X-Git-Url: http://gitweb.fperrin.net/?a=blobdiff_plain;f=src%2Fcom%2Fhughes%2Fandroid%2Fdictionary%2Fengine%2FDictionaryBuilderMain.java;h=75e3a08b00fdc113e6b9dc811c22f7835f510adb;hb=5a1b9f8a37d03dc675e1d895817422e9743a5b5b;hp=0faf144a37e474b755c569f02804ebf51d94999d;hpb=0eaf1a63bc6d1145490b64d8c68e5a545401ec16;p=DictionaryPC.git diff --git a/src/com/hughes/android/dictionary/engine/DictionaryBuilderMain.java b/src/com/hughes/android/dictionary/engine/DictionaryBuilderMain.java index 0faf144..75e3a08 100644 --- a/src/com/hughes/android/dictionary/engine/DictionaryBuilderMain.java +++ b/src/com/hughes/android/dictionary/engine/DictionaryBuilderMain.java @@ -14,6 +14,8 @@ package com.hughes.android.dictionary.engine; +import java.io.File; +import java.util.ArrayList; import java.util.Arrays; import java.util.LinkedHashMap; import java.util.LinkedHashSet; @@ -21,57 +23,31 @@ 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; +import junit.framework.TestCase; + public class DictionaryBuilderMain extends TestCase { - - static final String INPUTS = "data/inputs/"; - static final String STOPLISTS = "data/inputs/stoplists/"; - static final String OUTPUTS = "data/outputs/"; - - public static void main(final String[] args) throws Exception { - - // Builds all the dictionaries it can, outputs list to a text file. - - final Map isoToWikiName = new LinkedHashMap(WiktionaryLangs.isoCodeToWikiName); - isoToWikiName.remove("EN"); - isoToWikiName.remove("DE"); - - final Map isoToDedication = new LinkedHashMap(); - 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!"); - - final Map isoToStoplist = new LinkedHashMap(); - 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"); - - final Map isoToRegex = new LinkedHashMap(); - // HACK: The missing "e" prevents a full match, causing "Cantonese" to be appended to the entries. - isoToRegex.put("ZH", "Chinese|Mandarin|Cantones"); - - + + static final String INPUTS = "data/inputs/"; + static final String STOPLISTS = "data/inputs/stoplists/"; + static final String OUTPUTS = "data/outputs/"; + // Build the non EN ones. - - final String[][] nonEnPairs = new String[][] { - - /* + static final String[][] nonEnPairs = new String[][] { + {"EN"}, + {"DE"}, + {"IT"}, + // This one takes a really long time, and the result is too big for code.google.com + //{"FR"}, + + // The 3 I use most: + {"IT", "EN" }, + {"DE", "EN" }, + {"DE", "IT" }, + {"AR", "DE" }, {"AR", "ES" }, {"AR", "FR" }, @@ -80,8 +56,8 @@ public class DictionaryBuilderMain extends TestCase { {"AR", "JA" }, {"AR", "RU" }, {"AR", "TR" }, // Turkish - {"AR", "ZH" }, - + {"AR", "cmn" }, + {"DE", "AR" }, {"DE", "FR" }, {"DE", "CA" }, // Catalan @@ -94,13 +70,16 @@ public class DictionaryBuilderMain extends TestCase { {"DE", "IT" }, {"DE", "JA" }, {"DE", "LA" }, // Latin + {"DE", "NL" }, // Dutch {"DE", "PL" }, // Polish {"DE", "RU" }, {"DE", "SV" }, // Swedish {"DE", "TR" }, // Turkish - {"DE", "ZH" }, + {"DE", "cmn" }, + {"DE", "TA" }, // Tamil + + {"ES", "RU" }, // Spanish-Russian - {"FR", "BG" }, // Bulgarian {"FR", "CS" }, // Czech {"FR", "DE" }, @@ -111,7 +90,8 @@ public class DictionaryBuilderMain extends TestCase { {"FR", "NL" }, // Dutch {"FR", "RU" }, {"FR", "TR" }, // Turkish - {"FR", "ZH" }, + {"FR", "cmn" }, + {"FR", "EL" }, {"IT", "DE" }, {"IT", "EL" }, // Greek @@ -126,164 +106,244 @@ public class DictionaryBuilderMain extends TestCase { {"IT", "RU" }, {"IT", "SV" }, {"IT", "TR" }, // Turkish - {"IT", "ZH" }, + {"IT", "cmn" }, - {"JA", "ZH" }, + {"JA", "cmn" }, {"JA", "AR" }, {"JA", "KO" }, - {"ZH", "AR" }, - {"ZH", "DE" }, - {"ZH", "ES" }, - {"ZH", "FR" }, - {"ZH", "IT" }, - {"ZH", "KO" }, + {"cmn", "AR" }, + {"cmn", "DE" }, + {"cmn", "ES" }, + {"cmn", "FR" }, + {"cmn", "IT" }, + {"cmn", "KO" }, - {"NO", "SV" }, {"NO", "FI" }, {"FI", "SV" }, - + {"PL", "FR" }, // Polish {"PL", "RU" }, // Polish {"PL", "HU" }, // Polish {"PL", "ES" }, // Polish - */ - + {"TR", "EL" }, // Turkish, Greek + + {"FA", "HY" }, // Persian, Armenian, by request. + {"FA", "SV" }, // Persian, Swedish, by request. + {"NL", "PL" }, // Dutch, Polish, by request. }; - - final Set> done = new LinkedHashSet>(); - for (final String[] pair : nonEnPairs) { - Arrays.sort(pair); - final List pairList = Arrays.asList(pair); - if (done.contains(pairList)) { - continue; - } - done.add(pairList); - - final String lang1 = pair[0]; - final String lang2 = pair[1]; - - final String dictFile = String.format("%s/%s-%s_enwiktionary_BETA.quickdic", - OUTPUTS, lang1, lang2); - System.out.println("building dictFile: " + dictFile); - - if (!isoToStoplist.containsKey(lang1)) { - isoToStoplist.put(lang1, "empty.txt"); - } - if (!isoToStoplist.containsKey(lang2)) { - isoToStoplist.put(lang2, "empty.txt"); - } - - DictionaryBuilder.main(new String[] { - String.format("--dictOut=%s", dictFile), - String.format("--lang1=%s", lang1), - String.format("--lang2=%s", lang2), - String.format("--lang1Stoplist=%s", STOPLISTS + isoToStoplist.get(lang1)), - String.format("--lang2Stoplist=%s", STOPLISTS + isoToStoplist.get(lang2)), - String.format("--dictInfo=(EN)Wikitionary-based %s-%s dictionary.", lang1, lang2), - - String.format("--input2=%swikiSplit/en/EN.data", INPUTS), - String.format("--input2Name=BETA!enwiktionary.%s-%s", lang1, lang2), - String.format("--input2Format=%s", EnTranslationToTranslationParser.NAME), - String.format("--input2LangPattern1=%s", lang1), - String.format("--input2LangPattern2=%s", lang2), - }); + + + + static final Map isoToDedication = new LinkedHashMap(); + static { + isoToDedication.put("AF", "Wiktionary-based Afrikaans dictionary dedicated to Heiko and Mariëtte Horn."); + isoToDedication.put("HR", "Wiktionary-based Croatian dictionary dedicated to Ines Viskic and Miro Kresonja."); + isoToDedication.put("NL", "Wiktionary-based Dutch dictionary dedicated to Mike LeBeau."); + isoToDedication.put("DE", "@data/inputs/de-en_dedication.txt"); + isoToDedication.put("EL", "Wiktionary-based Greek dictionary dedicated to Noah Egge."); + isoToDedication.put("IT", "Wiktionary-based Italian dictionary dedicated to Carolina Tropini, my favorite stardust in the whole universe! Ti amo!"); + isoToDedication.put("KO", "Wiktionary-based Korean dictionary dedicated to Ande Elwood--fall fashion und Fernsehturms!"); + isoToDedication.put("PT", "Wiktionary-based Portuguese dictionary dedicated to Carlos Melo, one Tough Mudder."); + isoToDedication.put("RO", "Wiktionary-based Romanian dictionary dedicated to Radu Teodorescu."); + isoToDedication.put("RU", "Wiktionary-based Russian dictionary dedicated to Maxim Aronin--best friend always!."); + isoToDedication.put("SR", "Wiktionary-based Serbian dictionary dedicated to Filip Crnogorac--thanks for the honey!"); + isoToDedication.put("ES", "Wiktionary-based Spanish dictionary made especially for Carolina Tropini! <3 XoXoXXXXX!"); + isoToDedication.put("SV", "Wiktionary-based Swedish dictionary dedicated to Kajsa Palmblad--björn kramar!"); } - if (1==1) { - //return; + private static String getEnDictionaryInfo(String iso) { + return isoToDedication.containsKey(iso) ? isoToDedication.get(iso) : String.format("Wiktionary-based %s dictionary.", iso); } + static final Map isoToStoplist = new LinkedHashMap(); + 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"; + } + + static String getOtherLang(final String[] pair, final String first) { + assert Arrays.asList(pair).contains(first); + assert pair.length == 2; + return pair[0].equals(first) ? pair[1] : pair[0]; + } + + static List getMainArgs(final String[] pair) { + final List result = new ArrayList(); + + int i = 1; + + if (pair.length == 1) { + final String lang1 = pair[0]; + final String dictFile = String.format("%s/%s.quickdic", OUTPUTS, lang1); + result.add(String.format("--dictOut=%s", dictFile)); + result.add(String.format("--lang1=%s", lang1)); + result.add(String.format("--lang1Stoplist=%s", STOPLISTS + getStoplist(lang1))); + result.add(String.format("--dictInfo=Wikitionary-based %s dictionary.", lang1)); + + + final String wikiSplitFile = String.format("%s/wikiSplit/%s/%s.data", INPUTS, lang1.toLowerCase(), lang1); + if (new File(wikiSplitFile).canRead()) { + result.add(String.format("--input%d=%s", i, wikiSplitFile)); + result.add(String.format("--input%dName=%s.wiktionary.org", i, lang1.toLowerCase())); + result.add(String.format("--input%dFormat=%s", i, WholeSectionToHtmlParser.NAME)); + result.add(String.format("--input%dTitleIndex=%d", i, 1)); + result.add(String.format("--input%dWiktionaryLang=%s", i, lang1)); + result.add(String.format("--input%dSkipLang=%s", i, lang1)); + result.add(String.format("--input%dWebUrlTemplate=http://%s.wiktionary.org/wiki/%%s", i, lang1.toLowerCase())); + //result.add(String.format("--input%dPageLimit=100", i)); + ++i; + } else { + System.err.println("Can't read file: " + wikiSplitFile); + } + + if (lang1.equals("EN") && !lang1.equals("EN")) { + // Add a parser that tries to use the definitions. This is + // not very pretty yet. + result.add(String.format("--input%d=%s/wikiSplit/en/%s.data", i, INPUTS, lang1)); + result.add(String.format("--input%dName=ENWiktionary.%s", i, lang1)) ; + result.add(String.format("--input%dFormat=enwiktionary", i)); + result.add(String.format("--input%dWiktionaryType=EnEnglish", i)); + result.add(String.format("--input%dLangPattern=%s", i, "English")); + result.add(String.format("--input%dLangCodePattern=%s", i, lang1.toLowerCase())); + result.add(String.format("--input%dEnIndex=%d", i, 1)); + //result.add(String.format("--input%dPageLimit=100", i)); + ++i; + } - // Now build the EN ones. - -// isoToWikiName.keySet().retainAll(Arrays.asList("UK", "HR", "FI")); - //isoToWikiName.clear(); - boolean go = false; - for (final String foreignIso : isoToWikiName.keySet()) { - if (foreignIso.equals("SL")) { - go = true; - } - if (!go) { - continue; - } - - final String dictFile = String.format("%s/EN-%s_enwiktionary.quickdic", OUTPUTS, foreignIso); - System.out.println("building dictFile: " + dictFile); - - if (!isoToStoplist.containsKey(foreignIso)) { - isoToStoplist.put(foreignIso, "empty.txt"); + return result; + } // Single-lang dictionaries. + + 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))); + + // For a few langs, put the defs of the other language in DE/IT/FR using WholeSection. + for (final String wikitionaryLang : Arrays.asList("EN", "DE", "IT", "FR")) { + if (!Arrays.asList(pair).contains(wikitionaryLang)) { + continue; + } + final String foreignIso = getOtherLang(pair, wikitionaryLang); + final String wikiSplitFile = String.format("%s/wikiSplit/%s/%s.data", INPUTS, wikitionaryLang.toLowerCase(), foreignIso); + if (!new File(wikiSplitFile).canRead()) { + System.err.println("WARNING: Can't read file: " + wikiSplitFile); + continue; + } + result.add(String.format("--input%d=%s", i, wikiSplitFile)); + result.add(String.format("--input%dName=%s.wiktionary.org", i, wikitionaryLang.toLowerCase())); + result.add(String.format("--input%dFormat=%s", i, WholeSectionToHtmlParser.NAME)); + result.add(String.format("--input%dTitleIndex=%d", i, Arrays.asList(pair).indexOf(foreignIso) + 1)); + result.add(String.format("--input%dWiktionaryLang=%s", i, wikitionaryLang)); + result.add(String.format("--input%dSkipLang=%s", i, foreignIso)); + result.add(String.format("--input%dWebUrlTemplate=http://%s.wiktionary.org/wiki/%%s", i, wikitionaryLang.toLowerCase())); + ++i; } - if (!isoToDedication.containsKey(foreignIso)) { - isoToDedication.put(foreignIso, ""); + + // Deal with the pairs where one is English. + if (Arrays.asList(pair).contains("EN")) { + final String foreignIso = getOtherLang(pair, "EN"); + String foreignRegex = WiktionaryLangs.isoCodeToEnWikiName.get(foreignIso); + + result.add(String.format("--lang1=%s", lang1)); + result.add(String.format("--lang2=%s", lang2)); + result.add(String.format("--dictInfo=%s", getEnDictionaryInfo(foreignIso))); + + // Foreign section. + 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, Arrays.asList(pair).indexOf("EN") + 1)); + ++i; + + // Translation section. + 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, Arrays.asList(pair).indexOf("EN") + 1)); + ++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; + } + + } else { + // Pairs without English. + result.add(String.format("--lang1=%s", lang1)); + result.add(String.format("--lang2=%s", lang2)); + result.add(String.format("--dictInfo=Wikitionary-based %s-%s dictionary.", lang1, 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; + + // TODO: Could use FR translation section here too. } - if (!isoToRegex.containsKey(foreignIso)) { - isoToRegex.put(foreignIso, isoToWikiName.get(foreignIso)); + + return result; + } + + public static void main(final String[] args) throws Exception { + + final List allPairs = new ArrayList(); + + allPairs.addAll(Arrays.asList(nonEnPairs)); + // Add all the EN-XX pairs. + for (final String isoCode : WiktionaryLangs.isoCodeToEnWikiName.keySet()) { + if (!isoCode.equals("EN")) { + allPairs.add(new String[] {"EN", isoCode}); + } + } + + + final Set> done = new LinkedHashSet>(); + boolean go = true; + for (final String[] pair : allPairs) { + Arrays.sort(pair); + final List pairList = Arrays.asList(pair); + if (done.contains(pairList)) { + continue; + } + done.add(pairList); + + if (pairList.contains("EN") && pairList.contains("DE")) { + go = true; + } else { + go = false; + } + + if (!go) { + continue; + } + + DictionaryBuilder.main(getMainArgs(pair).toArray(new String[0])); } - - DictionaryBuilder.main(new String[] { - String.format("--dictOut=%s", dictFile), - String.format("--lang1=EN"), - String.format("--lang2=%s", foreignIso), - String.format("--lang1Stoplist=%s", STOPLISTS + isoToStoplist.get("EN")), - String.format("--lang2Stoplist=%s", STOPLISTS + isoToStoplist.get(foreignIso)), - String.format("--dictInfo=(EN)Wikitionary-based EN-%s dictionary.\n\n%s", foreignIso, isoToDedication.get(foreignIso)), - - "--input2=" + INPUTS + "wikiSplit/en/" + foreignIso + ".data", - "--input2Name=enwiktionary." + foreignIso, - "--input2Format=enwiktionary", - "--input2WiktionaryType=EnForeign", - "--input2LangPattern=" + isoToRegex.get(foreignIso), - "--input2LangCodePattern=" + foreignIso.toLowerCase(), - "--input2EnIndex=1", - - "--input3=" + INPUTS + "wikiSplit/en/EN.data", - "--input3Name=enwiktionary.english", - "--input3Format=enwiktionary", - "--input3WiktionaryType=EnToTranslation", - "--input3LangPattern=" + isoToRegex.get(foreignIso), - "--input3LangCodePattern=" + foreignIso.toLowerCase(), - "--input3EnIndex=1", - - }); - - } // foreignIso - - // Now special case German-English. - - final String dictFile = String.format("%s/DE-EN_chemnitz_enwiktionary.quickdic", OUTPUTS); - DictionaryBuilder.main(new String[] { - "--dictOut=" + dictFile, - "--lang1=DE", - "--lang2=EN", - String.format("--lang1Stoplist=%s", STOPLISTS + "de.txt"), - String.format("--lang2Stoplist=%s", STOPLISTS + "en.txt"), - "--dictInfo=@" + INPUTS + "de-en_chemnitz_enwiktionary.info", - - "--input4=" + INPUTS + "de-en_chemnitz.txt", - "--input4Name=chemnitz", - "--input4Charset=UTF8", - "--input4Format=chemnitz", - - "--input2=" + INPUTS + "wikiSplit/en/DE.data", - "--input2Name=enwiktionary.DE", - "--input2Format=enwiktionary", - "--input2WiktionaryType=EnForeign", - "--input2LangPattern=German", - "--input2LangCodePattern=de", - "--input2EnIndex=2", - - "--input3=" + INPUTS + "wikiSplit/en/EN.data", - "--input3Name=enwiktionary.english", - "--input3Format=enwiktionary", - "--input3WiktionaryType=EnToTranslation", - "--input3LangPattern=German", - "--input3LangCodePattern=de", - "--input3EnIndex=2", - }); - - } - + + } }