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/";
39 static final Map<String,String> isoToDedication = new LinkedHashMap<String, String>();
41 isoToDedication.put("AF", "Afrikaans dictionary dedicated to Heiko and Mariëtte Horn.");
42 isoToDedication.put("HR", "Croatian dictionary dedicated to Ines Viskic and Miro Kresonja.");
43 isoToDedication.put("NL", "Dutch dictionary dedicated to Mike LeBeau.");
44 // German handled in file.
45 isoToDedication.put("EL", "Greek dictionary dedicated to Noah Egge.");
46 isoToDedication.put("IT", "Italian dictionary dedicated to Carolina Tropini, my favorite stardust in the whole universe! Ti amo!");
47 isoToDedication.put("KO", "Korean dictionary dedicated to Ande Elwood--fall fashion und Fernsehturms!");
48 isoToDedication.put("PT", "Portuguese dictionary dedicated to Carlos Melo, one Tough Mudder.");
49 isoToDedication.put("RO", "Romanian dictionary dedicated to Radu Teodorescu.");
50 isoToDedication.put("RU", "Russian dictionary dedicated to Maxim Aronin--best friend always!.");
51 isoToDedication.put("SR", "Serbian dictionary dedicated to Filip Crnogorac--thanks for the honey.");
52 isoToDedication.put("ES", "Spanish dictionary made especially for Carolina Tropini! <3 XoXoXXXXX!");
53 isoToDedication.put("SV", "Swedish dictionary dedicated to Kajsa Palmblad--björn kramar!");
55 private static String getDedication(String iso) {
56 return isoToDedication.containsKey(iso) ? "\n\n" + isoToDedication.get(iso) : "";
59 static final Map<String,String> isoToStoplist = new LinkedHashMap<String, String>();
61 isoToStoplist.put("DE", "de.txt");
62 isoToStoplist.put("EN", "en.txt");
63 isoToStoplist.put("ES", "es.txt");
64 isoToStoplist.put("IT", "it.txt");
65 isoToStoplist.put("FR", "fr.txt");
67 private static String getStoplist(String iso) {
68 return isoToStoplist.containsKey(iso) ? isoToStoplist.get(iso) : "empty.txt";
71 static String getOtherLang(final String[] pair, final String first) {
72 assert Arrays.asList(pair).contains(first);
73 assert pair.length == 2;
74 return pair[0].equals(first) ? pair[1] : pair[0];
77 static List<String> getMainArgs(final String[] pair) {
78 final List<String> result = new ArrayList<String>();
80 final String lang1 = pair[0];
81 final String lang2 = pair[1];
83 final String dictFile = String.format("%s/%s-%s.quickdic",
84 OUTPUTS, lang1, lang2);
86 result.add(String.format("--dictOut=%s", dictFile));
87 result.add(String.format("--lang1Stoplist=%s", STOPLISTS + getStoplist(lang1)));
88 result.add(String.format("--lang2Stoplist=%s", STOPLISTS + getStoplist(lang2)));
92 // For a few langs, put the defs of the other language in DE/IT/FR using WholeSection.
93 for (final String wikitionaryLang : Arrays.asList("EN", "DE", "IT", "FR")) {
94 if (!Arrays.asList(pair).contains(wikitionaryLang)) {
97 final String foreignIso = getOtherLang(pair, wikitionaryLang);
98 final String wikiSplitFile = String.format("%s/wikiSplit/%s/%s.data", INPUTS, wikitionaryLang.toLowerCase(), foreignIso);
99 if (!new File(wikiSplitFile).canRead()) {
100 System.err.println("Can't read file: " + wikiSplitFile);
103 result.add(String.format("--input%d=%s", i, wikiSplitFile));
104 result.add(String.format("--input%dName=%s.wiktionary.org (%s)", i, wikitionaryLang, foreignIso));
105 result.add(String.format("--input%dFormat=%s", i, WholeSectionToHtmlParser.NAME));
106 result.add(String.format("--input%dTitleIndex=%d", i, Arrays.asList(pair).indexOf(foreignIso) + 1));
107 result.add(String.format("--input%dWiktionaryLang=%s", i, wikitionaryLang));
108 result.add(String.format("--input%dSkipLang=%s", i, foreignIso));
112 // Deal with the pairs where one is English.
113 if (Arrays.asList(pair).contains("EN")) {
114 final String foreignIso = getOtherLang(pair, "EN");
116 String foreignRegex = WiktionaryLangs.isoCodeToEnWikiName.get(foreignIso);
117 if (foreignIso.equals("ZH")) {
118 // HACK: The missing "e" prevents a full match, causing "Cantonese" to be appended to the entries.
119 foreignRegex = "Chinese|Mandarin|Cantones";
122 result.add(String.format("--lang1=%s", lang1));
123 result.add(String.format("--lang2=%s", lang2));
124 result.add(String.format("--dictInfo=(EN)Wikitionary-based EN-%s dictionary.%s", foreignIso, getDedication(foreignIso)));
126 result.add(String.format("--input%d=%s/wikiSplit/en/%s.data", i, INPUTS, foreignIso));
127 result.add(String.format("--input%dName=ENWiktionary.%s", i, foreignIso)) ;
128 result.add(String.format("--input%dFormat=enwiktionary", i));
129 result.add(String.format("--input%dWiktionaryType=EnForeign", i));
130 result.add(String.format("--input%dLangPattern=%s", i, foreignRegex));
131 result.add(String.format("--input%dLangCodePattern=%s", i, foreignIso.toLowerCase()));
132 result.add(String.format("--input%dEnIndex=%d", i, Arrays.asList(pair).indexOf("EN") + 1));
135 result.add(String.format("--input%d=%swikiSplit/en/EN.data", i, INPUTS));
136 result.add(String.format("--input%dName=enwiktionary.english", i));
137 result.add(String.format("--input%dFormat=enwiktionary", i));
138 result.add(String.format("--input%dWiktionaryType=EnToTranslation", i));
139 result.add(String.format("--input%dLangPattern=%s", i, foreignRegex));
140 result.add(String.format("--input%dLangCodePattern=%s", i, foreignIso.toLowerCase()));
141 result.add(String.format("--input%dEnIndex=%d", i, Arrays.asList(pair).indexOf("EN") + 1));
144 if (foreignIso.equals("DE")) {
145 result.add(String.format("--input%d=%sde-en_chemnitz.txt", i, INPUTS));
146 result.add(String.format("--input%dName=chemnitz", i));
147 result.add(String.format("--input%dCharset=UTF8", i));
148 result.add(String.format("--input%dFormat=chemnitz", i));
153 // Pairs without English.
154 result.add(String.format("--lang1=%s", lang1));
155 result.add(String.format("--lang2=%s", lang2));
156 result.add(String.format("--dictInfo=Wikitionary-based %s-%s dictionary.", lang1, lang2));
158 result.add(String.format("--input%d=%swikiSplit/en/EN.data", i, INPUTS));
159 result.add(String.format("--input%dName=BETA!enwiktionary.%s-%s", i, lang1, lang2));
160 result.add(String.format("--input%dFormat=%s", i, EnTranslationToTranslationParser.NAME));
161 result.add(String.format("--input%dLangPattern1=%s", i, lang1));
162 result.add(String.format("--input%dLangPattern2=%s", i, lang2));
169 public static void main(final String[] args) throws Exception {
171 final List<String[]> allPairs = new ArrayList<String[]>();
173 // Build the non EN ones.
174 final String[][] nonEnPairs = new String[][] {
189 {"AR", "TR" }, // Turkish
194 {"DE", "CA" }, // Catalan
195 {"DE", "CS" }, // Czech
196 {"DE", "EO" }, // Esperanto
200 {"DE", "HU" }, // Hungarian
203 {"DE", "LA" }, // Latin
204 {"DE", "NL" }, // Dutch
205 {"DE", "PL" }, // Polish
207 {"DE", "SV" }, // Swedish
208 {"DE", "TR" }, // Turkish
210 {"DE", "TA" }, // Tamil
212 {"ES", "RU" }, // Spanish-Russian
214 {"FR", "BG" }, // Bulgarian
215 {"FR", "CS" }, // Czech
221 {"FR", "NL" }, // Dutch
223 {"FR", "TR" }, // Turkish
227 {"IT", "EL" }, // Greek
232 {"IT", "LA" }, // Latin
233 {"IT", "LV" }, // Latvian
238 {"IT", "TR" }, // Turkish
257 {"PL", "FR" }, // Polish
258 {"PL", "RU" }, // Polish
259 {"PL", "HU" }, // Polish
260 {"PL", "ES" }, // Polish
262 {"TR", "EL" }, // Turkish, Greek
264 {"FA", "HY" }, // Persian, Armenian, by request.
265 {"FA", "SV" }, // Persian, Swedish, by request.
268 allPairs.addAll(Arrays.asList(nonEnPairs));
270 // Add all the EN-XX pairs.
271 for (final String isoCode : WiktionaryLangs.isoCodeToEnWikiName.keySet()) {
272 if (isoCode.equals("EN") || isoCode.equals("DE")) {
275 allPairs.add(new String[] {"EN", isoCode});
277 allPairs.add(new String[] {"EN", "DE"});
280 final Set<List<String>> done = new LinkedHashSet<List<String>>();
281 for (final String[] pair : allPairs) {
283 final List<String> pairList = Arrays.asList(pair);
284 if (done.contains(pairList)) {
289 if (!pairList.contains("IT") || !pairList.contains("EN")) {
293 DictionaryBuilder.main(getMainArgs(pair).toArray(new String[0]));