKey(iso) ? "\n\n" + isoToDedication.get(iso) : "";
160 static final Map<String,String> isoToStoplist = new LinkedHashMap<String, String>();
162 isoToStoplist.put("DE", "de.txt");
163 isoToStoplist.put("EN", "en.txt");
164 isoToStoplist.put("ES", "es.txt");
165 isoToStoplist.put("IT", "it.txt");
166 isoToStoplist.put("FR", "fr.txt");
168 private static String getStoplist(String iso) {
169 return isoToStoplist.containsKey(iso) ? isoToStoplist.get(iso) : "empty.txt";
172 static String getOtherLang(final String[] pair, final String first) {
173 assert Arrays.asList(pair).contains(first);
174 assert pair.length == 2;
175 return pair[0].equals(first) ? pair[1] : pair[0];
178 static List<String> getMainArgs(final String[] pair) {
179 final List<String> result = new ArrayList<String>();
183 if (pair.length == 1) {
184 final String lang1 = pair[0];
185 final String dictFile = String.format("%s/%s.quickdic", OUTPUTS, lang1);
186 result.add(String.format("--dictOut=%s", dictFile));
187 result.add(String.format("--lang1=%s", lang1));
188 result.add(String.format("--lang1Stoplist=%s", STOPLISTS + getStoplist(lang1)));
189 result.add(String.format("--dictInfo=Wikitionary-based %s dictionary.", lang1));
192 final String wikiSplitFile = String.format("%s/wikiSplit/%s/%s.data", INPUTS, lang1.toLowerCase(), lang1);
193 if (new File(wikiSplitFile).canRead()) {
194 result.add(String.format("--input%d=%s", i, wikiSplitFile));
195 result.add(String.format("--input%dName=%s.wiktionary.org", i, lang1.toLowerCase()));
196 result.add(String.format("--input%dFormat=%s", i, WholeSectionToHtmlParser.NAME));
197 result.add(String.format("--input%dTitleIndex=%d", i, 1));
198 result.add(String.format("--input%dWiktionaryLang=%s", i, lang1));
199 result.add(String.format("--input%dSkipLang=%s", i, lang1));
200 result.add(String.format("--input%dWebUrlTemplate=http://%s.wiktionary.org/wiki/%%s", i, lang1.toLowerCase()));
201 //result.add(String.format("--input%dPageLimit=100", i));
204 System.err.println("Can't read file: " + wikiSplitFile);
207 if (lang1.equals("EN")) {
208 result.add(String.format("--input%d=%s/wikiSplit/en/%s.data", i, INPUTS, lang1));
209 result.add(String.format("--input%dName=ENWiktionary.%s", i, lang1)) ;
210 result.add(String.format("--input%dFormat=enwiktionary", i));
211 result.add(String.format("--input%dWiktionaryType=EnEnglish", i));
212 result.add(String.format("--input%dLangPattern=%s", i, "English"));
213 result.add(String.format("--input%dLangCodePattern=%s", i, lang1.toLowerCase()));
214 result.add(String.format("--input%dEnIndex=%d", i, 1));
215 //result.add(String.format("--input%dPageLimit=100", i));
220 } // Single-lang dictionaries.
222 final String lang1 = pair[0];
223 final String lang2 = pair[1];
225 final String dictFile = String.format("%s/%s-%s.quickdic",
226 OUTPUTS, lang1, lang2);
228 result.add(String.format("--dictOut=%s", dictFile));
229 result.add(String.format("--lang1Stoplist=%s", STOPLISTS + getStoplist(lang1)));
230 result.add(String.format("--lang2Stoplist=%s", STOPLISTS + getStoplist(lang2)));
232 // For a few langs, put the defs of the other language in DE/IT/FR using WholeSection.
233 for (final String wikitionaryLang : Arrays.asList("EN", "DE", "IT", "FR")) {
234 if (!Arrays.asList(pair).contains(wikitionaryLang)) {
237 final String foreignIso = getOtherLang(pair, wikitionaryLang);
238 final String wikiSplitFile = String.format("%s/wikiSplit/%s/%s.data", INPUTS, wikitionaryLang.toLowerCase(), foreignIso);
239 if (!new File(wikiSplitFile).canRead()) {
240 System.err.println("Can't read file: " + wikiSplitFile);
243 result.add(String.format("--input%d=%s", i, wikiSplitFile));
244 result.add(String.format("--input%dName=%s.wiktionary.org", i, wikitionaryLang.toLowerCase()));
245 result.add(String.format("--input%dFormat=%s", i, WholeSectionToHtmlParser.NAME));
246 result.add(String.format("--input%dTitleIndex=%d", i, Arrays.asList(pair).indexOf(foreignIso) + 1));
247 result.add(String.format("--input%dWiktionaryLang=%s", i, wikitionaryLang));
248 result.add(String.format("--input%dSkipLang=%s", i, foreignIso));
249 result.add(String.format("--input%dWebUrlTemplate=http://%s.wiktionary.org/wiki/%%s", i, wikitionaryLang.toLowerCase()));
253 // Deal with the pairs where one is English.
254 if (Arrays.asList(pair).contains("EN")) {
255 final String foreignIso = getOtherLang(pair, "EN");
257 String foreignRegex = WiktionaryLangs.isoCodeToEnWikiName.get(foreignIso);
258 if (foreignIso.equals("ZH")) {
259 // HACK: The missing "e" prevents a full match, causing "Cantonese" to be appended to the entries.
260 foreignRegex = "Chinese|Mandarin|Cantones";
263 result.add(String.format("--lang1=%s", lang1));
264 result.add(String.format("--lang2=%s", lang2));
265 result.add(String.format("--dictInfo=(EN)Wikitionary-based EN-%s dictionary.%s", foreignIso, getDedication(foreignIso)));
267 result.add(String.format("--input%d=%s/wikiSplit/en/%s.data", i, INPUTS, foreignIso));
268 result.add(String.format("--input%dName=ENWiktionary.%s", i, foreignIso)) ;
269 result.add(String.format("--input%dFormat=enwiktionary", i));
270 result.add(String.format("--input%dWiktionaryType=EnForeign", i));
271 result.add(String.format("--input%dLangPattern=%s", i, foreignRegex));
272 result.add(String.format("--input%dLangCodePattern=%s", i, foreignIso.toLowerCase()));
273 result.add(String.format("--input%dEnIndex=%d", i, Arrays.asList(pair).indexOf("EN") + 1));
276 result.add(String.format("--input%d=%swikiSplit/en/EN.data", i, INPUTS));
277 result.add(String.format("--input%dName=enwiktionary.english", i));
278 result.add(String.format("--input%dFormat=enwiktionary", i));
279 result.add(String.format("--input%dWiktionaryType=EnToTranslation", i));
280 result.add(String.format("--input%dLangPattern=%s", i, foreignRegex));
281 result.add(String.format("--input%dLangCodePattern=%s", i, foreignIso.toLowerCase()));
282 result.add(String.format("--input%dEnIndex=%d", i, Arrays.asList(pair).indexOf("EN") + 1));
285 if (foreignIso.equals("DE")) {
286 result.add(String.format("--input%d=%sde-en_chemnitz.txt", i, INPUTS));
287 result.add(String.format("--input%dName=chemnitz", i));
288 result.add(String.format("--input%dCharset=UTF8", i));
289 result.add(String.format("--input%dFormat=chemnitz", i));
294 // Pairs without English.
295 result.add(String.format("--lang1=%s", lang1));
296 result.add(String.format("--lang2=%s", lang2));
297 result.add(String.format("--dictInfo=Wikitionary-based %s-%s dictionary.", lang1, lang2));
299 result.add(String.format("--input%d=%swikiSplit/en/EN.data", i, INPUTS));
300 result.add(String.format("--input%dName=BETA!enwiktionary.%s-%s", i, lang1, lang2));
301 result.add(String.format("--input%dFormat=%s", i, EnTranslationToTranslationParser.NAME));
302 result.add(String.format("--input%dLangPattern1=%s", i, lang1));
303 result.add(String.format("--input%dLangPattern2=%s", i, lang2));
310 public static void main(final String[] args) throws Exception {
312 final List<String[]> allPairs = new ArrayList<String[]>();
314 allPairs.addAll(Arrays.asList(nonEnPairs));
316 // Add all the EN-XX pairs.
317 for (final String isoCode : WiktionaryLangs.isoCodeToEnWikiName.keySet()) {
318 if (isoCode.equals("EN") || isoCode.equals("DE")) {
321 allPairs.add(new String[] {"EN", isoCode});
323 allPairs.add(new String[] {"EN", "DE"});
326 final Set<List<String>> done = new LinkedHashSet<List<String>>();
327 for (final String[] pair : allPairs) {
329 final List<String> pairList = Arrays.asList(pair);
330 if (done.contains(pairList)) {
335 if (!pairList.contains("IT") || !pairList.contains("EN")) {
339 DictionaryBuilder.main(getMainArgs(pair).toArray(new String[0]));