package com.hughes.android.dictionary.engine;
-import java.io.File;
-import java.io.PrintWriter;
-import java.io.RandomAccessFile;
import java.util.ArrayList;
-import java.util.Collections;
+import java.util.Arrays;
+import java.util.LinkedHashMap;
+import java.util.LinkedHashSet;
import java.util.List;
+import java.util.Map;
+import java.util.Set;
import junit.framework.TestCase;
+import com.hughes.android.dictionary.parser.wiktionary.EnTranslationToTranslationParser;
+import com.hughes.android.dictionary.parser.wiktionary.WholeSectionToHtmlParser;
+import com.hughes.android.dictionary.parser.wiktionary.WiktionaryLangs;
+
public class DictionaryBuilderMain extends TestCase {
- static final String INPUTS = "../DictionaryData/inputs/";
- static final String OUTPUTS = "../DictionaryData/outputs/";
+ static final String INPUTS = "data/inputs/";
+ static final String STOPLISTS = "data/inputs/stoplists/";
+ static final String OUTPUTS = "data/outputs/";
+
- static class Lang {
- final String nameRegex;
- final String isoCode;
- final String wikiSplit;
- public Lang(String nameRegex, String code, final String wikiSplit) {
- this.nameRegex = nameRegex;
- this.isoCode = code;
- this.wikiSplit = wikiSplit;
- }
+ static final Map<String,String> isoToDedication = new LinkedHashMap<String, String>();
+ static {
+ isoToDedication.put("AF", "Afrikaans dictionary dedicated to Heiko and Mariëtte Horn.");
+ isoToDedication.put("HR", "Croatian dictionary dedicated to Ines Viskic and Miro Kresonja.");
+ isoToDedication.put("NL", "Dutch dictionary dedicated to Mike LeBeau.");
+ // German handled in file.
+ isoToDedication.put("EL", "Greek dictionary dedicated to Noah Egge.");
+ isoToDedication.put("IT", "Italian dictionary dedicated to Carolina Tropini, my favorite stardust in the whole universe! Ti amo!");
+ isoToDedication.put("KO", "Korean dictionary dedicated to Ande Elwood--fall fashion und Fernsehturms!");
+ isoToDedication.put("PT", "Portuguese dictionary dedicated to Carlos Melo, one Tough Mudder.");
+ isoToDedication.put("RO", "Romanian dictionary dedicated to Radu Teodorescu.");
+ isoToDedication.put("RU", "Russian dictionary dedicated to Maxim Aronin--best friend always!.");
+ isoToDedication.put("SR", "Serbian dictionary dedicated to Filip Crnogorac--thanks for the honey.");
+ isoToDedication.put("ES", "Spanish dictionary made especially for Carolina Tropini! <3 XoXoXXXXX!");
+ isoToDedication.put("SV", "Swedish dictionary dedicated to Kajsa Palmblad--björn kramar!");
+ }
+ private static String getDedication(String iso) {
+ return isoToDedication.containsKey(iso) ? "\n\n" + isoToDedication.get(iso) : "";
}
+ static final Map<String,String> isoToStoplist = new LinkedHashMap<String, String>();
+ static {
+ isoToStoplist.put("DE", "de.txt");
+ isoToStoplist.put("EN", "en.txt");
+ isoToStoplist.put("ES", "es.txt");
+ isoToStoplist.put("IT", "it.txt");
+ isoToStoplist.put("FR", "fr.txt");
+ }
+ private static String getStoplist(String iso) {
+ return isoToStoplist.containsKey(iso) ? isoToStoplist.get(iso) : "empty.txt";
+ }
- public static void main(final String[] args) throws Exception {
+ static List<String> getMainArgs(final String[] pair) {
+ final List<String> result = new ArrayList<String>();
+
+ final String lang1 = pair[0];
+ final String lang2 = pair[1];
+
+ final String dictFile = String.format("%s/%s-%s.quickdic",
+ OUTPUTS, lang1, lang2);
+
+ result.add(String.format("--dictOut=%s", dictFile));
+ result.add(String.format("--lang1Stoplist=%s", STOPLISTS + getStoplist(lang1)));
+ result.add(String.format("--lang2Stoplist=%s", STOPLISTS + getStoplist(lang2)));
- Lang[] langs1 = new Lang[] {
- new Lang("^English$", "EN", null),
- };
- Lang[] langs2 = new Lang[] {
- new Lang("^.*Italian.*$", "IT", "italian.data"),
- new Lang("^.*Greek.*$", "EL", "greek.data"),
- new Lang("^.*Spanish.*$", "ES", "spanish.data"),
- /*
- new Lang("^German$", "DE"),
- new Lang("^Afrikaans$", "AF"),
- new Lang("^Armenian$", "HY"),
- new Lang("^Arabic$", "AR"),
- new Lang("^Chinese$|^Mandarin$", "ZH"),
- new Lang("^Croation$", "HR"),
- new Lang("^Czech$", "CS"),
- new Lang("^Dutch$", "NL"),
- new Lang("^English$", "EN"),
- new Lang("^Finnish$", "FI"),
- new Lang("^French$", "FR"),
- new Lang("^Hebrew$", "HE"),
- new Lang("^Hindi$", "HI"),
- new Lang("^Icelandic$", "IS"),
- new Lang("^Irish$", "GA"),
- new Lang("^Japanese$", "JA"),
- new Lang("^Korean$", "KO"),
- new Lang("^Kurdish$", "KU"),
- new Lang("^Lithuanian$", "LT"),
- new Lang("^Malay$", "MS"),
- new Lang("^Maori$", "MI"),
- new Lang("^Mongolian$", "MN"),
- new Lang("^Norwegian$", "NO"),
- new Lang("^Persian$", "FA"),
- new Lang("^Portuguese$", "PT"),
- new Lang("^Romanian$", "RO"),
- new Lang("^Russian$", "RU"),
- new Lang("^Sanskrit$", "SA"),
- new Lang("^Serbian$", "SR"),
- new Lang("^Somali$", "SO"),
- new Lang("^Sudanese$", "SU"),
- new Lang("^Swedish$", "SV"),
- new Lang("^Tajik$", "TG"),
- new Lang("^Thai$", "TH"),
- new Lang("^Tibetan$", "BO"),
- new Lang("^Turkish$", "TR"),
- new Lang("^Ukranian$", "UK"),
- new Lang("^Vietnamese$", "VI"),
- new Lang("^Welsh$", "CY"),
- new Lang("^Yiddish$", "YI"),
- new Lang("^Zulu$", "ZU"),*/
- };
+ int i = 2;
+
+ // Deal with the pairs where one is English.
+ if (Arrays.asList(pair).contains("EN")) {
+ final String foreignIso = pair[0].equals("EN") ? pair[1] : pair[0];
+
+
+ String foreignRegex = WiktionaryLangs.isoCodeToEnWikiName.get(foreignIso);
+ if (foreignIso.equals("ZH")) {
+ // HACK: The missing "e" prevents a full match, causing "Cantonese" to be appended to the entries.
+ foreignRegex = "Chinese|Mandarin|Cantones";
+ }
+
+ final int enIndex;
+ if (foreignIso.equals("DE")) {
+ // German-English is a special case since it was the first ever QuickDic!
+ result.add(String.format("--lang1=%s", "DE"));
+ result.add(String.format("--lang2=%s", "EN"));
+ result.add("--dictInfo=@" + INPUTS + "de-en_chemnitz_enwiktionary.info");
+
+ enIndex = 2;
+ } else {
+ result.add(String.format("--lang1=%s", "EN"));
+ result.add(String.format("--lang2=%s", foreignIso));
+ result.add(String.format("--dictInfo=(EN)Wikitionary-based EN-%s dictionary.%s", foreignIso, getDedication(foreignIso)));
+ enIndex = 1;
+ }
+
+ result.add(String.format("--input%d=%s/wikiSplit/en/%s.data", i, INPUTS, foreignIso));
+ result.add(String.format("--input%dName=ENWiktionary.%s", i, foreignIso)) ;
+ result.add(String.format("--input%dFormat=enwiktionary", i));
+ result.add(String.format("--input%dWiktionaryType=EnForeign", i));
+ result.add(String.format("--input%dLangPattern=%s", i, foreignRegex));
+ result.add(String.format("--input%dLangCodePattern=%s", i, foreignIso.toLowerCase()));
+ result.add(String.format("--input%dEnIndex=%d", i, enIndex));
+ ++i;
+
+ result.add(String.format("--input%d=%swikiSplit/en/EN.data", i, INPUTS));
+ result.add(String.format("--input%dName=enwiktionary.english", i));
+ result.add(String.format("--input%dFormat=enwiktionary", i));
+ result.add(String.format("--input%dWiktionaryType=EnToTranslation", i));
+ result.add(String.format("--input%dLangPattern=%s", i, foreignRegex));
+ result.add(String.format("--input%dLangCodePattern=%s", i, foreignIso.toLowerCase()));
+ result.add(String.format("--input%dEnIndex=%d", i, enIndex));
+ ++i;
+
+ if (foreignIso.equals("DE")) {
+ result.add(String.format("--input%d=%sde-en_chemnitz.txt", i, INPUTS));
+ result.add(String.format("--input%dName=chemnitz", i));
+ result.add(String.format("--input%dCharset=UTF8", i));
+ result.add(String.format("--input%dFormat=chemnitz", i));
+ ++i;
+ }
+
+ result.add(String.format("--input%d=%s/wikiSplit/en/%s.data", i, INPUTS, foreignIso));
+ result.add(String.format("--input%dName=%s", i, "ENWiktionary.WholeSections.%s", foreignIso));
+ result.add(String.format("--input%dFormat=%s", i, WholeSectionToHtmlParser.NAME));
+ result.add(String.format("--input%dTitleIndex=%d", i, 3 - enIndex));
+ ++i;
+
+ } else {
+ // Pairs without English.
+ result.add(String.format("--lang1=%s", lang1));
+ result.add(String.format("--lang2=%s", lang2));
+
+ result.add(String.format("--input%d=%swikiSplit/en/EN.data", i, INPUTS));
+ result.add(String.format("--input%dName=BETA!enwiktionary.%s-%s", i, lang1, lang2));
+ result.add(String.format("--input%dFormat=%s", i, EnTranslationToTranslationParser.NAME));
+ result.add(String.format("--input%dLangPattern1=%s", i, lang1));
+ result.add(String.format("--input%dLangPattern2=%s", i, lang2));
+ ++i;
+ }
+ return result;
+ }
+
+ public static void main(final String[] args) throws Exception {
+
+ final List<String[]> allPairs = new ArrayList<String[]>();
- for (final Lang lang1 : langs1) {
- for (final Lang lang2 : langs2) {
- if (lang1.nameRegex.equals(lang2.nameRegex)) {
- continue;
- }
+ // Build the non EN ones.
+ final String[][] nonEnPairs = new String[][] {
- int enIndex = -1;
- Lang nonEnglish = null;
- if (lang2.isoCode.equals("EN")) {
- enIndex = 2;
- nonEnglish = lang1;
- }
- if (lang1.isoCode.equals("EN")) {
- enIndex = 1;
- nonEnglish = lang2;
- }
- assert nonEnglish != null;
-
- final String dictFile = String.format(OUTPUTS + "/%s-%s_enwiktionary.quickdic", lang1.isoCode, lang2.isoCode);
- System.out.println("building dictFile: " + dictFile);
- DictionaryBuilder.main(new String[] {
- String.format("--dictOut=%s", dictFile),
- String.format("--lang1=%s", lang1.isoCode),
- String.format("--lang2=%s", lang2.isoCode),
- String.format("--dictInfo=(EN)Wikitionary-based %s-%s dictionary", lang1.isoCode, lang2.isoCode),
-
- "--input2=" + INPUTS + "wikiSplit/" + nonEnglish.wikiSplit,
- "--input2Name=enwiktionary." + nonEnglish.wikiSplit,
- "--input2Format=enwiktionary",
- "--input2LangPattern=" + nonEnglish.nameRegex,
- "--input2LangCodePattern=" + nonEnglish.isoCode.toLowerCase(),
- "--input2EnIndex=" + enIndex,
-
- "--input3=" + INPUTS + "wikiSplit/english.data",
- "--input3Name=enwiktionary.english",
- "--input3Format=enwiktionary",
- "--input3LangPattern=" + nonEnglish.nameRegex,
- "--input3LangCodePattern=" + (enIndex == 1 ? lang2.isoCode : lang1.isoCode).toLowerCase(),
- "--input3EnIndex=" + enIndex,
-
- });
+ {"AR", "DE" },
+ {"AR", "ES" },
+ {"AR", "FR" },
+ {"AR", "HE" },
+ {"AR", "IT" },
+ {"AR", "JA" },
+ {"AR", "RU" },
+ {"AR", "TR" }, // Turkish
+ {"AR", "ZH" },
- // Print the entries for diffing.
- final RandomAccessFile raf = new RandomAccessFile(new File(dictFile), "r");
- final Dictionary dict = new Dictionary(raf);
- final PrintWriter textOut = new PrintWriter(new File(dictFile + ".text"));
- final List<PairEntry> sorted = new ArrayList<PairEntry>(dict.pairEntries);
- Collections.sort(sorted);
- for (final PairEntry pairEntry : sorted) {
- textOut.println(pairEntry.getRawText(false));
- }
- textOut.close();
- raf.close();
-
- } // langs2
- } // langs1
-
- DictionaryBuilder.main(new String[] {
- "--dictOut=" + OUTPUTS + "DE-EN_chemnitz.quickdic",
- "--lang1=DE",
- "--lang2=EN",
- "--dictInfo=@" + INPUTS + "de-en_chemnitz.info",
-
- "--input1=" + INPUTS + "de-en_chemnitz.txt",
- "--input1Name=chemnitz",
- "--input1Charset=UTF8",
- "--input1Format=chemnitz",
- });
-
- DictionaryBuilder.main(new String[] {
- "--dictOut=" + OUTPUTS + "de-en_all.quickdic",
- "--lang1=DE",
- "--lang2=EN",
- "--dictInfo=@" + INPUTS + "de-en_all.info",
-
- "--input2=" + INPUTS + "de-en_chemnitz.txt",
- "--input2Name=dictcc",
- "--input2Charset=UTF8",
- "--input2Format=chemnitz",
-
- "--input3=" + INPUTS + "/copyrighted/de-en_dictcc.txt",
- "--input3Name=dictcc",
- "--input3Charset=UTF8",
- "--input3Format=dictcc",
+ {"DE", "AR" },
+ {"DE", "FR" },
+ {"DE", "CA" }, // Catalan
+ {"DE", "CS" }, // Czech
+ {"DE", "EO" }, // Esperanto
+ {"DE", "ES" },
+ {"DE", "FR" },
+ {"DE", "HE" },
+ {"DE", "HU" }, // Hungarian
+ {"DE", "IT" },
+ {"DE", "JA" },
+ {"DE", "LA" }, // Latin
+ {"DE", "PL" }, // Polish
+ {"DE", "RU" },
+ {"DE", "SV" }, // Swedish
+ {"DE", "TR" }, // Turkish
+ {"DE", "ZH" },
+
- // TODO: wiktionary
- });
+ {"FR", "BG" }, // Bulgarian
+ {"FR", "CS" }, // Czech
+ {"FR", "DE" },
+ {"FR", "ES" },
+ {"FR", "IT" },
+ {"FR", "JA" },
+ {"FR", "LA" },
+ {"FR", "NL" }, // Dutch
+ {"FR", "RU" },
+ {"FR", "TR" }, // Turkish
+ {"FR", "ZH" },
- }
-
+ {"IT", "DE" },
+ {"IT", "EL" }, // Greek
+ {"IT", "ES" },
+ {"IT", "FR" },
+ {"IT", "HU" },
+ {"IT", "JA" },
+ {"IT", "LA" }, // Latin
+ {"IT", "LV" }, // Latvian
+ {"IT", "NL" },
+ {"IT", "PL" },
+ {"IT", "RU" },
+ {"IT", "SV" },
+ {"IT", "TR" }, // Turkish
+ {"IT", "ZH" },
+
+ {"JA", "ZH" },
+ {"JA", "AR" },
+ {"JA", "KO" },
+
+ {"ZH", "AR" },
+ {"ZH", "DE" },
+ {"ZH", "ES" },
+ {"ZH", "FR" },
+ {"ZH", "IT" },
+ {"ZH", "KO" },
+
+
+ {"NO", "SV" },
+ {"NO", "FI" },
+ {"FI", "SV" },
+
+ {"PL", "FR" }, // Polish
+ {"PL", "RU" }, // Polish
+ {"PL", "HU" }, // Polish
+ {"PL", "ES" }, // Polish
+
+ //{"TR", "EL" }, // Turkish, Greek
+
+ };
+ allPairs.addAll(Arrays.asList(nonEnPairs));
+
+ // Add all the EN-XX pairs.
+ for (final String isoCode : WiktionaryLangs.isoCodeToEnWikiName.keySet()) {
+ if (isoCode.equals("EN") || isoCode.equals("DE")) {
+ continue;
+ }
+ allPairs.add(new String[] {"EN", isoCode});
+ }
+ allPairs.add(new String[] {"EN", "DE"});
+
+
+ final Set<List<String>> done = new LinkedHashSet<List<String>>();
+ for (final String[] pair : allPairs) {
+ Arrays.sort(pair);
+ final List<String> pairList = Arrays.asList(pair);
+ if (done.contains(pairList)) {
+ continue;
+ }
+ done.add(pairList);
+
+ if (!pairList.contains("IT") || !pairList.contains("EN")) {
+ continue;
+ }
+
+ DictionaryBuilder.main(getMainArgs(pair).toArray(new String[0]));
+ }
+
+ }
}