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[][] {
43 // This one takes a really long time, and the result is too big for code.google.com
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
97 {"IT", "EL" }, // Greek
102 {"IT", "LA" }, // Latin
103 {"IT", "LV" }, // Latvian
108 {"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.
135 {"NL", "PL" }, // Dutch, Polish, by request.
141 static final Map<String,String> isoToDedication = new LinkedHashMap<String, String>();
143 isoToDedication.put("AF", "Wiktionary-based Afrikaans dictionary dedicated to Heiko and Mariëtte Horn.");
144 isoToDedication.put("HR", "Wiktionary-based Croatian dictionary dedicated to Ines Viskic and Miro Kresonja.");
145 isoToDedication.put("NL", "Wiktionary-based Dutch dictionary dedicated to Mike LeBeau.");
146 isoToDedication.put("DE", "@data/inputs/de-en_dedication.txt");
147 isoToDedication.put("EL", "Wiktionary-based Greek dictionary dedicated to Noah Egge.");
148 isoToDedication.put("IT", "Wiktionary-based Italian dictionary dedicated to Carolina Tropini, my favorite stardust in the whole universe! Ti amo!");
149 isoToDedication.put("KO", "Wiktionary-based Korean dictionary dedicated to Ande Elwood--fall fashion und Fernsehturms!");
150 isoToDedication.put("PT", "Wiktionary-based Portuguese dictionary dedicated to Carlos Melo, one Tough Mudder.");
151 isoToDedication.put("RO", "Wiktionary-based Romanian dictionary dedicated to Radu Teodorescu.");
152 isoToDedication.put("RU", "Wiktionary-based Russian dictionary dedicated to Maxim Aronin--best friend always!.");
153 isoToDedication.put("SR", "Wiktionary-based Serbian dictionary dedicated to Filip Crnogorac--thanks for the honey!");
154 isoToDedication.put("ES", "Wiktionary-based Spanish dictionary made especially for Carolina Tropini! <3 XoXoXXXXX!");
155 isoToDedication.put("SV", "Wiktionary-based Swedish dictionary dedicated to Kajsa Palmblad--björn kramar!");
157 private static String getEnDictionaryInfo(String iso) {
158 return isoToDedication.containsKey(iso) ? isoToDedication.get(iso) : String.format("Wiktionary-based %s dictionary.", iso);
161 static final Map<String,String> isoToStoplist = new LinkedHashMap<String, String>();
163 isoToStoplist.put("DE", "de.txt");
164 isoToStoplist.put("EN", "en.txt");
165 isoToStoplist.put("ES", "es.txt");
166 isoToStoplist.put("IT", "it.txt");
167 isoToStoplist.put("FR", "fr.txt");
169 private static String getStoplist(String iso) {
170 return isoToStoplist.containsKey(iso) ? isoToStoplist.get(iso) : "empty.txt";
173 static String getOtherLang(final String[] pair, final String first) {
174 assert Arrays.asList(pair).contains(first);
175 assert pair.length == 2;
176 return pair[0].equals(first) ? pair[1] : pair[0];
179 static List<String> getMainArgs(final String[] pair) {
180 final List<String> result = new ArrayList<String>();
184 if (pair.length == 1) {
185 final String lang1 = pair[0];
186 final String dictFile = String.format("%s/%s.quickdic", OUTPUTS, lang1);
187 result.add(String.format("--dictOut=%s", dictFile));
188 result.add(String.format("--lang1=%s", lang1));
189 result.add(String.format("--lang1Stoplist=%s", STOPLISTS + getStoplist(lang1)));
190 result.add(String.format("--dictInfo=Wikitionary-based %s dictionary.", lang1));
193 final String wikiSplitFile = String.format("%s/wikiSplit/%s/%s.data", INPUTS, lang1.toLowerCase(), lang1);
194 if (new File(wikiSplitFile).canRead()) {
195 result.add(String.format("--input%d=%s", i, wikiSplitFile));
196 result.add(String.format("--input%dName=%s.wiktionary.org", i, lang1.toLowerCase()));
197 result.add(String.format("--input%dFormat=%s", i, WholeSectionToHtmlParser.NAME));
198 result.add(String.format("--input%dTitleIndex=%d", i, 1));
199 result.add(String.format("--input%dWiktionaryLang=%s", i, lang1));
200 result.add(String.format("--input%dSkipLang=%s", i, lang1));
201 result.add(String.format("--input%dWebUrlTemplate=http://%s.wiktionary.org/wiki/%%s", i, lang1.toLowerCase()));
202 //result.add(String.format("--input%dPageLimit=100", i));
205 System.err.println("Can't read file: " + wikiSplitFile);
208 if (lang1.equals("EN") && !lang1.equals("EN")) {
209 // Add a parser that tries to use the definitions. This is
210 // not very pretty yet.
211 result.add(String.format("--input%d=%s/wikiSplit/en/%s.data", i, INPUTS, lang1));
212 result.add(String.format("--input%dName=ENWiktionary.%s", i, lang1)) ;
213 result.add(String.format("--input%dFormat=enwiktionary", i));
214 result.add(String.format("--input%dWiktionaryType=EnEnglish", i));
215 result.add(String.format("--input%dLangPattern=%s", i, "English"));
216 result.add(String.format("--input%dLangCodePattern=%s", i, lang1.toLowerCase()));
217 result.add(String.format("--input%dEnIndex=%d", i, 1));
218 //result.add(String.format("--input%dPageLimit=100", i));
223 } // Single-lang dictionaries.
225 final String lang1 = pair[0];
226 final String lang2 = pair[1];
228 final String dictFile = String.format("%s/%s-%s.quickdic",
229 OUTPUTS, lang1, lang2);
231 result.add(String.format("--dictOut=%s", dictFile));
232 result.add(String.format("--lang1Stoplist=%s", STOPLISTS + getStoplist(lang1)));
233 result.add(String.format("--lang2Stoplist=%s", STOPLISTS + getStoplist(lang2)));
235 // For a few langs, put the defs of the other language in DE/IT/FR using WholeSection.
236 for (final String wikitionaryLang : Arrays.asList("EN", "DE", "IT", "FR")) {
237 if (!Arrays.asList(pair).contains(wikitionaryLang)) {
240 final String foreignIso = getOtherLang(pair, wikitionaryLang);
241 final String wikiSplitFile = String.format("%s/wikiSplit/%s/%s.data", INPUTS, wikitionaryLang.toLowerCase(), foreignIso);
242 if (!new File(wikiSplitFile).canRead()) {
243 System.err.println("WARNING: Can't read file: " + wikiSplitFile);
246 result.add(String.format("--input%d=%s", i, wikiSplitFile));
247 result.add(String.format("--input%dName=%s.wiktionary.org", i, wikitionaryLang.toLowerCase()));
248 result.add(String.format("--input%dFormat=%s", i, WholeSectionToHtmlParser.NAME));
249 result.add(String.format("--input%dTitleIndex=%d", i, Arrays.asList(pair).indexOf(foreignIso) + 1));
250 result.add(String.format("--input%dWiktionaryLang=%s", i, wikitionaryLang));
251 result.add(String.format("--input%dSkipLang=%s", i, foreignIso));
252 result.add(String.format("--input%dWebUrlTemplate=http://%s.wiktionary.org/wiki/%%s", i, wikitionaryLang.toLowerCase()));
256 // Deal with the pairs where one is English.
257 if (Arrays.asList(pair).contains("EN")) {
258 final String foreignIso = getOtherLang(pair, "EN");
259 String foreignRegex = WiktionaryLangs.isoCodeToEnWikiName.get(foreignIso);
261 result.add(String.format("--lang1=%s", lang1));
262 result.add(String.format("--lang2=%s", lang2));
263 result.add(String.format("--dictInfo=%s", getEnDictionaryInfo(foreignIso)));
266 result.add(String.format("--input%d=%s/wikiSplit/en/%s.data", i, INPUTS, foreignIso));
267 result.add(String.format("--input%dName=ENWiktionary.%s", i, foreignIso)) ;
268 result.add(String.format("--input%dFormat=enwiktionary", i));
269 result.add(String.format("--input%dWiktionaryType=EnForeign", i));
270 result.add(String.format("--input%dLangPattern=%s", i, foreignRegex));
271 result.add(String.format("--input%dLangCodePattern=%s", i, foreignIso.toLowerCase()));
272 result.add(String.format("--input%dEnIndex=%d", i, Arrays.asList(pair).indexOf("EN") + 1));
275 // Translation section.
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));
306 // TODO: Could use FR translation section here too.
312 public static void main(final String[] args) throws Exception {
314 final List<String[]> allPairs = new ArrayList<String[]>();
316 allPairs.addAll(Arrays.asList(nonEnPairs));
317 // Add all the EN-XX pairs.
318 for (final String isoCode : WiktionaryLangs.isoCodeToEnWikiName.keySet()) {
319 if (!isoCode.equals("EN")) {
320 allPairs.add(new String[] {"EN", isoCode});
325 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("EN") && pairList.contains("DE")) {
345 DictionaryBuilder.main(getMainArgs(pair).toArray(new String[0]));