--- /dev/null
+package org.apache.lucene.facet.taxonomy.directory;
+
+import java.io.IOException;
+
+import org.apache.lucene.index.CorruptIndexException;
+import org.apache.lucene.index.IndexReader;
+import org.apache.lucene.index.Term;
+import org.apache.lucene.index.TermPositions;
+
+import org.apache.lucene.facet.taxonomy.TaxonomyReader;
+
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements. See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+// getParent() needs to be extremely efficient, to the point that we need
+// to fetch all the data in advance into memory, and answer these calls
+// from memory. Currently we use a large integer array, which is
+// initialized when the taxonomy is opened, and potentially enlarged
+// when it is refresh()ed.
+/**
+ * @lucene.experimental
+ */
+class ParentArray {
+
+ // These arrays are not syncrhonized. Rather, the reference to the array
+ // is volatile, and the only writing operation (refreshPrefetchArrays)
+ // simply creates a new array and replaces the reference. The volatility
+ // of the reference ensures the correct atomic replacement and its
+ // visibility properties (the content of the array is visible when the
+ // new reference is visible).
+ private volatile int prefetchParentOrdinal[] = null;
+
+ public int[] getArray() {
+ return prefetchParentOrdinal;
+ }
+
+ /**
+ * refreshPrefetch() refreshes the parent array. Initially, it fills the
+ * array from the positions of an appropriate posting list. If called during
+ * a refresh(), when the arrays already exist, only values for new documents
+ * (those beyond the last one in the array) are read from the positions and
+ * added to the arrays (that are appropriately enlarged). We assume (and
+ * this is indeed a correct assumption in our case) that existing categories
+ * are never modified or deleted.
+ */
+ void refresh(IndexReader indexReader) throws IOException {
+ // Note that it is not necessary for us to obtain the read lock.
+ // The reason is that we are only called from refresh() (precluding
+ // another concurrent writer) or from the constructor (when no method
+ // could be running).
+ // The write lock is also not held during the following code, meaning
+ // that reads *can* happen while this code is running. The "volatile"
+ // property of the prefetchParentOrdinal and prefetchDepth array
+ // references ensure the correct visibility property of the assignment
+ // but other than that, we do *not* guarantee that a reader will not
+ // use an old version of one of these arrays (or both) while a refresh
+ // is going on. But we find this acceptable - until a refresh has
+ // finished, the reader should not expect to see new information
+ // (and the old information is the same in the old and new versions).
+ int first;
+ int num = indexReader.maxDoc();
+ if (prefetchParentOrdinal==null) {
+ prefetchParentOrdinal = new int[num];
+ // Starting Lucene 2.9, following the change LUCENE-1542, we can
+ // no longer reliably read the parent "-1" (see comment in
+ // LuceneTaxonomyWriter.SinglePositionTokenStream). We have no way
+ // to fix this in indexing without breaking backward-compatibility
+ // with existing indexes, so what we'll do instead is just
+ // hard-code the parent of ordinal 0 to be -1, and assume (as is
+ // indeed the case) that no other parent can be -1.
+ if (num>0) {
+ prefetchParentOrdinal[0] = TaxonomyReader.INVALID_ORDINAL;
+ }
+ first = 1;
+ } else {
+ first = prefetchParentOrdinal.length;
+ if (first==num) {
+ return; // nothing to do - no category was added
+ }
+ // In Java 6, we could just do Arrays.copyOf()...
+ int[] newarray = new int[num];
+ System.arraycopy(prefetchParentOrdinal, 0, newarray, 0,
+ prefetchParentOrdinal.length);
+ prefetchParentOrdinal = newarray;
+ }
+
+ // Read the new part of the parents array from the positions:
+ TermPositions positions = indexReader.termPositions(
+ new Term(Consts.FIELD_PAYLOADS, Consts.PAYLOAD_PARENT));
+ try {
+ if (!positions.skipTo(first) && first < num) {
+ throw new CorruptIndexException("Missing parent data for category " + first);
+ }
+ for (int i=first; i<num; i++) {
+ // Note that we know positions.doc() >= i (this is an
+ // invariant kept throughout this loop)
+ if (positions.doc()==i) {
+ if (positions.freq() == 0) { // shouldn't happen
+ throw new CorruptIndexException(
+ "Missing parent data for category "+i);
+ }
+
+ // TODO (Facet): keep a local (non-volatile) copy of the prefetchParentOrdinal
+ // reference, because access to volatile reference is slower (?).
+ // Note: The positions we get here are one less than the position
+ // increment we added originally, so we get here the right numbers:
+ prefetchParentOrdinal[i] = positions.nextPosition();
+
+ if (!positions.next()) {
+ if ( i+1 < num ) {
+ throw new CorruptIndexException(
+ "Missing parent data for category "+(i+1));
+ }
+ break;
+ }
+ } else { // this shouldn't happen
+ throw new CorruptIndexException(
+ "Missing parent data for category "+i);
+ }
+ }
+ } finally {
+ positions.close(); // to be on the safe side.
+ }
+ }
+
+ /**
+ * add() is used in LuceneTaxonomyWriter, not in LuceneTaxonomyReader.
+ * It is only called from a synchronized method, so it is not reentrant,
+ * and also doesn't need to worry about reads happening at the same time.
+ *
+ * NOTE: add() and refresh() CANNOT be used together. If you call add(),
+ * this changes the arrays and refresh() can no longer be used.
+ */
+ void add(int ordinal, int parentOrdinal) throws IOException {
+ if (ordinal >= prefetchParentOrdinal.length) {
+ // grow the array, if necessary.
+ // In Java 6, we could just do Arrays.copyOf()...
+ int[] newarray = new int[ordinal*2+1];
+ System.arraycopy(prefetchParentOrdinal, 0, newarray, 0,
+ prefetchParentOrdinal.length);
+ prefetchParentOrdinal = newarray;
+ }
+ prefetchParentOrdinal[ordinal] = parentOrdinal;
+ }
+
+}
fnp / pylucene.git / blobdiff