+
+ static List<String> getMainArgs(final String[] pair) {
+ final List<String> 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;
+ }
+
+ 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;
+ }
+
+ // 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.
+ }
+
+ return result;
+ }
+
+ public static void main(final String[] args) throws Exception {
+
+ final List<String[]> allPairs = new ArrayList<>(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<List<String>> done = new LinkedHashSet<>();
+ boolean go = true;
+ for (final String[] pair : allPairs) {
+ Arrays.sort(pair);
+ final List<String> pairList = Arrays.asList(pair);
+ if (done.contains(pairList)) {
+ continue;
+ }
+ done.add(pairList);
+
+ go = pairList.contains("EN") && pairList.contains("DE");
+
+ if (!go) {
+ continue;
+ }
+
+ DictionaryBuilder.main(getMainArgs(pair).toArray(new String[0]));
+ }
+