Split out the lucene query parser from the core repository.

git-svn-id: https://svn.alfresco.com/repos/alfresco-enterprise/alfresco/HEAD/root@20978 c4b6b30b-aa2e-2d43-bbcb-ca4b014f7261
This commit is contained in:
Andrew Hind
2010-07-07 12:31:10 +00:00
parent 8105f39e33
commit 839f421e5f
28 changed files with 4 additions and 8534 deletions

View File

@@ -1,323 +0,0 @@
/*
* Copyright (C) 2005-2010 Alfresco Software Limited.
*
* This file is part of Alfresco
*
* Alfresco is free software: you can redistribute it and/or modify
* it under the terms of the GNU Lesser General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* Alfresco is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public License
* along with Alfresco. If not, see <http://www.gnu.org/licenses/>.
*/
package org.alfresco.util;
import org.alfresco.repo.search.impl.lucene.LuceneQueryParser;
import org.apache.lucene.queryParser.QueryParser;
/**
* Helper class to provide conversions between different search languages
*
* @author Derek Hulley
*/
public class SearchLanguageConversion
{
/**
* SQL like query language summary:
* <ul>
* <li>Escape: \</li>
* <li>Single char search: _</li>
* <li>Multiple char search: %</li>
* <li>Reserved: \%_</li>
* </ul>
*/
public static LanguageDefinition DEF_SQL_LIKE = new SimpleLanguageDef('\\', "%", "_", "\\%_[]");
/**
* XPath like query language summary:
* <ul>
* <li>Escape: \</li>
* <li>Single char search: _</li>
* <li>Multiple char search: %</li>
* <li>Reserved: \%_</li>
* </ul>
*/
public static LanguageDefinition DEF_XPATH_LIKE = new SimpleLanguageDef('\\', "%", "_", "\\%_[]");
/**
* Regular expression query language summary:
* <ul>
* <li>Escape: \</li>
* <li>Single char search: .</li>
* <li>Multiple char search: .*</li>
* <li>Reserved: \*.+?^$(){}|</li>
* </ul>
*/
public static LanguageDefinition DEF_REGEX = new SimpleLanguageDef('\\', ".*", ".", "\\*.+?^$(){}|");
/**
* Lucene syntax summary: {@link QueryParser#escape(String) Lucene Query Parser}
*/
public static LanguageDefinition DEF_LUCENE = new LuceneLanguageDef(true);
public static LanguageDefinition DEF_LUCENE_INTERNAL = new LuceneLanguageDef(false);
/**
* CIFS name patch query language summary:
* <ul>
* <li>Escape: \ (but not used)</li>
* <li>Single char search: ?</li>
* <li>Multiple char search: *</li>
* <li>Reserved: "*\<>?/:|£%&+;</li>
* </ul>
*/
public static LanguageDefinition DEF_CIFS = new SimpleLanguageDef('\\', "*", "?", "\"*\\<>?/:|£%&+;");
/**
* Escape a string according to the <b>XPath</b> like function syntax.
*
* @param str
* the string to escape
* @return Returns the escaped string
*/
public static String escapeForXPathLike(String str)
{
return escape(DEF_XPATH_LIKE, str);
}
/**
* Escape a string according to the <b>regex</b> language syntax.
*
* @param str
* the string to escape
* @return Returns the escaped string
*/
public static String escapeForRegex(String str)
{
return escape(DEF_REGEX, str);
}
/**
* Escape a string according to the <b>Lucene</b> query syntax.
*
* @param str
* the string to escape
* @return Returns the escaped string
*/
public static String escapeForLucene(String str)
{
return escape(DEF_LUCENE, str);
}
/**
* Generic escaping using the language definition
*/
private static String escape(LanguageDefinition def, String str)
{
StringBuilder sb = new StringBuilder(str.length() * 2);
char[] chars = str.toCharArray();
for (int i = 0; i < chars.length; i++)
{
// first check for reserved chars
if (def.isReserved(chars[i]))
{
// escape it
sb.append(def.escapeChar);
}
sb.append(chars[i]);
}
return sb.toString();
}
/**
* Convert an <b>xpath</b> like function clause into a <b>regex</b> query.
*
* @param xpathLikeClause
* @return Returns a valid regular expression that is equivalent to the given <b>xpath</b> like clause.
*/
public static String convertXPathLikeToRegex(String xpathLikeClause)
{
return "(?s)" + convert(DEF_XPATH_LIKE, DEF_REGEX, xpathLikeClause);
}
/**
* Convert an <b>xpath</b> like function clause into a <b>Lucene</b> query.
*
* @param xpathLikeClause
* @return Returns a valid <b>Lucene</b> expression that is equivalent to the given <b>xpath</b> like clause.
*/
public static String convertXPathLikeToLucene(String xpathLikeClause)
{
return convert(DEF_XPATH_LIKE, DEF_LUCENE, xpathLikeClause);
}
/**
* Convert a <b>sql</b> like function clause into a <b>Lucene</b> query.
*
* @param sqlLikeClause
* @return Returns a valid <b>Lucene</b> expression that is equivalent to the given <b>sql</b> like clause.
*/
public static String convertSQLLikeToLucene(String sqlLikeClause)
{
return convert(DEF_SQL_LIKE, DEF_LUCENE_INTERNAL, sqlLikeClause);
}
/**
* Convert a <b>sql</b> like function clause into a <b>regex</b> query.
*
* @param sqlLikeClause
* @return Returns a valid regular expression that is equivalent to the given <b>sql</b> like clause.
*/
public static String convertSQLLikeToRegex(String sqlLikeClause)
{
return "(?s)" + convert(DEF_SQL_LIKE, DEF_REGEX, sqlLikeClause);
}
/**
* Convert a <b>CIFS</b> name path into the equivalent <b>Lucene</b> query.
*
* @param cifsNamePath
* the CIFS named path
* @return Returns a valid <b>Lucene</b> expression that is equivalent to the given CIFS name path
*/
public static String convertCifsToLucene(String cifsNamePath)
{
return convert(DEF_CIFS, DEF_LUCENE, cifsNamePath);
}
public static String convert(LanguageDefinition from, LanguageDefinition to, String query)
{
char[] chars = query.toCharArray();
StringBuilder sb = new StringBuilder(chars.length * 2);
boolean escaping = false;
for (int i = 0; i < chars.length; i++)
{
if (escaping) // if we are currently escaping, just escape the current character
{
if(to.isReserved(chars[i]))
{
sb.append(to.escapeChar); // the to format escape char
}
sb.append(chars[i]); // the current char
escaping = false;
}
else if (chars[i] == from.escapeChar) // not escaping and have escape char
{
escaping = true;
}
else if (query.startsWith(from.multiCharWildcard, i)) // not escaping but have multi-char wildcard
{
// translate the wildcard
sb.append(to.multiCharWildcard);
}
else if (query.startsWith(from.singleCharWildcard, i)) // have single-char wildcard
{
// translate the wildcard
sb.append(to.singleCharWildcard);
}
else if (to.isReserved(chars[i])) // reserved character
{
sb.append(to.escapeChar).append(chars[i]);
}
else
// just a normal char in both
{
sb.append(chars[i]);
}
}
return sb.toString();
}
/**
* Simple store of special characters for a given query language
*/
public static abstract class LanguageDefinition
{
public final char escapeChar;
public final String multiCharWildcard;
public final String singleCharWildcard;
public LanguageDefinition(char escapeChar, String multiCharWildcard, String singleCharWildcard)
{
this.escapeChar = escapeChar;
this.multiCharWildcard = multiCharWildcard;
this.singleCharWildcard = singleCharWildcard;
}
public abstract boolean isReserved(char ch);
}
private static class SimpleLanguageDef extends LanguageDefinition
{
private String reserved;
public SimpleLanguageDef(char escapeChar, String multiCharWildcard, String singleCharWildcard, String reserved)
{
super(escapeChar, multiCharWildcard, singleCharWildcard);
this.reserved = reserved;
}
@Override
public boolean isReserved(char ch)
{
return (reserved.indexOf(ch) > -1);
}
}
private static class LuceneLanguageDef extends LanguageDefinition
{
private String reserved;
public LuceneLanguageDef(boolean reserve)
{
super('\\', "*", "?");
if (reserve)
{
init();
}
else
{
reserved = "";
}
}
/**
* Discovers all the reserved chars
*/
private void init()
{
StringBuilder sb = new StringBuilder(20);
for (char ch = 0; ch < 256; ch++)
{
char[] chars = new char[] { ch };
String unescaped = new String(chars);
// check it
String escaped = LuceneQueryParser.escape(unescaped);
if (!escaped.equals(unescaped))
{
// it was escaped
sb.append(ch);
}
}
reserved = sb.toString();
}
@Override
public boolean isReserved(char ch)
{
return (reserved.indexOf(ch) > -1);
}
}
}

View File

@@ -1,122 +0,0 @@
/*
* Copyright (C) 2005-2010 Alfresco Software Limited.
*
* This file is part of Alfresco
*
* Alfresco is free software: you can redistribute it and/or modify
* it under the terms of the GNU Lesser General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* Alfresco is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public License
* along with Alfresco. If not, see <http://www.gnu.org/licenses/>.
*/
package org.alfresco.util;
import junit.framework.TestCase;
/**
* @see org.alfresco.util.SearchLanguageConversion
*
* @author Derek Hulley
*/
public class SearchLanguageConversionTest extends TestCase
{
/**
* A string with a whole lod of badness to stress test with
*/
private static final String BAD_STRING =
"\\ | ! \" £ " +
"$ % ^ & * ( " +
") _ { } [ ] " +
"@ # ~ ' : ; " +
", . < > + ? " +
"/ \\\\ \\* \\? \\_";
public void testEscapeXPathLike()
{
String good = SearchLanguageConversion.escapeForXPathLike(BAD_STRING);
assertEquals("Escaping for xpath failed",
"\\\\ | ! \" £ " +
"$ \\% ^ & * ( " +
") \\_ { } \\[ \\] " +
"@ # ~ ' : ; " +
", . < > + ? " +
"/ \\\\\\\\ \\\\* \\\\? \\\\\\_",
good);
}
public void testEscapeRegex()
{
String good = SearchLanguageConversion.escapeForRegex(BAD_STRING);
assertEquals("Escaping for regex failed",
"\\\\ \\| ! \" £ " +
"\\$ % \\^ & \\* \\( " +
"\\) _ \\{ \\} [ ] " +
"@ # ~ ' : ; " +
", \\. < > \\+ \\? " +
"/ \\\\\\\\ \\\\\\* \\\\\\? \\\\_",
good);
}
public void testEscapeLucene()
{
String good = SearchLanguageConversion.escapeForLucene(BAD_STRING);
assertEquals("Escaping for Lucene failed",
"\\\\ \\| \\! \\\" £ " +
"$ % \\^ \\& \\* \\( " +
"\\) _ \\{ \\} \\[ \\] " +
"@ # \\~ ' \\: ; " +
", . < > \\+ \\? " +
"/ \\\\\\\\ \\\\\\* \\\\\\? \\\\_",
good);
}
public void testConvertXPathLikeToRegex()
{
String good = SearchLanguageConversion.convertXPathLikeToRegex(BAD_STRING);
assertEquals("XPath like to regex failed",
"(?s) \\| ! \" £ " +
"\\$ .* \\^ & \\* \\( " +
"\\) . \\{ \\} [ ] " +
"@ # ~ ' : ; " +
", \\. < > \\+ \\? " +
"/ \\\\ \\* \\? _",
good);
}
public void testConvertXPathLikeToLucene()
{
String good = SearchLanguageConversion.convertXPathLikeToLucene(BAD_STRING);
assertEquals("XPath like to Lucene failed",
" \\| \\! \\\" £ " +
"$ * \\^ \\& \\* \\( " +
"\\) ? \\{ \\} \\[ \\] " +
"@ # \\~ ' \\: ; " +
", . < > \\+ \\? " +
"/ \\\\ \\* \\? _",
good);
}
public void testSqlToLucene()
{
String sqlLike = "AB%_*?\\%\\_";
String lucene = "AB*?\\*\\?%_";
String converted = SearchLanguageConversion.convert(SearchLanguageConversion.DEF_SQL_LIKE, SearchLanguageConversion.DEF_LUCENE, sqlLike);
assertEquals(lucene, converted);
}
public void testLuceneToRegexp()
{
String lucene = "AB*?\\*\\?.*.";
String regexp = "AB.*.\\*\\?\\..*\\.";
String converted = SearchLanguageConversion.convert(SearchLanguageConversion.DEF_LUCENE, SearchLanguageConversion.DEF_REGEX, lucene);
assertEquals(regexp, converted);
}
}