/**
* Get ParsedTuv from list of Tuv for specific language.
*
* <p>Language choosed by:<br>
* - with the same language+country<br>
* - if not exist, then with the same language but without country<br>
* - if not exist, then with the same language with whatever country<br>
*/
protected ParsedTuv getTuvByLang(Language lang) {
String langLanguage = lang.getLanguageCode();
String langCountry = lang.getCountryCode();
ParsedTuv tuvLC = null; // Tuv with the same language+country
ParsedTuv tuvL = null; // Tuv with the same language only, without country
ParsedTuv tuvLW = null; // Tuv with the same language+whatever country
for (int i = 0; i < currentTu.tuvs.size(); i++) {
ParsedTuv tuv = currentTu.tuvs.get(i);
String tuvLang = tuv.lang;
if (!langLanguage.regionMatches(true, 0, tuvLang, 0, 2)) {
// language not equals - there is no sense to processing
continue;
}
if (tuvLang.length() < 3) {
// language only, without country
tuvL = tuv;
} else if (langCountry.regionMatches(true, 0, tuvLang, 3, 2)) {
// the same country
tuvLC = tuv;
} else {
// other country
tuvLW = tuv;
}
}
ParsedTuv bestTuv;
if (tuvLC != null) {
bestTuv = tuvLC;
} else if (tuvL != null) {
bestTuv = tuvL;
} else {
bestTuv = tuvLW;
}
return bestTuv;
}
/**
* Glues the sentences back to paragraph.
*
* <p>As sentences are returned by {@link #segment(String, List)} without spaces before and after
* them, this method adds spaces if needed:
*
* <ul>
* <li>For translation to Japanese does <b>not</b> add any spaces. <br>
* A special exceptions are the Break SRX rules that break on space, i.e. before and after
* patterns consist of spaces (they get trimmed to an empty string). For such rules all the
* spaces are added
* <li>For translation from Japanese adds one space
* <li>For all other language combinations adds those spaces as were in the paragraph before.
* </ul>
*
* @param sentences list of translated sentences
* @param spaces information about spaces in original paragraph
* @param brules rules that account to breaks
* @return glued translated paragraph
*/
public static String glue(
Language sourceLang,
Language targetLang,
List<String> sentences,
List<StringBuffer> spaces,
List<Rule> brules) {
if (sentences.size() <= 0) return "";
StringBuffer res = new StringBuffer();
res.append(sentences.get(0));
for (int i = 1; i < sentences.size(); i++) {
StringBuffer sp = new StringBuffer();
sp.append(spaces.get(2 * i - 1));
sp.append(spaces.get(2 * i));
if (CJK_LANGUAGES.contains(targetLang.getLanguageCode().toUpperCase(Locale.ENGLISH))) {
Rule rule = brules.get(i - 1);
char lastChar = res.charAt(res.length() - 1);
if ((lastChar != '.')
&& (!PatternConsts.SPACY_REGEX.matcher(rule.getBeforebreak()).matches()
|| !PatternConsts.SPACY_REGEX.matcher(rule.getAfterbreak()).matches()))
sp.setLength(0);
} else if (CJK_LANGUAGES.contains(sourceLang.getLanguageCode().toUpperCase(Locale.ENGLISH))
&& sp.length() == 0) sp.append(" ");
res.append(sp);
res.append(sentences.get(i));
}
return res.toString();
}
/** Indicates whether some other language is "equal to" this one. */
@Override
public boolean equals(Object lang) {
if (this == lang) return true;
if (lang == null) return false;
if (!(lang instanceof Language)) return false;
Language that = (Language) lang;
return this.getLocaleCode().equals(that.getLocaleCode());
}
public static void doImport(String sourcelang, String targetlang, String folder, String tmxfile)
throws IOException {
Language source = new Language(sourcelang);
Language target = new Language(targetlang);
TMXSaver saver = new TMXSaver(source.getLanguage(), target.getLanguage(), tmxfile);
FolderScanner scanner = new FolderScanner(folder, target.getLocale(), saver);
scanner.scan();
saver.save();
}
protected void parseHeader(StartElement element, final Language sourceLanguage) {
isParagraphSegtype = SEG_PARAGRAPH.equals(getAttributeValue(element, "segtype"));
isOmegaT = CT_OMEGAT.equals(getAttributeValue(element, "creationtool"));
// log some details
Log.logRB("TMXR_INFO_CREATION_TOOL", new Object[] {getAttributeValue(element, "creationtool")});
Log.logRB(
"TMXR_INFO_CREATION_TOOL_VERSION",
new Object[] {getAttributeValue(element, "creationtoolversion")});
Log.logRB("TMXR_INFO_SEG_TYPE", new Object[] {getAttributeValue(element, "segtype")});
Log.logRB("TMXR_INFO_SOURCE_LANG", new Object[] {getAttributeValue(element, "srclang")});
// give a warning if the TMX source language is
// different from the project source language
String tmxSourceLanguage = getAttributeValue(element, "srclang");
if (!tmxSourceLanguage.equalsIgnoreCase(sourceLanguage.getLanguage())) {
Log.logWarningRB(
"TMXR_WARNING_INCORRECT_SOURCE_LANG", new Object[] {tmxSourceLanguage, sourceLanguage});
}
// give a warning that TMX file will be upgraded to sentence segmentation
if (isSegmentingEnabled && isParagraphSegtype) {
Log.logWarningRB("TMXR_WARNING_UPGRADE_SENTSEG");
}
}
@Override
protected String translate(Language sLang, Language tLang, String text) throws Exception {
if (mvYandexKey == null) {
return "";
}
String lvSourceLang = sLang.getLanguageCode().substring(0, 2).toLowerCase();
String lvTargetLang = tLang.getLanguageCode().substring(0, 2).toLowerCase();
String lvShorText = text.length() > 10000 ? text.substring(0, 9999) + "…" : text;
String prev = getFromCache(sLang, tLang, lvShorText);
if (prev != null) {
return prev;
}
// ----------------------------------------------------------------------
Map<String, String> p = new TreeMap<String, String>();
p.put("key", mvYandexKey);
p.put("lang", lvSourceLang + '-' + lvTargetLang);
p.put("text", lvShorText);
YTPPostResponse response = requestTranslate(p); // WikiGet.post(GT_URL, p);
XPathFactory xPathFactory = XPathFactory.newInstance();
// XPath xPathCode = xPathFactory.newXPath();
// String pathCode = "/Translation/@code";
// String code = (String) xPathCode.evaluate(pathCode, new InputSource(new
// StringReader(lvResponse)));
switch (response.code) {
case ERR_OK:
break;
case ERR_KEY_INVALID:
return response.code + ": " + OStrings.getString("MT_ENGINE_YANDEX_INVALID_KEY");
case ERR_KEY_BLOCKED:
return response.code + ": " + OStrings.getString("MT_ENGINE_YANDEX_API_BLOCKED");
case ERR_DAILY_REQ_LIMIT_EXCEEDED:
return response.code + ": " + OStrings.getString("MT_ENGINE_YANDEX_DAILY_LIMIT_DETECT");
case ERR_DAILY_CHAR_LIMIT_EXCEEDED:
return response.code + ": " + OStrings.getString("MT_ENGINE_YANDEX_DAILY_LIMIT_VOLUME");
case ERR_TEXT_TOO_LONG:
return response.code + ": " + OStrings.getString("MT_ENGINE_YANDEX_MAZ_SIZE");
case ERR_UNPROCESSABLE_TEXT:
return response.code
+ ": "
+ OStrings.getString("MT_ENGINE_YANDEX_TRANSLATION_NOT_POSSIBLE");
case ERR_LANG_NOT_SUPPORTED:
return response.code
+ ": "
+ OStrings.getString("MT_ENGINE_YANDEX_DIRECTION_NOT_SUPPORTED");
default:
break;
}
XPath xPathText = xPathFactory.newXPath();
String pathText = "/Translation[@code='200']/text";
String result =
(String) xPathText.evaluate(pathText, new InputSource(new StringReader(response.response)));
putToCache(sLang, tLang, lvShorText, result);
return result;
}
@Override
protected String getDisplayText(Language value) {
return value == null ? null : value + " - " + value.getDisplayName();
}