public void testCompareLists() throws IOException {
AnalyzedSentence sentence1 = langTool.getAnalyzedSentence("Hier ein Test");
assertTrue(
rule.compareLists(
sentence1.getTokensWithoutWhitespace(), 0, 2, new String[] {"", "Hier", "ein"}));
assertTrue(
rule.compareLists(
sentence1.getTokensWithoutWhitespace(), 1, 2, new String[] {"Hier", "ein"}));
assertTrue(
rule.compareLists(
sentence1.getTokensWithoutWhitespace(),
0,
3,
new String[] {"", "Hier", "ein", "Test"}));
assertFalse(
rule.compareLists(
sentence1.getTokensWithoutWhitespace(),
0,
4,
new String[] {"", "Hier", "ein", "Test"}));
AnalyzedSentence sentence2 = langTool.getAnalyzedSentence("das Heilige Römische Reich");
assertTrue(
rule.compareLists(
sentence2.getTokensWithoutWhitespace(),
0,
4,
new String[] {"", "das", "Heilige", "Römische", "Reich"}));
assertFalse(
rule.compareLists(
sentence2.getTokensWithoutWhitespace(),
8,
11,
new String[] {"", "das", "Heilige", "Römische", "Reich"}));
}
private void assertBadWithMessage(String s, String expectedErrorSubstring) throws IOException {
assertEquals(1, rule.match(langTool.getAnalyzedSentence(s)).length);
final String errorMessage = rule.match(langTool.getAnalyzedSentence(s))[0].getMessage();
assertTrue(
"Got error '" + errorMessage + "', expected substring '" + expectedErrorSubstring + "'",
errorMessage.contains(expectedErrorSubstring));
}
public void testRule() throws IOException {
// correct sentences:
// assertCorrect("els etiquetadors sobre els etiquetats.");
assertCorrect("tot tenyit amb llum de nostàlgia");
assertCorrect("Ho van fer per duplicat.");
assertCorrect("Assecat el braç del riu");
assertCorrect("el llibre empaquetat");
assertCorrect("un resultat equilibrat");
assertCorrect("el nostre equip era bastant equilibrat");
assertCorrect("un llibre ben empaquetat");
assertCorrect("l'informe filtrat pel ministre");
assertCorrect("L'informe filtrat és terrible");
assertCorrect("ha liderat la batalla");
assertCorrect("Els tinc empaquetats");
assertCorrect("amb tractament unitari i equilibrat");
assertCorrect("Processat després de la mort de Carles II");
assertCorrect("Processat diverses vegades");
assertCorrect("moltes vegades empaquetat amb pressa");
assertCorrect("és llavors embotellat i llançat al mercat");
assertCorrect("la comercialització de vi embotellat amb les firmes comercials");
assertCorrect("eixia al mercat el vi blanc embotellat amb la marca");
assertCorrect("que arribi a un equilibrat matrimoni");
assertCorrect("És un cafè amb molt de cos i molt equilibrat.");
assertCorrect("i per tant etiquetat com a observat");
assertCorrect("Molt equilibrat en les seves característiques");
assertCorrect("filtrat per Wikileaks");
assertCorrect("una vegada filtrat");
assertCorrect("no equilibrat");
// errors:
assertIncorrect("Assecat del braç del riu");
assertIncorrect("Cal vigilar el filtrat del vi");
assertIncorrect("El procés d'empaquetat");
assertIncorrect("Els equilibrats de les rodes");
// assertIncorrect("Duplicat de claus");
assertIncorrect("El procés d'etiquetat de les ampolles");
assertIncorrect("El rentat de cotes");
RuleMatch[] matches = rule.match(langTool.getAnalyzedSentence("El repicat i el rejuntat."));
assertEquals(2, matches.length);
matches = rule.match(langTool.getAnalyzedSentence("El procés de relligat dels llibres."));
assertEquals(1, matches.length);
assertEquals("relligadura", matches[0].getSuggestedReplacements().get(0));
assertEquals("relligament", matches[0].getSuggestedReplacements().get(1));
assertEquals("relligada", matches[0].getSuggestedReplacements().get(2));
matches = rule.match(langTool.getAnalyzedSentence("Els rentats de cervell."));
assertEquals(1, matches.length);
assertEquals("rentades", matches[0].getSuggestedReplacements().get(0));
assertEquals("rentatges", matches[0].getSuggestedReplacements().get(1));
assertEquals("rentaments", matches[0].getSuggestedReplacements().get(2));
}
@Test
public void testRule() throws IOException {
final MixedAlphabetsRule rule = new MixedAlphabetsRule(TestTools.getMessages("uk"));
final JLanguageTool langTool = new JLanguageTool(new Ukrainian());
// correct sentences:
assertEquals(0, rule.match(langTool.getAnalyzedSentence("сміття")).length);
// incorrect sentences:
RuleMatch[] matches = rule.match(langTool.getAnalyzedSentence("смi\u00ADття"));
// check match positions:
assertEquals(1, matches.length);
assertEquals(Arrays.asList("сміття"), matches[0].getSuggestedReplacements());
}
private static void tagText(final String contents, final JLanguageTool lt) throws IOException {
AnalyzedSentence analyzedText;
final List<String> sentences = lt.sentenceTokenize(contents);
for (final String sentence : sentences) {
analyzedText = lt.getAnalyzedSentence(sentence);
System.out.println(getSentence(analyzedText));
}
}
public void testChunker() throws Exception {
JLanguageTool lt = new JLanguageTool(new Ukrainian());
AnalyzedSentence analyzedSentence = lt.getAnalyzedSentence("Для годиться.");
AnalyzedSentence disambiguated = chunker.disambiguate(analyzedSentence);
AnalyzedTokenReadings[] tokens = disambiguated.getTokens();
assertTrue(tokens[1].getReadings().toString().contains("<adv>"));
assertTrue(tokens[4].getReadings().toString().contains("</adv>"));
}
private void assertBad(String s, String... expectedSuggestions) throws IOException {
RuleMatch[] matches = rule.match(langTool.getAnalyzedSentence(s));
assertEquals("Did not find one match in sentence '" + s + "'", 1, matches.length);
if (expectedSuggestions.length > 0) {
RuleMatch match = matches[0];
List<String> suggestions = match.getSuggestedReplacements();
assertThat(suggestions, is(Arrays.asList(expectedSuggestions)));
}
}
public void testPositions() throws IOException {
final AccentuationCheckRule rule = new AccentuationCheckRule(TestTools.getEnglishMessages());
final RuleMatch[] matches;
final JLanguageTool langTool = new JLanguageTool(new Catalan());
matches = rule.match(langTool.getAnalyzedSentence("Són circumstancies extraordinàries."));
assertEquals(4, matches[0].getFromPos());
assertEquals(18, matches[0].getToPos());
}
@Test
public void testRule() throws IOException {
DifferentPunctuationRule rule = new DifferentPunctuationRule();
RuleMatch[] matches;
JLanguageTool srcLangTool = new JLanguageTool(TestTools.getDemoLanguage());
JLanguageTool trgLangTool = new JLanguageTool(new FakeLanguage());
rule.setSourceLanguage(TestTools.getDemoLanguage());
// correct sentences:
matches =
rule.match(
srcLangTool.getAnalyzedSentence("This is a test sentence!"),
trgLangTool.getAnalyzedSentence("C'est la vie!"));
assertEquals(0, matches.length);
matches =
rule.match(
srcLangTool.getAnalyzedSentence("one sentence"),
trgLangTool.getAnalyzedSentence("jedno zdanie"));
assertEquals(0, matches.length);
// incorrect sentences:
matches =
rule.match(
srcLangTool.getAnalyzedSentence("This this is a test sentence."),
trgLangTool.getAnalyzedSentence("This this is a test sentence!"));
assertEquals(1, matches.length);
}
private void assertBad(String s, int n, String... expectedSuggestions) throws IOException {
RuleMatch[] matches = rule.match(langTool.getAnalyzedSentence(s));
assertEquals("Did not find " + n + " match(es) in sentence '" + s + "'", n, matches.length);
if (expectedSuggestions.length > 0) {
RuleMatch match = matches[0];
// When two errors are reported by the rule (so TODO above), it might happen that the first
// match does not have the suggestions, but the second one
if (matches.length > 1 && match.getSuggestedReplacements().size() == 0) {
match = matches[1];
}
List<String> suggestions = match.getSuggestedReplacements();
assertThat(suggestions, is(Arrays.asList(expectedSuggestions)));
}
}
private List<MatchingSentence> findMatchingSentences(
IndexSearcher indexSearcher, TopDocs topDocs, JLanguageTool languageTool) throws IOException {
final List<MatchingSentence> matchingSentences = new ArrayList<>();
for (ScoreDoc match : topDocs.scoreDocs) {
final Document doc = indexSearcher.doc(match.doc);
final String sentence = doc.get(FIELD_NAME);
final List<RuleMatch> ruleMatches = languageTool.check(sentence);
if (ruleMatches.size() > 0) {
final String source = doc.get(SOURCE_FIELD_NAME);
final AnalyzedSentence analyzedSentence = languageTool.getAnalyzedSentence(sentence);
final MatchingSentence matchingSentence =
new MatchingSentence(sentence, source, analyzedSentence, ruleMatches);
matchingSentences.add(matchingSentence);
}
}
return matchingSentences;
}
private void assertGood(String s) throws IOException {
assertEquals(
"Found unexpected match in sentence '" + s + "'",
0,
rule.match(langTool.getAnalyzedSentence(s)).length);
}
private void assertBad(String input) throws IOException {
assertEquals(
"Did not find expected error in: '" + input + "'",
1,
rule.match(langTool.getAnalyzedSentence(input)).length);
}
private void assertCorrect(String sentence) throws IOException {
final RuleMatch[] matches = rule.match(langTool.getAnalyzedSentence(sentence));
assertEquals(0, matches.length);
}
private void assertBad(String s, int n) throws IOException {
assertEquals(n, rule.match(langTool.getAnalyzedSentence(s)).length);
}
public void testRule() throws IOException {
// correct sentences:
assertCorrect("(l'execució a mans d'\"especialistes\")");
assertCorrect("\"Vine\", li va dir.");
assertCorrect("(Una frase de prova).");
assertCorrect("Aquesta és la paraula 'prova'.");
assertCorrect("This is a sentence with a smiley :-)");
assertCorrect("This is a sentence with a smiley ;-) and so on...");
assertCorrect("Aquesta és l'hora de les decisions.");
assertCorrect("Aquesta és l’hora de les decisions.");
assertCorrect("(fig. 20)");
assertCorrect("\"Sóc la teva filla. El corcó no et rosegarà més.\"\n\n");
assertCorrect("–\"Club dels llagoters\" –va repetir en Ron.");
assertCorrect("—\"Club dels llagoters\" –va repetir en Ron.");
assertCorrect("»Això em porta a demanar-t'ho.");
assertCorrect("»Això em porta (sí) a demanar-t'ho.");
assertCorrect("al capítol 12 \"Llavors i fruits oleaginosos\"");
assertCorrect("\"Per què serveixen les forquilles?\" i aquest respon \"per menjar\".");
assertCorrect("És a 60º 50' 23\"");
assertCorrect("És a 60º 50' 23'");
assertCorrect("60° 50' 23'");
assertCorrect("60° 50'");
// assertCorrect("el grau en 60 parts iguals, tenim el minut (1'):");
// assertCorrect("el minut en 60 parts iguals, tenim el segon (1\"):");
// assertCorrect("The screen is 20\" wide.");
assertCorrect("This is a [test] sentence...");
assertCorrect(
"The plight of Tamil refugees caused a surge of support from most of the Tamil political parties.[90]");
assertCorrect("This is what he said: \"We believe in freedom. This is what we do.\"");
assertCorrect("(([20] [20] [20]))");
// test for a case that created a false alarm after disambiguation
assertCorrect("This is a \"special test\", right?");
// numerical bullets
assertCorrect("We discussed this in Chapter 1).");
assertCorrect("The jury recommended that: (1) Four additional deputies be employed.");
assertCorrect("We discussed this in section 1a).");
assertCorrect("We discussed this in section iv).");
// inches exception shouldn't match " here:
assertCorrect(
"In addition, the government would pay a $1,000 \"cost of education\" grant to the schools.");
// assertCorrect("Paradise lost to the alleged water needs of Texas' big cities Thursday.");
assertCorrect("Porta'l cap ací.");
assertCorrect("Porta-me'n cinquanta!");
// incorrect sentences:
assertIncorrect("Ploraria.\"");
assertIncorrect("Aquesta és l555’hora de les decisions.");
assertIncorrect("Vine\", li va dir.");
assertIncorrect("Aquesta és l‘hora de les decisions.");
assertIncorrect("(This is a test sentence.");
assertIncorrect("This is a test with an apostrophe &'.");
assertIncorrect("&'");
assertIncorrect("!'");
assertIncorrect("What?'");
// this is currently considered incorrect... although people often use smileys this way:
assertIncorrect("Some text (and some funny remark :-) with more text to follow");
RuleMatch[] matches;
matches = rule.match(langTool.getAnalyzedSentence("(This is a test” sentence."));
assertEquals(2, matches.length);
matches = rule.match(langTool.getAnalyzedSentence("This [is (a test} sentence."));
assertEquals(3, matches.length);
}