Author: natalia
Date: Tue Aug 21 18:58:31 2007
New Revision: 568372
URL: http://svn.apache.org/viewvc?rev=568372&view=rev
Log:
Configurable default field for text indexer
Modified:
xml/xindice/trunk/java/src/org/apache/xindice/core/indexer/LuceneIndexer.java
xml/xindice/trunk/java/tests/src/org/apache/xindice/core/indexer/LuceneIndexerTest.java
Modified:
xml/xindice/trunk/java/src/org/apache/xindice/core/indexer/LuceneIndexer.java
URL:
http://svn.apache.org/viewvc/xml/xindice/trunk/java/src/org/apache/xindice/core/indexer/LuceneIndexer.java?rev=568372&r1=568371&r2=568372&view=diff
==============================================================================
---
xml/xindice/trunk/java/src/org/apache/xindice/core/indexer/LuceneIndexer.java
(original)
+++
xml/xindice/trunk/java/src/org/apache/xindice/core/indexer/LuceneIndexer.java
Tue Aug 21 18:58:31 2007
@@ -85,6 +85,7 @@
private static final String NAME = "name";
private static final String PATTERN = "pattern";
+ private static final String DEFAULT = "default";
private static final String ANALYZER = "analyzer";
private static final String PATTERN_STRING = "pattern";
private static final String PATTERN_ALIAS = "alias";
@@ -112,6 +113,8 @@
private int docsAdded;
private int docsDeleted;
+ private String defaultField = "";
+
private void setFile(File f) {
idxFile = f;
}
@@ -150,11 +153,18 @@
* org.apache.lucene.analysis.SimpleAnalyzer by default.</ul>
*
* <dl><dt>pattern
- * <dd>Child element. Its attributes:
- * <ul><li>pattern - IndexPattern. Required. For acceptable formats, see
+ * <dd>Child element. Indexer must have at least one pattern. Its
+ * attributes:
+ * <ul><li>pattern - IndexPattern. For acceptable formats, see
* [EMAIL PROTECTED] org.apache.xindice.core.indexer.Indexer#getPatterns()}
* <li>alias - Name of the field to store/search values for that pattern.
- * Required.</ul></dl>
+ * </dl>
+ * <dl><dt>default
+ * <dd>Child element. Optional. Its attributes:
+ * <li>alias - Indicates the pattern alias that will be used as
+ * the default field for search. If omitted, search query has to include
+ * field name for all terms, there will be no default.
+ * </ul></dl>
* </dl>
*
* @param config Configuration to apply
@@ -182,6 +192,18 @@
this.patterns.put(new IndexPattern(collection.getSymbols(),
name, null), alias);
}
+ Configuration[] defaults = config.getChildren(DEFAULT);
+ if (defaults.length > 1) {
+ throw new CannotCreateException("There may be only one default
field");
+ } else if (defaults.length == 1) {
+ String alias = defaults[0].getAttribute(PATTERN_ALIAS);
+ if (this.patterns.values().contains(alias)) {
+ defaultField = alias;
+ } else {
+ throw new CannotCreateException("Alias '" + alias + "' is
undefined in configuration");
+ }
+ }
+
setFile(new File(collection.getCollectionRoot(), name));
} catch (Exception e) {
throw new XindiceException(e);
@@ -461,7 +483,7 @@
String textQuery = query.getValue(0).toString();
try {
- return queryMatches(new QueryParser("",
getAnalyzer()).parse(textQuery));
+ return queryMatches(new QueryParser(defaultField,
getAnalyzer()).parse(textQuery));
} catch (ParseException e) {
throw new CompilationException("Failed to parse query '" +
textQuery + "'", e);
}
Modified:
xml/xindice/trunk/java/tests/src/org/apache/xindice/core/indexer/LuceneIndexerTest.java
URL:
http://svn.apache.org/viewvc/xml/xindice/trunk/java/tests/src/org/apache/xindice/core/indexer/LuceneIndexerTest.java?rev=568372&r1=568371&r2=568372&view=diff
==============================================================================
---
xml/xindice/trunk/java/tests/src/org/apache/xindice/core/indexer/LuceneIndexerTest.java
(original)
+++
xml/xindice/trunk/java/tests/src/org/apache/xindice/core/indexer/LuceneIndexerTest.java
Tue Aug 21 18:58:31 2007
@@ -79,7 +79,7 @@
db.close();
}
- private LuceneIndexer createIndex(String name, HashMap patterns) throws
Exception {
+ private LuceneIndexer createIndex(String name, HashMap patterns, String
defaultField) throws Exception {
String config = "<index name='" + name + "' " +
"class='" + indexClass + "' " +
"analyzer='org.apache.lucene.analysis.SimpleAnalyzer'>";
@@ -87,6 +87,9 @@
String alias = (String) i.next();
config += "<pattern pattern='" + patterns.get(alias) + "' alias='"
+ alias + "' />";
}
+ if (defaultField != null) {
+ config += "<default alias='" + defaultField + "' />";
+ }
config += "</index>";
LuceneIndexer ind = (LuceneIndexer) collection.createIndexer(new
Configuration(DOMParser.toDocument(config)));
@@ -101,7 +104,7 @@
public void testSingleIndex() throws Exception {
HashMap patterns = new HashMap();
patterns.put("test", "a");
- LuceneIndexer idx = createIndex("test", patterns);
+ LuceneIndexer idx = createIndex("test", patterns, null);
String query = "test:test";
IndexMatch[] match = query(idx, query);
@@ -114,9 +117,9 @@
HashMap patterns = new HashMap();
patterns.put("test1", "[EMAIL PROTECTED]");
patterns.put("test2", "[EMAIL PROTECTED]");
- createIndex("test1", patterns);
- createIndex("test2", patterns);
- assertTrue("There can be only one full text index per collection",
false);
+ createIndex("test1", patterns, null);
+ createIndex("test2", patterns, null);
+ fail("There can be only one full text index per collection");
} catch (DuplicateIndexException e) {
// correct behavior
}
@@ -126,27 +129,27 @@
HashMap patterns = new HashMap();
patterns.put("test1", "[EMAIL PROTECTED]");
patterns.put("test2", "[EMAIL PROTECTED]");
- LuceneIndexer idx = createIndex("test", patterns);
+ LuceneIndexer idx = createIndex("test", patterns, "test1");
- IndexMatch[] match = query(idx, "test1:test");
+ IndexMatch[] match = query(idx, "test");
assertEquals(3, match.length);
match = query(idx, "test2:test");
assertEquals(3, match.length);
- match = query(idx, "test1:test AND test2:test");
+ match = query(idx, "test AND test2:test");
assertEquals(1, match.length);
- match = query(idx, "test1:test OR test2:test");
+ match = query(idx, "test OR test2:test");
assertEquals(5, match.length);
}
public void testNestedContent() throws Exception {
HashMap patterns = new HashMap();
patterns.put("doc", "doc");
- LuceneIndexer idx = createIndex("test", patterns);
+ LuceneIndexer idx = createIndex("test", patterns, "doc");
- String query = "doc:foo";
+ String query = "foo";
IndexMatch[] match = query(idx, query);
assertEquals(2, match.length);
