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[][] {
59 {"AR", "TR" }, // Turkish
64 {"DE", "CA" }, // Catalan
65 {"DE", "CS" }, // Czech
66 {"DE", "EO" }, // Esperanto
70 {"DE", "HU" }, // Hungarian
73 {"DE", "LA" }, // Latin
74 {"DE", "NL" }, // Dutch
75 {"DE", "PL" }, // Polish
77 {"DE", "SV" }, // Swedish
78 {"DE", "TR" }, // Turkish
80 {"DE", "TA" }, // Tamil
82 {"ES", "RU" }, // Spanish-Russian
84 {"FR", "BG" }, // Bulgarian
85 {"FR", "CS" }, // Czech
91 {"FR", "NL" }, // Dutch
93 {"FR", "TR" }, // Turkish
98 {"IT", "EL" }, // Greek
103 {"IT", "LA" }, // Latin
104 {"IT", "LV" }, // Latvian
109 {"IT", "TR" }, // Turkish
128 {"PL", "FR" }, // Polish
129 {"PL", "RU" }, // Polish
130 {"PL", "HU" }, // Polish
131 {"PL", "ES" }, // Polish
133 {"TR", "EL" }, // Turkish, Greek
135 {"FA", "HY" }, // Persian, Armenian, by request.
136 {"FA", "SV" }, // Persian, Swedish, by request.
137 {"NL", "PL" }, // Dutch, Polish, by request.
145 static final Map<String,String> isoToDedication = new LinkedHashMap<String, String>();
147 isoToDedication.put("AF", "Afrikaans dictionary dedicated to Heiko and Mariëtte Horn.");
148 isoToDedication.put("HR", "Croatian dictionary dedicated to Ines Viskic and Miro Kresonja.");
149 isoToDedication.put("NL", "Dutch dictionary dedicated to Mike LeBeau.");
150 // German handled in file.
151 isoToDedication.put("EL", "Greek dictionary dedicated to Noah Egge.");
152 isoToDedication.put("IT", "Italian dictionary dedicated to Carolina Tropini, my favorite stardust in the whole universe! Ti amo!");
153 isoToDedication.put("KO", "Korean dictionary dedicated to Ande Elwood--fall fashion und Fernsehturms!");
154 isoToDedication.put("PT", "Portuguese dictionary dedicated to Carlos Melo, one Tough Mudder.");
155 isoToDedication.put("RO", "Romanian dictionary dedicated to Radu Teodorescu.");
156 isoToDedication.put("RU", "Russian dictionary dedicated to Maxim Aronin--best friend always!.");
157 isoToDedication.put("SR", "Serbian dictionary dedicated to Filip Crnogorac--thanks for the honey.");
158 isoToDedication.put("ES", "Spanish dictionary made especially for Carolina Tropini! <3 XoXoXXXXX!");
159 isoToDedication.put("SV", "Swedish dictionary dedicated to Kajsa Palmblad--björn kramar!");
161 private static String getDedication(String iso) {
162 return isoToDedication.containsKey(iso) ? "\n\n" + isoToDedication.get(iso) : "";
165 static final Map<String,String> isoToStoplist = new LinkedHashMap<String, String>();
167 isoToStoplist.put("DE", "de.txt");
168 isoToStoplist.put("EN", "en.txt");
169 isoToStoplist.put("ES", "es.txt");
170 isoToStoplist.put("IT", "it.txt");
171 isoToStoplist.put("FR", "fr.txt");
173 private static String getStoplist(String iso) {
174 return isoToStoplist.containsKey(iso) ? isoToStoplist.get(iso) : "empty.txt";
177 static String getOtherLang(final String[] pair, final String first) {
178 assert Arrays.asList(pair).contains(first);
179 assert pair.length == 2;
180 return pair[0].equals(first) ? pair[1] : pair[0];
183 static List<String> getMainArgs(final String[] pair) {
184 final List<String> result = new ArrayList<String>();
188 if (pair.length == 1) {
189 final String lang1 = pair[0];
190 final String dictFile = String.format("%s/%s.quickdic", OUTPUTS, lang1);
191 result.add(String.format("--dictOut=%s", dictFile));
192 result.add(String.format("--lang1=%s", lang1));
193 result.add(String.format("--lang1Stoplist=%s", STOPLISTS + getStoplist(lang1)));
194 result.add(String.format("--dictInfo=Wikitionary-based %s dictionary.", lang1));
197 final String wikiSplitFile = String.format("%s/wikiSplit/%s/%s.data", INPUTS, lang1.toLowerCase(), lang1);
198 if (new File(wikiSplitFile).canRead()) {
199 result.add(String.format("--input%d=%s", i, wikiSplitFile));
200 result.add(String.format("--input%dName=%s.wiktionary.org", i, lang1.toLowerCase()));
201 result.add(String.format("--input%dFormat=%s", i, WholeSectionToHtmlParser.NAME));
202 result.add(String.format("--input%dTitleIndex=%d", i, 1));
203 result.add(String.format("--input%dWiktionaryLang=%s", i, lang1));
204 result.add(String.format("--input%dSkipLang=%s", i, lang1));
205 result.add(String.format("--input%dWebUrlTemplate=http://%s.wiktionary.org/wiki/%%s", i, lang1.toLowerCase()));
206 //result.add(String.format("--input%dPageLimit=100", i));
209 System.err.println("Can't read file: " + wikiSplitFile);
212 if (lang1.equals("EN")) {
213 result.add(String.format("--input%d=%s/wikiSplit/en/%s.data", i, INPUTS, lang1));
214 result.add(String.format("--input%dName=ENWiktionary.%s", i, lang1)) ;
215 result.add(String.format("--input%dFormat=enwiktionary", i));
216 result.add(String.format("--input%dWiktionaryType=EnEnglish", i));
217 result.add(String.format("--input%dLangPattern=%s", i, "English"));
218 result.add(String.format("--input%dLangCodePattern=%s", i, lang1.toLowerCase()));
219 result.add(String.format("--input%dEnIndex=%d", i, 1));
220 //result.add(String.format("--input%dPageLimit=100", i));
225 } // Single-lang dictionaries.
227 final String lang1 = pair[0];
228 final String lang2 = pair[1];
230 final String dictFile = String.format("%s/%s-%s.quickdic",
231 OUTPUTS, lang1, lang2);
233 result.add(String.format("--dictOut=%s", dictFile));
234 result.add(String.format("--lang1Stoplist=%s", STOPLISTS + getStoplist(lang1)));
235 result.add(String.format("--lang2Stoplist=%s", STOPLISTS + getStoplist(lang2)));
237 // For a few langs, put the defs of the other language in DE/IT/FR using WholeSection.
238 for (final String wikitionaryLang : Arrays.asList("EN", "DE", "IT", "FR")) {
239 if (!Arrays.asList(pair).contains(wikitionaryLang)) {
242 final String foreignIso = getOtherLang(pair, wikitionaryLang);
243 final String wikiSplitFile = String.format("%s/wikiSplit/%s/%s.data", INPUTS, wikitionaryLang.toLowerCase(), foreignIso);
244 if (!new File(wikiSplitFile).canRead()) {
245 System.err.println("Can't read file: " + wikiSplitFile);
248 result.add(String.format("--input%d=%s", i, wikiSplitFile));
249 result.add(String.format("--input%dName=%s.wiktionary.org", i, wikitionaryLang.toLowerCase()));
250 result.add(String.format("--input%dFormat=%s", i, WholeSectionToHtmlParser.NAME));
251 result.add(String.format("--input%dTitleIndex=%d", i, Arrays.asList(pair).indexOf(foreignIso) + 1));
252 result.add(String.format("--input%dWiktionaryLang=%s", i, wikitionaryLang));
253 result.add(String.format("--input%dSkipLang=%s", i, foreignIso));
254 result.add(String.format("--input%dWebUrlTemplate=http://%s.wiktionary.org/wiki/%%s", i, wikitionaryLang.toLowerCase()));
258 // Deal with the pairs where one is English.
259 if (Arrays.asList(pair).contains("EN")) {
260 final String foreignIso = getOtherLang(pair, "EN");
262 String foreignRegex = WiktionaryLangs.isoCodeToEnWikiName.get(foreignIso);
263 if (foreignIso.equals("ZH")) {
264 // HACK: The missing "e" prevents a full match, causing "Cantonese" to be appended to the entries.
265 foreignRegex = "Chinese|Mandarin|Cantones";
268 result.add(String.format("--lang1=%s", lang1));
269 result.add(String.format("--lang2=%s", lang2));
270 result.add(String.format("--dictInfo=(EN)Wikitionary-based EN-%s dictionary.%s", foreignIso, getDedication(foreignIso)));
272 result.add(String.format("--input%d=%s/wikiSplit/en/%s.data", i, INPUTS, foreignIso));
273 result.add(String.format("--input%dName=ENWiktionary.%s", i, foreignIso)) ;
274 result.add(String.format("--input%dFormat=enwiktionary", i));
275 result.add(String.format("--input%dWiktionaryType=EnForeign", i));
276 result.add(String.format("--input%dLangPattern=%s", i, foreignRegex));
277 result.add(String.format("--input%dLangCodePattern=%s", i, foreignIso.toLowerCase()));
278 result.add(String.format("--input%dEnIndex=%d", i, Arrays.asList(pair).indexOf("EN") + 1));
281 result.add(String.format("--input%d=%swikiSplit/en/EN.data", i, INPUTS));
282 result.add(String.format("--input%dName=enwiktionary.english", i));
283 result.add(String.format("--input%dFormat=enwiktionary", i));
284 result.add(String.format("--input%dWiktionaryType=EnToTranslation", i));
285 result.add(String.format("--input%dLangPattern=%s", i, foreignRegex));
286 result.add(String.format("--input%dLangCodePattern=%s", i, foreignIso.toLowerCase()));
287 result.add(String.format("--input%dEnIndex=%d", i, Arrays.asList(pair).indexOf("EN") + 1));
290 if (foreignIso.equals("DE")) {
291 result.add(String.format("--input%d=%sde-en_chemnitz.txt", i, INPUTS));
292 result.add(String.format("--input%dName=chemnitz", i));
293 result.add(String.format("--input%dCharset=UTF8", i));
294 result.add(String.format("--input%dFormat=chemnitz", i));
299 // Pairs without English.
300 result.add(String.format("--lang1=%s", lang1));
301 result.add(String.format("--lang2=%s", lang2));
302 result.add(String.format("--dictInfo=Wikitionary-based %s-%s dictionary.", lang1, lang2));
304 result.add(String.format("--input%d=%swikiSplit/en/EN.data", i, INPUTS));
305 result.add(String.format("--input%dName=BETA!enwiktionary.%s-%s", i, lang1, lang2));
306 result.add(String.format("--input%dFormat=%s", i, EnTranslationToTranslationParser.NAME));
307 result.add(String.format("--input%dLangPattern1=%s", i, lang1));
308 result.add(String.format("--input%dLangPattern2=%s", i, lang2));
315 public static void main(final String[] args) throws Exception {
317 final List<String[]> allPairs = new ArrayList<String[]>();
319 allPairs.addAll(Arrays.asList(nonEnPairs));
320 // Add all the EN-XX pairs.
321 for (final String isoCode : WiktionaryLangs.isoCodeToEnWikiName.keySet()) {
322 allPairs.add(new String[] {"EN", isoCode});
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("EN") && !pairList.contains("EL")) {
339 DictionaryBuilder.main(getMainArgs(pair).toArray(new String[0]));