use of org.omegat.core.data.ITMXEntry in project omegat by omegat-org.
the class Searcher method searchEntries.
/**
* Loops over collection of TMXEntries and checks every entry.
* If max nr of hits have been reached or search has been stopped,
* the function stops and returns false. Else it finishes and returns true;
*
* @param tmEn collection of TMX Entries to check.
* @param tmxID identifier of the TMX. E.g. the filename or language code
* @throws SearchLimitReachedException when nr of found matches exceeds requested nr of results
*/
private void searchEntries(Iterable<? extends ITMXEntry> tmEn, int origin, final String tmxID) throws SearchLimitReachedException {
for (ITMXEntry tm : tmEn) {
// stop searching if the max. nr of hits has been reached
if (m_numFinds >= searchExpression.numberOfResults) {
throw new SearchLimitReachedException();
}
// for alternative translations:
// - it is not feasible to get the SourceTextEntry that matches the tm.source, so we cannot get the entryNum
// and real translation
// - although the 'translation' is used as 'source', we search it as translation, else we cannot show to
// which real source it belongs
checkEntry(tm.getSourceText(), tm.getTranslationText(), tm.getNote(), null, null, origin, tmxID);
checkStop.checkInterrupted();
}
}
use of org.omegat.core.data.ITMXEntry in project omegat by omegat-org.
the class POFilterTest method testLoad.
@Test
public void testLoad() throws Exception {
String f = "test/data/filters/po/file-POFilter-multiple.po";
Map<String, String> options = new TreeMap<String, String>();
options.put("skipHeader", "true");
TestFileInfo fi = loadSourceFiles(new PoFilter(), f, options);
String comment = OStrings.getString("POFILTER_TRANSLATOR_COMMENTS") + "\n" + "A valid comment\nAnother valid comment\n\n" + OStrings.getString("POFILTER_EXTRACTED_COMMENTS") + "\n" + "Some extracted comments\nMore extracted comments\n\n" + OStrings.getString("POFILTER_REFERENCES") + "\n" + "/my/source/file\n/my/source/file2\n\n";
checkMultiStart(fi, f);
checkMulti("source1", null, "some context", null, null, comment);
checkMulti("source2", null, "", null, null, null);
checkMulti("source3", null, "", null, null, null);
checkMulti("source1", null, "", null, null, null);
checkMulti("source1", null, "other context", null, null, null);
checkMulti("source4", null, "one more context", null, null, null);
checkMulti("source4", null, "one more context[1]", null, null, StringUtil.format(OStrings.getString("POFILTER_PLURAL_FORM_COMMENT"), 1) + "\n");
checkMulti("source4", null, "one more context[2]", null, null, StringUtil.format(OStrings.getString("POFILTER_PLURAL_FORM_COMMENT"), 2) + "\n");
checkMulti("source5", null, "", null, null, null);
checkMulti("source6", null, "", null, null, null);
checkMultiEnd();
ExternalTMX tmEntries = fi.referenceEntries;
assertEquals(2, tmEntries.getEntries().size());
{
ITMXEntry entry = tmEntries.getEntries().get(0);
assertEquals("True fuzzy!", entry.getSourceText());
assertEquals("trans5", entry.getTranslationText());
}
{
ITMXEntry entry = tmEntries.getEntries().get(1);
assertEquals("True fuzzy 2!", entry.getSourceText());
assertEquals("trans6", entry.getTranslationText());
}
}
use of org.omegat.core.data.ITMXEntry in project omegat by omegat-org.
the class FindMatches method search.
public List<NearString> search(final String searchText, final boolean requiresTranslation, final boolean fillSimilarityData, final IStopped stop) throws StoppedException {
result = new ArrayList<>(OConsts.MAX_NEAR_STRINGS + 1);
srcText = searchText;
removedText = "";
// of the translatable text
if (removePattern != null) {
StringBuilder removedBuffer = new StringBuilder();
Matcher removeMatcher = removePattern.matcher(srcText);
while (removeMatcher.find()) {
removedBuffer.append(removeMatcher.group());
}
srcText = removeMatcher.replaceAll("");
removedText = removedBuffer.toString();
}
// get tokens for original string
strTokensStem = tokenizeStem(srcText);
strTokensNoStem = tokenizeNoStem(srcText);
strTokensAll = tokenizeAll(srcText);
// travel by project entries, including orphaned
if (project.getProjectProperties().isSupportDefaultTranslations()) {
project.iterateByDefaultTranslations(new DefaultTranslationsIterator() {
public void iterate(String source, TMXEntry trans) {
checkStopped(stop);
if (!searchExactlyTheSame && source.equals(searchText)) {
// skip original==original entry comparison
return;
}
if (requiresTranslation && trans.translation == null) {
return;
}
String fileName = project.isOrphaned(source) ? ORPHANED_FILE_NAME : null;
processEntry(null, source, trans.translation, NearString.MATCH_SOURCE.MEMORY, false, 0, fileName, trans.creator, trans.creationDate, trans.changer, trans.changeDate, null);
}
});
}
project.iterateByMultipleTranslations(new MultipleTranslationsIterator() {
public void iterate(EntryKey source, TMXEntry trans) {
checkStopped(stop);
if (!searchExactlyTheSame && source.sourceText.equals(searchText)) {
// skip original==original entry comparison
return;
}
if (requiresTranslation && trans.translation == null) {
return;
}
String fileName = project.isOrphaned(source) ? ORPHANED_FILE_NAME : null;
processEntry(source, source.sourceText, trans.translation, NearString.MATCH_SOURCE.MEMORY, false, 0, fileName, trans.creator, trans.creationDate, trans.changer, trans.changeDate, null);
}
});
/*
Penalty applied for fuzzy matches in another language (if no match in the
target language was found).
*/
int foreignPenalty = Preferences.getPreferenceDefault(Preferences.PENALTY_FOR_FOREIGN_MATCHES, Preferences.PENALTY_FOR_FOREIGN_MATCHES_DEFAULT);
// travel by translation memories
for (Map.Entry<String, ExternalTMX> en : project.getTransMemories().entrySet()) {
int penalty = 0;
Matcher matcher = SEARCH_FOR_PENALTY.matcher(en.getKey());
if (matcher.find()) {
penalty = Integer.parseInt(matcher.group(1));
}
for (ITMXEntry tmen : en.getValue().getEntries()) {
checkStopped(stop);
if (tmen.getSourceText() == null) {
// Not all TMX entries have a source; in that case there can be no meaningful match, so skip.
continue;
}
if (requiresTranslation && tmen.getTranslationText() == null) {
continue;
}
int tmenPenalty = penalty;
if (tmen.hasPropValue(ExternalTMFactory.TMXLoader.PROP_FOREIGN_MATCH, "true")) {
tmenPenalty += foreignPenalty;
}
processEntry(null, tmen.getSourceText(), tmen.getTranslationText(), NearString.MATCH_SOURCE.TM, false, tmenPenalty, en.getKey(), tmen.getCreator(), tmen.getCreationDate(), tmen.getChanger(), tmen.getChangeDate(), tmen.getProperties());
}
}
// travel by all entries for check source file translations
for (SourceTextEntry ste : project.getAllEntries()) {
checkStopped(stop);
if (ste.getSourceTranslation() != null) {
processEntry(ste.getKey(), ste.getSrcText(), ste.getSourceTranslation(), NearString.MATCH_SOURCE.MEMORY, ste.isSourceTranslationFuzzy(), 0, ste.getKey().file, "", 0, "", 0, null);
}
}
if (separateSegmentMatcher != null) {
// split paragraph even when segmentation disabled, then find matches for every segment
List<StringBuilder> spaces = new ArrayList<StringBuilder>();
List<Rule> brules = new ArrayList<Rule>();
Language sourceLang = project.getProjectProperties().getSourceLanguage();
Language targetLang = project.getProjectProperties().getTargetLanguage();
List<String> segments = Core.getSegmenter().segment(sourceLang, srcText, spaces, brules);
if (segments.size() > 1) {
List<String> fsrc = new ArrayList<String>(segments.size());
List<String> ftrans = new ArrayList<String>(segments.size());
// multiple segments
for (short i = 0; i < segments.size(); i++) {
String onesrc = segments.get(i);
// find match for separate segment
List<NearString> segmentMatch = separateSegmentMatcher.search(onesrc, requiresTranslation, false, stop);
if (!segmentMatch.isEmpty() && segmentMatch.get(0).scores[0].score >= SUBSEGMENT_MATCH_THRESHOLD) {
fsrc.add(segmentMatch.get(0).source);
ftrans.add(segmentMatch.get(0).translation);
} else {
fsrc.add("");
ftrans.add("");
}
}
// glue found sources
String foundSrc = Core.getSegmenter().glue(sourceLang, sourceLang, fsrc, spaces, brules);
// glue found translations
String foundTrans = Core.getSegmenter().glue(sourceLang, targetLang, ftrans, spaces, brules);
processEntry(null, foundSrc, foundTrans, NearString.MATCH_SOURCE.TM, false, 0, "", "", 0, "", 0, null);
}
}
if (fillSimilarityData) {
// fill similarity data only for result
for (NearString near : result) {
// fix for bug 1586397
byte[] similarityData = FuzzyMatcher.buildSimilarityData(strTokensAll, tokenizeAll(near.source));
near.attr = similarityData;
}
}
return result;
}
Aggregations