2 * Licensed to the Apache Software Foundation (ASF) under one or more
3 * contributor license agreements. See the NOTICE file distributed with
4 * this work for additional information regarding copyright ownership.
5 * The ASF licenses this file to You under the Apache License, Version 2.0
6 * (the "License"); you may not use this file except in compliance with
7 * the License. You may obtain a copy of the License at
9 * http://www.apache.org/licenses/LICENSE-2.0
11 * Unless required by applicable law or agreed to in writing, software
12 * distributed under the License is distributed on an "AS IS" BASIS,
13 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
14 * See the License for the specific language governing permissions and
15 * limitations under the License.
20 JAVA_UNICODE_ESCAPE=true;
21 USER_CHAR_STREAM=true;
24 PARSER_BEGIN(QueryParser)
26 package org.apache.lucene.queryParser;
28 import java.io.IOException;
29 import java.io.StringReader;
30 import java.text.Collator;
31 import java.text.DateFormat;
32 import java.util.ArrayList;
33 import java.util.Calendar;
34 import java.util.Date;
35 import java.util.HashMap;
36 import java.util.List;
37 import java.util.Locale;
40 import org.apache.lucene.analysis.Analyzer;
41 import org.apache.lucene.analysis.CachingTokenFilter;
42 import org.apache.lucene.analysis.TokenStream;
43 import org.apache.lucene.analysis.tokenattributes.PositionIncrementAttribute;
44 import org.apache.lucene.analysis.tokenattributes.CharTermAttribute;
45 import org.apache.lucene.document.DateField;
46 import org.apache.lucene.document.DateTools;
47 import org.apache.lucene.index.Term;
48 import org.apache.lucene.search.BooleanClause;
49 import org.apache.lucene.search.BooleanQuery;
50 import org.apache.lucene.search.FuzzyQuery;
51 import org.apache.lucene.search.MultiTermQuery;
52 import org.apache.lucene.search.MatchAllDocsQuery;
53 import org.apache.lucene.search.MultiPhraseQuery;
54 import org.apache.lucene.search.PhraseQuery;
55 import org.apache.lucene.search.PrefixQuery;
56 import org.apache.lucene.search.Query;
57 import org.apache.lucene.search.TermRangeQuery;
58 import org.apache.lucene.search.TermQuery;
59 import org.apache.lucene.search.WildcardQuery;
60 import org.apache.lucene.util.Version;
61 import org.apache.lucene.util.VirtualMethod;
64 * This class is generated by JavaCC. The most important method is
65 * {@link #parse(String)}.
67 * The syntax for query strings is as follows:
68 * A Query is a series of clauses.
69 * A clause may be prefixed by:
71 * <li> a plus (<code>+</code>) or a minus (<code>-</code>) sign, indicating
72 * that the clause is required or prohibited respectively; or
73 * <li> a term followed by a colon, indicating the field to be searched.
74 * This enables one to construct queries which search multiple fields.
77 * A clause may be either:
79 * <li> a term, indicating all the documents that contain this term; or
80 * <li> a nested query, enclosed in parentheses. Note that this may be used
81 * with a <code>+</code>/<code>-</code> prefix to require any of a set of
85 * Thus, in BNF, the query grammar is:
87 * Query ::= ( Clause )*
88 * Clause ::= ["+", "-"] [<TERM> ":"] ( <TERM> | "(" Query ")" )
92 * Examples of appropriately formatted queries can be found in the <a
93 * href="../../../../../../queryparsersyntax.html">query syntax
98 * In {@link TermRangeQuery}s, QueryParser tries to detect date values, e.g.
99 * <tt>date:[6/1/2005 TO 6/4/2005]</tt> produces a range query that searches
100 * for "date" fields between 2005-06-01 and 2005-06-04. Note that the format
101 * of the accepted input depends on {@link #setLocale(Locale) the locale}.
102 * By default a date is converted into a search term using the deprecated
103 * {@link DateField} for compatibility reasons.
104 * To use the new {@link DateTools} to convert dates, a
105 * {@link org.apache.lucene.document.DateTools.Resolution} has to be set.
108 * The date resolution that shall be used for RangeQueries can be set
109 * using {@link #setDateResolution(DateTools.Resolution)}
110 * or {@link #setDateResolution(String, DateTools.Resolution)}. The former
111 * sets the default date resolution for all fields, whereas the latter can
112 * be used to set field specific date resolutions. Field specific date
113 * resolutions take, if set, precedence over the default date resolution.
116 * If you use neither {@link DateField} nor {@link DateTools} in your
117 * index, you can create your own
118 * query parser that inherits QueryParser and overwrites
119 * {@link #getRangeQuery(String, String, String, boolean)} to
120 * use a different method for date conversion.
123 * <p>Note that QueryParser is <em>not</em> thread-safe.</p>
125 * <p><b>NOTE</b>: there is a new QueryParser in contrib, which matches
126 * the same syntax as this class, but is more modular,
127 * enabling substantial customization to how a query is created.
129 * <a name="version"/>
130 * <p><b>NOTE</b>: You must specify the required {@link Version}
131 * compatibility when creating QueryParser:
133 * <li> As of 2.9, {@link #setEnablePositionIncrements} is true by
135 * <li> As of 3.1, {@link #setAutoGeneratePhraseQueries} is false by
139 public class QueryParser {
141 private static final int CONJ_NONE = 0;
142 private static final int CONJ_AND = 1;
143 private static final int CONJ_OR = 2;
145 private static final int MOD_NONE = 0;
146 private static final int MOD_NOT = 10;
147 private static final int MOD_REQ = 11;
149 // make it possible to call setDefaultOperator() without accessing
151 /** Alternative form of QueryParser.Operator.AND */
152 public static final Operator AND_OPERATOR = Operator.AND;
153 /** Alternative form of QueryParser.Operator.OR */
154 public static final Operator OR_OPERATOR = Operator.OR;
156 /** The actual operator that parser uses to combine query terms */
157 private Operator operator = OR_OPERATOR;
159 boolean lowercaseExpandedTerms = true;
160 MultiTermQuery.RewriteMethod multiTermRewriteMethod = MultiTermQuery.CONSTANT_SCORE_AUTO_REWRITE_DEFAULT;
161 boolean allowLeadingWildcard = false;
162 boolean enablePositionIncrements = true;
167 float fuzzyMinSim = FuzzyQuery.defaultMinSimilarity;
168 int fuzzyPrefixLength = FuzzyQuery.defaultPrefixLength;
169 Locale locale = Locale.getDefault();
171 // the default date resolution
172 DateTools.Resolution dateResolution = null;
173 // maps field names to date resolutions
174 Map<String,DateTools.Resolution> fieldToDateResolution = null;
176 // The collator to use when determining range inclusion,
177 // for use when constructing RangeQuerys.
178 Collator rangeCollator = null;
180 /** @deprecated remove when getFieldQuery is removed */
182 private static final VirtualMethod<QueryParser> getFieldQueryMethod =
183 new VirtualMethod<QueryParser>(QueryParser.class, "getFieldQuery", String.class, String.class);
184 /** @deprecated remove when getFieldQuery is removed */
186 private static final VirtualMethod<QueryParser> getFieldQueryWithQuotedMethod =
187 new VirtualMethod<QueryParser>(QueryParser.class, "getFieldQuery", String.class, String.class, boolean.class);
188 /** @deprecated remove when getFieldQuery is removed */
190 private final boolean hasNewAPI =
191 VirtualMethod.compareImplementationDistance(getClass(),
192 getFieldQueryWithQuotedMethod, getFieldQueryMethod) >= 0; // its ok for both to be overridden
194 private boolean autoGeneratePhraseQueries;
196 /** The default operator for parsing queries.
197 * Use {@link QueryParser#setDefaultOperator} to change it.
199 static public enum Operator { OR, AND }
201 /** Constructs a query parser.
202 * @param matchVersion Lucene version to match. See <a href="#version">above</a>.
203 * @param f the default field for query terms.
204 * @param a used to find terms in the query text.
206 public QueryParser(Version matchVersion, String f, Analyzer a) {
207 this(new FastCharStream(new StringReader("")));
210 if (matchVersion.onOrAfter(Version.LUCENE_29)) {
211 enablePositionIncrements = true;
213 enablePositionIncrements = false;
215 if (matchVersion.onOrAfter(Version.LUCENE_31)) {
216 setAutoGeneratePhraseQueries(false);
218 setAutoGeneratePhraseQueries(true);
222 /** Parses a query string, returning a {@link org.apache.lucene.search.Query}.
223 * @param query the query string to be parsed.
224 * @throws ParseException if the parsing fails
226 public Query parse(String query) throws ParseException {
227 ReInit(new FastCharStream(new StringReader(query)));
229 // TopLevelQuery is a Query followed by the end-of-input (EOF)
230 Query res = TopLevelQuery(field);
231 return res!=null ? res : newBooleanQuery(false);
233 catch (ParseException tme) {
234 // rethrow to include the original query:
235 ParseException e = new ParseException("Cannot parse '" +query+ "': " + tme.getMessage());
239 catch (TokenMgrError tme) {
240 ParseException e = new ParseException("Cannot parse '" +query+ "': " + tme.getMessage());
244 catch (BooleanQuery.TooManyClauses tmc) {
245 ParseException e = new ParseException("Cannot parse '" +query+ "': too many boolean clauses");
252 * @return Returns the analyzer.
254 public Analyzer getAnalyzer() {
259 * @return Returns the field.
261 public String getField() {
266 * @see #setAutoGeneratePhraseQueries(boolean)
268 public final boolean getAutoGeneratePhraseQueries() {
269 return autoGeneratePhraseQueries;
273 * Set to true if phrase queries will be automatically generated
274 * when the analyzer returns more than one term from whitespace
276 * NOTE: this behavior may not be suitable for all languages.
278 * Set to false if phrase queries should only be generated when
279 * surrounded by double quotes.
281 public final void setAutoGeneratePhraseQueries(boolean value) {
282 if (value == false && !hasNewAPI)
283 throw new IllegalArgumentException("You must implement the new API: getFieldQuery(String,String,boolean)"
284 + " to use setAutoGeneratePhraseQueries(false)");
285 this.autoGeneratePhraseQueries = value;
289 * Get the minimal similarity for fuzzy queries.
291 public float getFuzzyMinSim() {
296 * Set the minimum similarity for fuzzy queries.
299 public void setFuzzyMinSim(float fuzzyMinSim) {
300 this.fuzzyMinSim = fuzzyMinSim;
304 * Get the prefix length for fuzzy queries.
305 * @return Returns the fuzzyPrefixLength.
307 public int getFuzzyPrefixLength() {
308 return fuzzyPrefixLength;
312 * Set the prefix length for fuzzy queries. Default is 0.
313 * @param fuzzyPrefixLength The fuzzyPrefixLength to set.
315 public void setFuzzyPrefixLength(int fuzzyPrefixLength) {
316 this.fuzzyPrefixLength = fuzzyPrefixLength;
320 * Sets the default slop for phrases. If zero, then exact phrase matches
321 * are required. Default value is zero.
323 public void setPhraseSlop(int phraseSlop) {
324 this.phraseSlop = phraseSlop;
328 * Gets the default slop for phrases.
330 public int getPhraseSlop() {
336 * Set to <code>true</code> to allow leading wildcard characters.
338 * When set, <code>*</code> or <code>?</code> are allowed as
339 * the first character of a PrefixQuery and WildcardQuery.
340 * Note that this can produce very slow
341 * queries on big indexes.
345 public void setAllowLeadingWildcard(boolean allowLeadingWildcard) {
346 this.allowLeadingWildcard = allowLeadingWildcard;
350 * @see #setAllowLeadingWildcard(boolean)
352 public boolean getAllowLeadingWildcard() {
353 return allowLeadingWildcard;
357 * Set to <code>true</code> to enable position increments in result query.
359 * When set, result phrase and multi-phrase queries will
360 * be aware of position increments.
361 * Useful when e.g. a StopFilter increases the position increment of
362 * the token that follows an omitted token.
366 public void setEnablePositionIncrements(boolean enable) {
367 this.enablePositionIncrements = enable;
371 * @see #setEnablePositionIncrements(boolean)
373 public boolean getEnablePositionIncrements() {
374 return enablePositionIncrements;
378 * Sets the boolean operator of the QueryParser.
379 * In default mode (<code>OR_OPERATOR</code>) terms without any modifiers
380 * are considered optional: for example <code>capital of Hungary</code> is equal to
381 * <code>capital OR of OR Hungary</code>.<br/>
382 * In <code>AND_OPERATOR</code> mode terms are considered to be in conjunction: the
383 * above mentioned query is parsed as <code>capital AND of AND Hungary</code>
385 public void setDefaultOperator(Operator op) {
391 * Gets implicit operator setting, which will be either AND_OPERATOR
394 public Operator getDefaultOperator() {
400 * Whether terms of wildcard, prefix, fuzzy and range queries are to be automatically
401 * lower-cased or not. Default is <code>true</code>.
403 public void setLowercaseExpandedTerms(boolean lowercaseExpandedTerms) {
404 this.lowercaseExpandedTerms = lowercaseExpandedTerms;
409 * @see #setLowercaseExpandedTerms(boolean)
411 public boolean getLowercaseExpandedTerms() {
412 return lowercaseExpandedTerms;
416 * By default QueryParser uses {@link MultiTermQuery#CONSTANT_SCORE_AUTO_REWRITE_DEFAULT}
417 * when creating a PrefixQuery, WildcardQuery or RangeQuery. This implementation is generally preferable because it
418 * a) Runs faster b) Does not have the scarcity of terms unduly influence score
419 * c) avoids any "TooManyBooleanClauses" exception.
420 * However, if your application really needs to use the
421 * old-fashioned BooleanQuery expansion rewriting and the above
422 * points are not relevant then use this to change
423 * the rewrite method.
425 public void setMultiTermRewriteMethod(MultiTermQuery.RewriteMethod method) {
426 multiTermRewriteMethod = method;
431 * @see #setMultiTermRewriteMethod
433 public MultiTermQuery.RewriteMethod getMultiTermRewriteMethod() {
434 return multiTermRewriteMethod;
438 * Set locale used by date range parsing.
440 public void setLocale(Locale locale) {
441 this.locale = locale;
445 * Returns current locale, allowing access by subclasses.
447 public Locale getLocale() {
452 * Sets the default date resolution used by RangeQueries for fields for which no
453 * specific date resolutions has been set. Field specific resolutions can be set
454 * with {@link #setDateResolution(String, DateTools.Resolution)}.
456 * @param dateResolution the default date resolution to set
458 public void setDateResolution(DateTools.Resolution dateResolution) {
459 this.dateResolution = dateResolution;
463 * Sets the date resolution used by RangeQueries for a specific field.
465 * @param fieldName field for which the date resolution is to be set
466 * @param dateResolution date resolution to set
468 public void setDateResolution(String fieldName, DateTools.Resolution dateResolution) {
469 if (fieldName == null) {
470 throw new IllegalArgumentException("Field cannot be null.");
473 if (fieldToDateResolution == null) {
474 // lazily initialize HashMap
475 fieldToDateResolution = new HashMap<String,DateTools.Resolution>();
478 fieldToDateResolution.put(fieldName, dateResolution);
482 * Returns the date resolution that is used by RangeQueries for the given field.
483 * Returns null, if no default or field specific date resolution has been set
484 * for the given field.
487 public DateTools.Resolution getDateResolution(String fieldName) {
488 if (fieldName == null) {
489 throw new IllegalArgumentException("Field cannot be null.");
492 if (fieldToDateResolution == null) {
493 // no field specific date resolutions set; return default date resolution instead
494 return this.dateResolution;
497 DateTools.Resolution resolution = fieldToDateResolution.get(fieldName);
498 if (resolution == null) {
499 // no date resolutions set for the given field; return default date resolution instead
500 resolution = this.dateResolution;
507 * Sets the collator used to determine index term inclusion in ranges
510 * <strong>WARNING:</strong> Setting the rangeCollator to a non-null
511 * collator using this method will cause every single index Term in the
512 * Field referenced by lowerTerm and/or upperTerm to be examined.
513 * Depending on the number of index Terms in this Field, the operation could
516 * @param rc the collator to use when constructing RangeQuerys
518 public void setRangeCollator(Collator rc) {
523 * @return the collator used to determine index term inclusion in ranges
526 public Collator getRangeCollator() {
527 return rangeCollator;
530 protected void addClause(List<BooleanClause> clauses, int conj, int mods, Query q) {
531 boolean required, prohibited;
533 // If this term is introduced by AND, make the preceding term required,
534 // unless it's already prohibited
535 if (clauses.size() > 0 && conj == CONJ_AND) {
536 BooleanClause c = clauses.get(clauses.size()-1);
537 if (!c.isProhibited())
538 c.setOccur(BooleanClause.Occur.MUST);
541 if (clauses.size() > 0 && operator == AND_OPERATOR && conj == CONJ_OR) {
542 // If this term is introduced by OR, make the preceding term optional,
543 // unless it's prohibited (that means we leave -a OR b but +a OR b-->a OR b)
544 // notice if the input is a OR b, first term is parsed as required; without
545 // this modification a OR b would parsed as +a OR b
546 BooleanClause c = clauses.get(clauses.size()-1);
547 if (!c.isProhibited())
548 c.setOccur(BooleanClause.Occur.SHOULD);
551 // We might have been passed a null query; the term might have been
552 // filtered away by the analyzer.
556 if (operator == OR_OPERATOR) {
557 // We set REQUIRED if we're introduced by AND or +; PROHIBITED if
558 // introduced by NOT or -; make sure not to set both.
559 prohibited = (mods == MOD_NOT);
560 required = (mods == MOD_REQ);
561 if (conj == CONJ_AND && !prohibited) {
565 // We set PROHIBITED if we're introduced by NOT or -; We set REQUIRED
566 // if not PROHIBITED and not introduced by OR
567 prohibited = (mods == MOD_NOT);
568 required = (!prohibited && conj != CONJ_OR);
570 if (required && !prohibited)
571 clauses.add(newBooleanClause(q, BooleanClause.Occur.MUST));
572 else if (!required && !prohibited)
573 clauses.add(newBooleanClause(q, BooleanClause.Occur.SHOULD));
574 else if (!required && prohibited)
575 clauses.add(newBooleanClause(q, BooleanClause.Occur.MUST_NOT));
577 throw new RuntimeException("Clause cannot be both required and prohibited");
581 * @deprecated Use {@link #getFieldQuery(String,String,boolean)} instead.
584 protected Query getFieldQuery(String field, String queryText) throws ParseException {
585 // treat the text as if it was quoted, to drive phrase logic with old versions.
586 return getFieldQuery(field, queryText, true);
590 * @exception ParseException throw in overridden method to disallow
592 protected Query getFieldQuery(String field, String queryText, boolean quoted) throws ParseException {
593 // Use the analyzer to get all the tokens, and then build a TermQuery,
594 // PhraseQuery, or nothing based on the term count
598 source = analyzer.reusableTokenStream(field, new StringReader(queryText));
600 } catch (IOException e) {
601 source = analyzer.tokenStream(field, new StringReader(queryText));
603 CachingTokenFilter buffer = new CachingTokenFilter(source);
604 CharTermAttribute termAtt = null;
605 PositionIncrementAttribute posIncrAtt = null;
608 boolean success = false;
612 } catch (IOException e) {
613 // success==false if we hit an exception
616 if (buffer.hasAttribute(CharTermAttribute.class)) {
617 termAtt = buffer.getAttribute(CharTermAttribute.class);
619 if (buffer.hasAttribute(PositionIncrementAttribute.class)) {
620 posIncrAtt = buffer.getAttribute(PositionIncrementAttribute.class);
624 int positionCount = 0;
625 boolean severalTokensAtSamePosition = false;
627 boolean hasMoreTokens = false;
628 if (termAtt != null) {
630 hasMoreTokens = buffer.incrementToken();
631 while (hasMoreTokens) {
633 int positionIncrement = (posIncrAtt != null) ? posIncrAtt.getPositionIncrement() : 1;
634 if (positionIncrement != 0) {
635 positionCount += positionIncrement;
637 severalTokensAtSamePosition = true;
639 hasMoreTokens = buffer.incrementToken();
641 } catch (IOException e) {
646 // rewind the buffer stream
649 // close original stream - all tokens buffered
652 catch (IOException e) {
658 else if (numTokens == 1) {
661 boolean hasNext = buffer.incrementToken();
662 assert hasNext == true;
663 term = termAtt.toString();
664 } catch (IOException e) {
665 // safe to ignore, because we know the number of tokens
667 return newTermQuery(new Term(field, term));
669 if (severalTokensAtSamePosition || (!quoted && !autoGeneratePhraseQueries)) {
670 if (positionCount == 1 || (!quoted && !autoGeneratePhraseQueries)) {
672 BooleanQuery q = newBooleanQuery(positionCount == 1);
674 BooleanClause.Occur occur = positionCount > 1 && operator == AND_OPERATOR ?
675 BooleanClause.Occur.MUST : BooleanClause.Occur.SHOULD;
677 for (int i = 0; i < numTokens; i++) {
680 boolean hasNext = buffer.incrementToken();
681 assert hasNext == true;
682 term = termAtt.toString();
683 } catch (IOException e) {
684 // safe to ignore, because we know the number of tokens
687 Query currentQuery = newTermQuery(
688 new Term(field, term));
689 q.add(currentQuery, occur);
695 MultiPhraseQuery mpq = newMultiPhraseQuery();
696 mpq.setSlop(phraseSlop);
697 List<Term> multiTerms = new ArrayList<Term>();
699 for (int i = 0; i < numTokens; i++) {
701 int positionIncrement = 1;
703 boolean hasNext = buffer.incrementToken();
704 assert hasNext == true;
705 term = termAtt.toString();
706 if (posIncrAtt != null) {
707 positionIncrement = posIncrAtt.getPositionIncrement();
709 } catch (IOException e) {
710 // safe to ignore, because we know the number of tokens
713 if (positionIncrement > 0 && multiTerms.size() > 0) {
714 if (enablePositionIncrements) {
715 mpq.add(multiTerms.toArray(new Term[0]),position);
717 mpq.add(multiTerms.toArray(new Term[0]));
721 position += positionIncrement;
722 multiTerms.add(new Term(field, term));
724 if (enablePositionIncrements) {
725 mpq.add(multiTerms.toArray(new Term[0]),position);
727 mpq.add(multiTerms.toArray(new Term[0]));
733 PhraseQuery pq = newPhraseQuery();
734 pq.setSlop(phraseSlop);
738 for (int i = 0; i < numTokens; i++) {
740 int positionIncrement = 1;
743 boolean hasNext = buffer.incrementToken();
744 assert hasNext == true;
745 term = termAtt.toString();
746 if (posIncrAtt != null) {
747 positionIncrement = posIncrAtt.getPositionIncrement();
749 } catch (IOException e) {
750 // safe to ignore, because we know the number of tokens
753 if (enablePositionIncrements) {
754 position += positionIncrement;
755 pq.add(new Term(field, term),position);
757 pq.add(new Term(field, term));
768 * Base implementation delegates to {@link #getFieldQuery(String,String,boolean)}.
769 * This method may be overridden, for example, to return
770 * a SpanNearQuery instead of a PhraseQuery.
772 * @exception ParseException throw in overridden method to disallow
774 protected Query getFieldQuery(String field, String queryText, int slop)
775 throws ParseException {
776 Query query = hasNewAPI ? getFieldQuery(field, queryText, true) : getFieldQuery(field, queryText);
778 if (query instanceof PhraseQuery) {
779 ((PhraseQuery) query).setSlop(slop);
781 if (query instanceof MultiPhraseQuery) {
782 ((MultiPhraseQuery) query).setSlop(slop);
790 * @exception ParseException throw in overridden method to disallow
792 protected Query getRangeQuery(String field,
795 boolean inclusive) throws ParseException
797 if (lowercaseExpandedTerms) {
798 part1 = part1.toLowerCase();
799 part2 = part2.toLowerCase();
802 DateFormat df = DateFormat.getDateInstance(DateFormat.SHORT, locale);
804 Date d1 = df.parse(part1);
805 Date d2 = df.parse(part2);
807 // The user can only specify the date, not the time, so make sure
808 // the time is set to the latest possible time of that date to really
809 // include all documents:
810 Calendar cal = Calendar.getInstance(locale);
812 cal.set(Calendar.HOUR_OF_DAY, 23);
813 cal.set(Calendar.MINUTE, 59);
814 cal.set(Calendar.SECOND, 59);
815 cal.set(Calendar.MILLISECOND, 999);
818 DateTools.Resolution resolution = getDateResolution(field);
819 if (resolution == null) {
820 // no default or field specific date resolution has been set,
821 // use deprecated DateField to maintain compatibility with
822 // pre-1.9 Lucene versions.
823 part1 = DateField.dateToString(d1);
824 part2 = DateField.dateToString(d2);
826 part1 = DateTools.dateToString(d1, resolution);
827 part2 = DateTools.dateToString(d2, resolution);
830 catch (Exception e) { }
832 return newRangeQuery(field, part1, part2, inclusive);
836 * Builds a new BooleanQuery instance
837 * @param disableCoord disable coord
838 * @return new BooleanQuery instance
840 protected BooleanQuery newBooleanQuery(boolean disableCoord) {
841 return new BooleanQuery(disableCoord);
845 * Builds a new BooleanClause instance
847 * @param occur how this clause should occur when matching documents
848 * @return new BooleanClause instance
850 protected BooleanClause newBooleanClause(Query q, BooleanClause.Occur occur) {
851 return new BooleanClause(q, occur);
855 * Builds a new TermQuery instance
857 * @return new TermQuery instance
859 protected Query newTermQuery(Term term){
860 return new TermQuery(term);
864 * Builds a new PhraseQuery instance
865 * @return new PhraseQuery instance
867 protected PhraseQuery newPhraseQuery(){
868 return new PhraseQuery();
872 * Builds a new MultiPhraseQuery instance
873 * @return new MultiPhraseQuery instance
875 protected MultiPhraseQuery newMultiPhraseQuery(){
876 return new MultiPhraseQuery();
880 * Builds a new PrefixQuery instance
881 * @param prefix Prefix term
882 * @return new PrefixQuery instance
884 protected Query newPrefixQuery(Term prefix){
885 PrefixQuery query = new PrefixQuery(prefix);
886 query.setRewriteMethod(multiTermRewriteMethod);
891 * Builds a new FuzzyQuery instance
893 * @param minimumSimilarity minimum similarity
894 * @param prefixLength prefix length
895 * @return new FuzzyQuery Instance
897 protected Query newFuzzyQuery(Term term, float minimumSimilarity, int prefixLength) {
898 // FuzzyQuery doesn't yet allow constant score rewrite
899 return new FuzzyQuery(term,minimumSimilarity,prefixLength);
903 * Builds a new TermRangeQuery instance
907 * @param inclusive true if range is inclusive
908 * @return new TermRangeQuery instance
910 protected Query newRangeQuery(String field, String part1, String part2, boolean inclusive) {
911 final TermRangeQuery query = new TermRangeQuery(field, part1, part2, inclusive, inclusive, rangeCollator);
912 query.setRewriteMethod(multiTermRewriteMethod);
917 * Builds a new MatchAllDocsQuery instance
918 * @return new MatchAllDocsQuery instance
920 protected Query newMatchAllDocsQuery() {
921 return new MatchAllDocsQuery();
925 * Builds a new WildcardQuery instance
926 * @param t wildcard term
927 * @return new WildcardQuery instance
929 protected Query newWildcardQuery(Term t) {
930 WildcardQuery query = new WildcardQuery(t);
931 query.setRewriteMethod(multiTermRewriteMethod);
936 * Factory method for generating query, given a set of clauses.
937 * By default creates a boolean query composed of clauses passed in.
939 * Can be overridden by extending classes, to modify query being
942 * @param clauses List that contains {@link BooleanClause} instances
945 * @return Resulting {@link Query} object.
946 * @exception ParseException throw in overridden method to disallow
948 protected Query getBooleanQuery(List<BooleanClause> clauses) throws ParseException {
949 return getBooleanQuery(clauses, false);
953 * Factory method for generating query, given a set of clauses.
954 * By default creates a boolean query composed of clauses passed in.
956 * Can be overridden by extending classes, to modify query being
959 * @param clauses List that contains {@link BooleanClause} instances
961 * @param disableCoord true if coord scoring should be disabled.
963 * @return Resulting {@link Query} object.
964 * @exception ParseException throw in overridden method to disallow
966 protected Query getBooleanQuery(List<BooleanClause> clauses, boolean disableCoord)
967 throws ParseException
969 if (clauses.size()==0) {
970 return null; // all clause words were filtered away by the analyzer.
972 BooleanQuery query = newBooleanQuery(disableCoord);
973 for(final BooleanClause clause: clauses) {
980 * Factory method for generating a query. Called when parser
981 * parses an input term token that contains one or more wildcard
982 * characters (? and *), but is not a prefix term token (one
983 * that has just a single * character at the end)
985 * Depending on settings, prefix term may be lower-cased
986 * automatically. It will not go through the default Analyzer,
987 * however, since normal Analyzers are unlikely to work properly
988 * with wildcard templates.
990 * Can be overridden by extending classes, to provide custom handling for
991 * wildcard queries, which may be necessary due to missing analyzer calls.
993 * @param field Name of the field query will use.
994 * @param termStr Term token that contains one or more wild card
995 * characters (? or *), but is not simple prefix term
997 * @return Resulting {@link Query} built for the term
998 * @exception ParseException throw in overridden method to disallow
1000 protected Query getWildcardQuery(String field, String termStr) throws ParseException
1002 if ("*".equals(field)) {
1003 if ("*".equals(termStr)) return newMatchAllDocsQuery();
1005 if (!allowLeadingWildcard && (termStr.startsWith("*") || termStr.startsWith("?")))
1006 throw new ParseException("'*' or '?' not allowed as first character in WildcardQuery");
1007 if (lowercaseExpandedTerms) {
1008 termStr = termStr.toLowerCase();
1010 Term t = new Term(field, termStr);
1011 return newWildcardQuery(t);
1015 * Factory method for generating a query (similar to
1016 * {@link #getWildcardQuery}). Called when parser parses an input term
1017 * token that uses prefix notation; that is, contains a single '*' wildcard
1018 * character as its last character. Since this is a special case
1019 * of generic wildcard term, and such a query can be optimized easily,
1020 * this usually results in a different query object.
1022 * Depending on settings, a prefix term may be lower-cased
1023 * automatically. It will not go through the default Analyzer,
1024 * however, since normal Analyzers are unlikely to work properly
1025 * with wildcard templates.
1027 * Can be overridden by extending classes, to provide custom handling for
1028 * wild card queries, which may be necessary due to missing analyzer calls.
1030 * @param field Name of the field query will use.
1031 * @param termStr Term token to use for building term for the query
1032 * (<b>without</b> trailing '*' character!)
1034 * @return Resulting {@link Query} built for the term
1035 * @exception ParseException throw in overridden method to disallow
1037 protected Query getPrefixQuery(String field, String termStr) throws ParseException
1039 if (!allowLeadingWildcard && termStr.startsWith("*"))
1040 throw new ParseException("'*' not allowed as first character in PrefixQuery");
1041 if (lowercaseExpandedTerms) {
1042 termStr = termStr.toLowerCase();
1044 Term t = new Term(field, termStr);
1045 return newPrefixQuery(t);
1049 * Factory method for generating a query (similar to
1050 * {@link #getWildcardQuery}). Called when parser parses
1051 * an input term token that has the fuzzy suffix (~) appended.
1053 * @param field Name of the field query will use.
1054 * @param termStr Term token to use for building term for the query
1056 * @return Resulting {@link Query} built for the term
1057 * @exception ParseException throw in overridden method to disallow
1059 protected Query getFuzzyQuery(String field, String termStr, float minSimilarity) throws ParseException
1061 if (lowercaseExpandedTerms) {
1062 termStr = termStr.toLowerCase();
1064 Term t = new Term(field, termStr);
1065 return newFuzzyQuery(t, minSimilarity, fuzzyPrefixLength);
1069 * Returns a String where the escape char has been
1070 * removed, or kept only once if there was a double escape.
1072 * Supports escaped unicode characters, e. g. translates
1073 * <code>\\u0041</code> to <code>A</code>.
1076 private String discardEscapeChar(String input) throws ParseException {
1077 // Create char array to hold unescaped char sequence
1078 char[] output = new char[input.length()];
1080 // The length of the output can be less than the input
1081 // due to discarded escape chars. This variable holds
1082 // the actual length of the output
1085 // We remember whether the last processed character was
1086 // an escape character
1087 boolean lastCharWasEscapeChar = false;
1089 // The multiplier the current unicode digit must be multiplied with.
1090 // E. g. the first digit must be multiplied with 16^3, the second with 16^2...
1091 int codePointMultiplier = 0;
1093 // Used to calculate the codepoint of the escaped unicode character
1096 for (int i = 0; i < input.length(); i++) {
1097 char curChar = input.charAt(i);
1098 if (codePointMultiplier > 0) {
1099 codePoint += hexToInt(curChar) * codePointMultiplier;
1100 codePointMultiplier >>>= 4;
1101 if (codePointMultiplier == 0) {
1102 output[length++] = (char)codePoint;
1105 } else if (lastCharWasEscapeChar) {
1106 if (curChar == 'u') {
1107 // found an escaped unicode character
1108 codePointMultiplier = 16 * 16 * 16;
1110 // this character was escaped
1111 output[length] = curChar;
1114 lastCharWasEscapeChar = false;
1116 if (curChar == '\\') {
1117 lastCharWasEscapeChar = true;
1119 output[length] = curChar;
1125 if (codePointMultiplier > 0) {
1126 throw new ParseException("Truncated unicode escape sequence.");
1129 if (lastCharWasEscapeChar) {
1130 throw new ParseException("Term can not end with escape character.");
1133 return new String(output, 0, length);
1136 /** Returns the numeric value of the hexadecimal character */
1137 private static final int hexToInt(char c) throws ParseException {
1138 if ('0' <= c && c <= '9') {
1140 } else if ('a' <= c && c <= 'f'){
1141 return c - 'a' + 10;
1142 } else if ('A' <= c && c <= 'F') {
1143 return c - 'A' + 10;
1145 throw new ParseException("None-hex character in unicode escape sequence: " + c);
1150 * Returns a String where those characters that QueryParser
1151 * expects to be escaped are escaped by a preceding <code>\</code>.
1153 public static String escape(String s) {
1154 StringBuilder sb = new StringBuilder();
1155 for (int i = 0; i < s.length(); i++) {
1156 char c = s.charAt(i);
1157 // These characters are part of the query syntax and must be escaped
1158 if (c == '\\' || c == '+' || c == '-' || c == '!' || c == '(' || c == ')' || c == ':'
1159 || c == '^' || c == '[' || c == ']' || c == '\"' || c == '{' || c == '}' || c == '~'
1160 || c == '*' || c == '?' || c == '|' || c == '&') {
1165 return sb.toString();
1169 * Command line tool to test QueryParser, using {@link org.apache.lucene.analysis.SimpleAnalyzer}.
1171 * <code>java org.apache.lucene.queryParser.QueryParser <input></code>
1173 public static void main(String[] args) throws Exception {
1174 if (args.length == 0) {
1175 System.out.println("Usage: java org.apache.lucene.queryParser.QueryParser <input>");
1178 QueryParser qp = new QueryParser(Version.LUCENE_CURRENT, "field",
1179 new org.apache.lucene.analysis.SimpleAnalyzer());
1180 Query q = qp.parse(args[0]);
1181 System.out.println(q.toString("field"));
1185 PARSER_END(QueryParser)
1187 /* ***************** */
1188 /* Token Definitions */
1189 /* ***************** */
1192 <#_NUM_CHAR: ["0"-"9"] >
1193 // every character that follows a backslash is considered as an escaped character
1194 | <#_ESCAPED_CHAR: "\\" ~[] >
1195 | <#_TERM_START_CHAR: ( ~[ " ", "\t", "\n", "\r", "\u3000", "+", "-", "!", "(", ")", ":", "^",
1196 "[", "]", "\"", "{", "}", "~", "*", "?", "\\" ]
1197 | <_ESCAPED_CHAR> ) >
1198 | <#_TERM_CHAR: ( <_TERM_START_CHAR> | <_ESCAPED_CHAR> | "-" | "+" ) >
1199 | <#_WHITESPACE: ( " " | "\t" | "\n" | "\r" | "\u3000") >
1200 | <#_QUOTED_CHAR: ( ~[ "\"", "\\" ] | <_ESCAPED_CHAR> ) >
1203 <DEFAULT, RangeIn, RangeEx> SKIP : {
1208 <AND: ("AND" | "&&") >
1209 | <OR: ("OR" | "||") >
1210 | <NOT: ("NOT" | "!") >
1217 | <CARAT: "^" > : Boost
1218 | <QUOTED: "\"" (<_QUOTED_CHAR>)* "\"">
1219 | <TERM: <_TERM_START_CHAR> (<_TERM_CHAR>)* >
1220 | <FUZZY_SLOP: "~" ( (<_NUM_CHAR>)+ ( "." (<_NUM_CHAR>)+ )? )? >
1221 | <PREFIXTERM: ("*") | ( <_TERM_START_CHAR> (<_TERM_CHAR>)* "*" ) >
1222 | <WILDTERM: (<_TERM_START_CHAR> | [ "*", "?" ]) (<_TERM_CHAR> | ( [ "*", "?" ] ))* >
1223 | <RANGEIN_START: "[" > : RangeIn
1224 | <RANGEEX_START: "{" > : RangeEx
1228 <NUMBER: (<_NUM_CHAR>)+ ( "." (<_NUM_CHAR>)+ )? > : DEFAULT
1233 | <RANGEIN_END: "]"> : DEFAULT
1234 | <RANGEIN_QUOTED: "\"" (~["\""] | "\\\"")+ "\"">
1235 | <RANGEIN_GOOP: (~[ " ", "]" ])+ >
1240 | <RANGEEX_END: "}"> : DEFAULT
1241 | <RANGEEX_QUOTED: "\"" (~["\""] | "\\\"")+ "\"">
1242 | <RANGEEX_GOOP: (~[ " ", "}" ])+ >
1245 // * Query ::= ( Clause )*
1246 // * Clause ::= ["+", "-"] [<TERM> ":"] ( <TERM> | "(" Query ")" )
1248 int Conjunction() : {
1249 int ret = CONJ_NONE;
1253 <AND> { ret = CONJ_AND; }
1254 | <OR> { ret = CONJ_OR; }
1264 <PLUS> { ret = MOD_REQ; }
1265 | <MINUS> { ret = MOD_NOT; }
1266 | <NOT> { ret = MOD_NOT; }
1271 // This makes sure that there is no garbage after the query string
1272 Query TopLevelQuery(String field) :
1277 q=Query(field) <EOF>
1283 Query Query(String field) :
1285 List<BooleanClause> clauses = new ArrayList<BooleanClause>();
1286 Query q, firstQuery=null;
1290 mods=Modifiers() q=Clause(field)
1292 addClause(clauses, CONJ_NONE, mods, q);
1293 if (mods == MOD_NONE)
1297 conj=Conjunction() mods=Modifiers() q=Clause(field)
1298 { addClause(clauses, conj, mods, q); }
1301 if (clauses.size() == 1 && firstQuery != null)
1304 return getBooleanQuery(clauses);
1309 Query Clause(String field) : {
1311 Token fieldToken=null, boost=null;
1317 fieldToken=<TERM> <COLON> {field=discardEscapeChar(fieldToken.image);}
1318 | <STAR> <COLON> {field="*";}
1324 | <LPAREN> q=Query(field) <RPAREN> (<CARAT> boost=<NUMBER>)?
1328 if (boost != null) {
1329 float f = (float)1.0;
1331 f = Float.valueOf(boost.image).floatValue();
1333 } catch (Exception ignored) { }
1340 Query Term(String field) : {
1341 Token term, boost=null, fuzzySlop=null, goop1, goop2;
1342 boolean prefix = false;
1343 boolean wildcard = false;
1344 boolean fuzzy = false;
1351 | term=<STAR> { wildcard=true; }
1352 | term=<PREFIXTERM> { prefix=true; }
1353 | term=<WILDTERM> { wildcard=true; }
1356 [ fuzzySlop=<FUZZY_SLOP> { fuzzy=true; } ]
1357 [ <CARAT> boost=<NUMBER> [ fuzzySlop=<FUZZY_SLOP> { fuzzy=true; } ] ]
1359 String termImage=discardEscapeChar(term.image);
1361 q = getWildcardQuery(field, termImage);
1362 } else if (prefix) {
1363 q = getPrefixQuery(field,
1364 discardEscapeChar(term.image.substring
1365 (0, term.image.length()-1)));
1367 float fms = fuzzyMinSim;
1369 fms = Float.valueOf(fuzzySlop.image.substring(1)).floatValue();
1370 } catch (Exception ignored) { }
1371 if(fms < 0.0f || fms > 1.0f){
1372 throw new ParseException("Minimum similarity for a FuzzyQuery has to be between 0.0f and 1.0f !");
1374 q = getFuzzyQuery(field, termImage,fms);
1376 q = hasNewAPI ? getFieldQuery(field, termImage, false) : getFieldQuery(field, termImage);
1379 | ( <RANGEIN_START> ( goop1=<RANGEIN_GOOP>|goop1=<RANGEIN_QUOTED> )
1380 [ <RANGEIN_TO> ] ( goop2=<RANGEIN_GOOP>|goop2=<RANGEIN_QUOTED> )
1382 [ <CARAT> boost=<NUMBER> ]
1384 if (goop1.kind == RANGEIN_QUOTED) {
1385 goop1.image = goop1.image.substring(1, goop1.image.length()-1);
1387 if (goop2.kind == RANGEIN_QUOTED) {
1388 goop2.image = goop2.image.substring(1, goop2.image.length()-1);
1390 q = getRangeQuery(field, discardEscapeChar(goop1.image), discardEscapeChar(goop2.image), true);
1392 | ( <RANGEEX_START> ( goop1=<RANGEEX_GOOP>|goop1=<RANGEEX_QUOTED> )
1393 [ <RANGEEX_TO> ] ( goop2=<RANGEEX_GOOP>|goop2=<RANGEEX_QUOTED> )
1395 [ <CARAT> boost=<NUMBER> ]
1397 if (goop1.kind == RANGEEX_QUOTED) {
1398 goop1.image = goop1.image.substring(1, goop1.image.length()-1);
1400 if (goop2.kind == RANGEEX_QUOTED) {
1401 goop2.image = goop2.image.substring(1, goop2.image.length()-1);
1404 q = getRangeQuery(field, discardEscapeChar(goop1.image), discardEscapeChar(goop2.image), false);
1407 [ fuzzySlop=<FUZZY_SLOP> ]
1408 [ <CARAT> boost=<NUMBER> ]
1412 if (fuzzySlop != null) {
1414 s = Float.valueOf(fuzzySlop.image.substring(1)).intValue();
1416 catch (Exception ignored) { }
1418 q = getFieldQuery(field, discardEscapeChar(term.image.substring(1, term.image.length()-1)), s);
1422 if (boost != null) {
1423 float f = (float) 1.0;
1425 f = Float.valueOf(boost.image).floatValue();
1427 catch (Exception ignored) {
1428 /* Should this be handled somehow? (defaults to "no boost", if
1429 * boost number is invalid)
1433 // avoid boosting null queries, such as those caused by stop words