public void testLeadingWildcardType() throws Exception {
AqpQueryParser qp = getParser(null);
qp.setAllowLeadingWildcard(true);
assertEquals(WildcardQuery.class, qp.parse("t*erm*", "field").getClass());
assertEquals(WildcardQuery.class, qp.parse("?term*", "field").getClass());
assertEquals(WildcardQuery.class, qp.parse("*term*", "field").getClass());
}
public void testStopwords() throws Exception {
AqpQueryParser qp = getParser();
qp.setAnalyzer(
new StopAnalyzer(
TEST_VERSION_CURRENT, StopFilter.makeStopSet(TEST_VERSION_CURRENT, "the", "foo")));
Query result = qp.parse("a:the OR a:foo", "a");
assertNotNull("result is null and it shouldn't be", result);
assertTrue("result is not a BooleanQuery", result instanceof BooleanQuery);
assertTrue(
((BooleanQuery) result).clauses().size() + " does not equal: " + 0,
((BooleanQuery) result).clauses().size() == 0);
result = qp.parse("a:woo OR a:the", "a");
assertNotNull("result is null and it shouldn't be", result);
assertTrue("result is not a TermQuery", result instanceof TermQuery);
result = qp.parse("(fieldX:xxxxx OR fieldy:xxxxxxxx)^2 AND (fieldx:the OR fieldy:foo)", "a");
assertNotNull("result is null and it shouldn't be", result);
assertTrue("result is not a BooleanQuery", result instanceof BooleanQuery);
if (VERBOSE) System.out.println("Result: " + result);
assertTrue(
((BooleanQuery) result).clauses().size() + " does not equal: " + 2,
((BooleanQuery) result).clauses().size() == 2);
}
public static AqpQueryParser init(Analyzer a) throws Exception {
AqpQueryParser p = AqpStandardLuceneParser.init();
((QueryNodeProcessorPipeline) p.getQueryNodeProcessor())
.add(new QPTestParserQueryNodeProcessor());
p.setAnalyzer(a);
return p;
}
public void testMatchAllDocs() throws Exception {
AqpQueryParser qp = getParser();
qp.setAnalyzer(new WhitespaceAnalyzer(TEST_VERSION_CURRENT));
assertEquals(new MatchAllDocsQuery(), qp.parse("*:*", "field"));
assertEquals(new MatchAllDocsQuery(), qp.parse("(*:*)", "field"));
BooleanQuery bq = (BooleanQuery) qp.parse("+*:* -*:*", "field");
assertTrue(bq.getClauses()[0].getQuery() instanceof MatchAllDocsQuery);
assertTrue(bq.getClauses()[1].getQuery() instanceof MatchAllDocsQuery);
}
public void testBoost() throws Exception {
StandardAnalyzer oneStopAnalyzer =
new StandardAnalyzer(TEST_VERSION_CURRENT, Collections.singleton("on"));
AqpQueryParser qp = getParser();
qp.setAnalyzer(oneStopAnalyzer);
Query q = qp.parse("on^1.0", "field");
assertNotNull(q);
q = qp.parse("\"hello\"^2.0", "field");
assertNotNull(q);
assertEquals(q.getBoost(), (float) 2.0, (float) 0.5);
q = qp.parse("hello^2.0", "field");
assertNotNull(q);
assertEquals(q.getBoost(), (float) 2.0, (float) 0.5);
q = qp.parse("\"on\"^1.0", "field");
assertNotNull(q);
AqpQueryParser qp2 = AqpStandardLuceneParser.init();
qp2.setAnalyzer(new StandardAnalyzer(TEST_VERSION_CURRENT));
q = qp2.parse("the^3", "field");
// "the" is a stop word so the result is an empty query:
assertNotNull(q);
assertEquals("", q.toString());
assertEquals(1.0f, q.getBoost(), 0.01f);
}
public void testBooleanQuery() throws Exception {
BooleanQuery.setMaxClauseCount(2);
try {
AqpQueryParser qp = getParser();
qp.setAnalyzer(new WhitespaceAnalyzer(TEST_VERSION_CURRENT));
qp.parse("one two three", "field");
fail("ParseException expected due to too many boolean clauses");
} catch (QueryNodeException expected) {
// too many boolean clauses, so ParseException is expected
}
}
public void testDateRange() throws Exception {
String startDate = getLocalizedDate(2002, 1, 1);
String endDate = getLocalizedDate(2002, 1, 4);
Calendar endDateExpected = new GregorianCalendar();
endDateExpected.clear();
endDateExpected.set(2002, 1, 4, 23, 59, 59);
endDateExpected.set(Calendar.MILLISECOND, 999);
final String defaultField = "default";
final String monthField = "month";
final String hourField = "hour";
AqpQueryParser qp = getParser();
// Don't set any date resolution and verify if DateField is used
assertDateRangeQueryEquals(
qp, defaultField, startDate, endDate, endDateExpected.getTime(), null);
Map<CharSequence, DateTools.Resolution> dateRes =
new HashMap<CharSequence, DateTools.Resolution>();
// set a field specific date resolution
dateRes.put(monthField, DateTools.Resolution.MONTH);
qp.setDateResolution(dateRes);
// DateField should still be used for defaultField
assertDateRangeQueryEquals(
qp, defaultField, startDate, endDate, endDateExpected.getTime(), null);
// set default date resolution to MILLISECOND
qp.setDateResolution(DateTools.Resolution.MILLISECOND);
// set second field specific date resolution
dateRes.put(hourField, DateTools.Resolution.HOUR);
qp.setDateResolution(dateRes);
// for this field no field specific date resolution has been set,
// so verify if the default resolution is used
assertDateRangeQueryEquals(
qp,
defaultField,
startDate,
endDate,
endDateExpected.getTime(),
DateTools.Resolution.MILLISECOND);
// verify if field specific date resolutions are used for these two
// fields
assertDateRangeQueryEquals(
qp, monthField, startDate, endDate, endDateExpected.getTime(), DateTools.Resolution.MONTH);
assertDateRangeQueryEquals(
qp, hourField, startDate, endDate, endDateExpected.getTime(), DateTools.Resolution.HOUR);
}
public void testFarsiRangeCollating() throws Exception {
Directory ramDir = newDirectory();
IndexWriter iw =
new IndexWriter(
ramDir,
newIndexWriterConfig(
TEST_VERSION_CURRENT, new WhitespaceAnalyzer(TEST_VERSION_CURRENT)));
Document doc = new Document();
doc.add(newField("content", "\u0633\u0627\u0628", Field.Store.YES, Field.Index.NOT_ANALYZED));
iw.addDocument(doc);
iw.close();
IndexSearcher is = new IndexSearcher(ramDir, true);
AqpQueryParser qp = getParser();
qp.setAnalyzer(new WhitespaceAnalyzer(TEST_VERSION_CURRENT));
// Neither Java 1.4.2 nor 1.5.0 has Farsi Locale collation available in
// RuleBasedCollator. However, the Arabic Locale seems to order the
// Farsi
// characters properly.
Collator c = Collator.getInstance(new Locale("ar"));
qp.setRangeCollator(c);
// Unicode order would include U+0633 in [ U+062F - U+0698 ], but Farsi
// orders the U+0698 character before the U+0633 character, so the
// single
// index Term below should NOT be returned by a ConstantScoreRangeQuery
// with a Farsi Collator (or an Arabic one for the case when Farsi is
// not
// supported).
// Test ConstantScoreRangeQuery
qp.setMultiTermRewriteMethod(MultiTermQuery.CONSTANT_SCORE_FILTER_REWRITE);
ScoreDoc[] result =
is.search(qp.parse("[ \u062F TO \u0698 ]", "content"), null, 1000).scoreDocs;
assertEquals("The index Term should not be included.", 0, result.length);
result = is.search(qp.parse("[ \u0633 TO \u0638 ]", "content"), null, 1000).scoreDocs;
assertEquals("The index Term should be included.", 1, result.length);
// Test RangeQuery
qp.setMultiTermRewriteMethod(MultiTermQuery.SCORING_BOOLEAN_QUERY_REWRITE);
result = is.search(qp.parse("[ \u062F TO \u0698 ]", "content"), null, 1000).scoreDocs;
assertEquals("The index Term should not be included.", 0, result.length);
result = is.search(qp.parse("[ \u0633 TO \u0638 ]", "content"), null, 1000).scoreDocs;
assertEquals("The index Term should be included.", 1, result.length);
is.close();
ramDir.close();
}
public void testConstantScoreAutoRewrite() throws Exception {
AqpQueryParser qp = AqpStandardLuceneParser.init(getGrammarName());
qp.setAnalyzer(new WhitespaceAnalyzer(TEST_VERSION_CURRENT));
Query q = qp.parse("foo*bar", "field");
assertTrue(q instanceof WildcardQuery);
assertEquals(
MultiTermQuery.CONSTANT_SCORE_AUTO_REWRITE_DEFAULT,
((MultiTermQuery) q).getRewriteMethod());
q = qp.parse("foo*", "field");
assertTrue(q instanceof PrefixQuery);
assertEquals(
MultiTermQuery.CONSTANT_SCORE_AUTO_REWRITE_DEFAULT,
((MultiTermQuery) q).getRewriteMethod());
q = qp.parse("[a TO z]", "field");
assertTrue(q instanceof TermRangeQuery);
assertEquals(
MultiTermQuery.CONSTANT_SCORE_AUTO_REWRITE_DEFAULT,
((MultiTermQuery) q).getRewriteMethod());
}
public void testPositionIncrement() throws Exception {
AqpQueryParser qp = getParser();
qp.setAnalyzer(
new StopAnalyzer(
TEST_VERSION_CURRENT,
StopFilter.makeStopSet(TEST_VERSION_CURRENT, "the", "in", "are", "this")));
qp.setEnablePositionIncrements(true);
String qtxt = "\"the words in poisitions pos02578 are stopped in this phrasequery\"";
// 0 2 5 7 8
int expectedPositions[] = {1, 3, 4, 6, 9};
PhraseQuery pq = (PhraseQuery) qp.parse(qtxt, "a");
// System.out.println("Query text: "+qtxt);
// System.out.println("Result: "+pq);
Term t[] = pq.getTerms();
int pos[] = pq.getPositions();
for (int i = 0; i < t.length; i++) {
// System.out.println(i+". "+t[i]+" pos: "+pos[i]);
assertEquals(
"term " + i + " = " + t[i] + " has wrong term-position!", expectedPositions[i], pos[i]);
}
}
public void testRange() throws Exception {
assertQueryEquals("[ a TO z]", null, "[a TO z]");
assertEquals(
MultiTermQuery.CONSTANT_SCORE_AUTO_REWRITE_DEFAULT,
((TermRangeQuery) getQuery("[ a TO z]", null)).getRewriteMethod());
AqpQueryParser qp = getParser();
qp.setMultiTermRewriteMethod(MultiTermQuery.SCORING_BOOLEAN_QUERY_REWRITE);
assertEquals(
MultiTermQuery.SCORING_BOOLEAN_QUERY_REWRITE,
((TermRangeQuery) qp.parse("[ a TO z]", "field")).getRewriteMethod());
assertQueryEquals("[ a TO z ]", null, "[a TO z]");
assertQueryEquals("{ a TO z}", null, "{a TO z}");
assertQueryEquals("{ a TO z }", null, "{a TO z}");
assertQueryEquals("{ a TO z }^2.0", null, "{a TO z}^2.0");
assertQueryEquals("[ a TO z] OR bar", null, "[a TO z] bar");
assertQueryEquals("[ a TO z] AND bar", null, "+[a TO z] +bar");
assertQueryEquals("( bar blar { a TO z}) ", null, "bar blar {a TO z}");
// the original expected value was: gack (bar blar {a TO z})
assertQueryEquals("gack ( bar blar { a TO z}) ", null, "gack bar blar {a TO z}");
}
/** This test differs from TestPrecedenceQueryParser */
public void testPrecedence() throws Exception {
AqpQueryParser qp1 = getParser();
qp1.setAnalyzer(new WhitespaceAnalyzer(TEST_VERSION_CURRENT));
AqpQueryParser qp2 = getParser();
qp2.setAnalyzer(new WhitespaceAnalyzer(TEST_VERSION_CURRENT));
// TODO: to achieve standard lucene behaviour (no operator precedence)
// modify the GroupQueryNodeProcessor to recognize our new BooleanQN classes
// then do:
QueryNodeProcessorPipeline processor = (QueryNodeProcessorPipeline) qp1.getQueryNodeProcessor();
processor.add(new GroupQueryNodeProcessor());
Query query1 = qp1.parse("A AND B OR C AND D", "field");
Query query2 = qp2.parse("+A +B +C +D", "field");
assertEquals(query1, query2);
}
public void testSimple() throws Exception {
assertQueryEquals("\"term germ\"~2", null, "\"term germ\"~2");
assertQueryEquals("term term term", null, "term term term");
assertQueryEquals(
"t�rm term term", new WhitespaceAnalyzer(TEST_VERSION_CURRENT), "t�rm term term");
assertQueryEquals("�mlaut", new WhitespaceAnalyzer(TEST_VERSION_CURRENT), "�mlaut");
// XXX: not allowed, TODO???
// assertQueryEquals("\"\"", new KeywordAnalyzer(), "");
// assertQueryEquals("foo:\"\"", new KeywordAnalyzer(), "foo:");
assertQueryEquals("a AND b", null, "+a +b");
assertQueryEquals("(a AND b)", null, "+a +b");
assertQueryEquals("c OR (a AND b)", null, "c (+a +b)");
assertQueryEquals("a AND NOT b", null, "+a -b");
assertQueryEquals("a NOT b", null, "+a -b");
assertQueryEquals("a AND -b", null, "+a -b");
assertQueryEquals("a AND !b", null, "+a -b");
assertQueryEquals("a && b", null, "+a +b");
assertQueryEquals("a && ! b", null, "+a -b");
assertQueryEquals("a OR b", null, "a b");
assertQueryEquals("a || b", null, "a b");
assertQueryEquals("a OR !b", null, "a -b");
assertQueryEquals("a OR ! b", null, "a -b");
assertQueryEquals("a OR -b", null, "a -b");
assertQueryEquals("+term -term term", null, "+term -term term");
assertQueryEquals("foo:term AND field:anotherTerm", null, "+foo:term +anotherterm");
assertQueryEquals("term AND \"phrase phrase\"", null, "+term +\"phrase phrase\"");
assertQueryEquals("\"hello there\"", null, "\"hello there\"");
assertTrue(getQuery("a AND b", null) instanceof BooleanQuery);
assertTrue(getQuery("hello", null) instanceof TermQuery);
assertTrue(getQuery("\"hello there\"", null) instanceof PhraseQuery);
assertQueryEquals("germ term^2.0", null, "germ term^2.0");
assertQueryEquals("(term)^2.0", null, "term^2.0");
assertQueryEquals("(germ term)^2.0", null, "(germ term)^2.0");
assertQueryEquals("term^2.0", null, "term^2.0");
assertQueryEquals("term^2", null, "term^2.0");
assertQueryEquals("\"germ term\"^2.0", null, "\"germ term\"^2.0");
assertQueryEquals("\"term germ\"^2", null, "\"term germ\"^2.0");
assertQueryEquals("(foo OR bar) AND (baz OR boo)", null, "+(foo bar) +(baz boo)");
assertQueryEquals("((a OR b) AND NOT c) OR d", null, "(+(a b) -c) d");
assertQueryEquals("((a OR b) NOT c) OR d", null, "(+(a b) -c) d");
assertQueryEquals(
"+(apple \"steve jobs\") -(foo bar baz)", null, "+(apple \"steve jobs\") -(foo bar baz)");
assertQueryEquals(
"+title:(dog OR cat) -author:\"bob dole\"",
null,
"+(title:dog title:cat) -author:\"bob dole\"");
AqpQueryParser qp = getParser();
qp.setDefaultOperator(Operator.OR);
assertQueryMatch(qp, "title:(+a -b c)", "text", "+title:a -title:b title:c");
qp.setDefaultOperator(Operator.AND);
assertQueryMatch(qp, "title:(+a -b c)", "text", "+title:a -title:b +title:c");
}
