You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@lucene.apache.org by GitBox <gi...@apache.org> on 2022/09/07 02:27:37 UTC
[GitHub] [lucene] jtibshirani opened a new pull request, #11756: LUCENE-10577: Remove LeafReader#searchNearestVectorsExhaustively
jtibshirani opened a new pull request, #11756:
URL: https://github.com/apache/lucene/pull/11756
This PR removes the recently added function on LeafReader to exhaustively
search through vectors, plus the helper function
KnnVectorsReader#searchExhaustively. Instead it performs the exact search
within KnnVectorQuery, using a new helper class called VectorScorer.
Follow-up to #1054
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] jtibshirani commented on a diff in pull request #11756: LUCENE-10577: Remove LeafReader#searchNearestVectorsExhaustively
Posted by GitBox <gi...@apache.org>.
jtibshirani commented on code in PR #11756:
URL: https://github.com/apache/lucene/pull/11756#discussion_r965154356
##########
lucene/core/src/java/org/apache/lucene/search/KnnVectorQuery.java:
##########
@@ -175,9 +176,42 @@ private TopDocs approximateSearch(LeafReaderContext context, Bits acceptDocs, in
}
// We allow this to be overridden so that tests can check what search strategy is used
- protected TopDocs exactSearch(LeafReaderContext context, DocIdSetIterator acceptDocs)
+ protected TopDocs exactSearch(LeafReaderContext context, DocIdSetIterator acceptIterator)
throws IOException {
- return context.reader().searchNearestVectorsExhaustively(field, target, k, acceptDocs);
+ FieldInfo fi = context.reader().getFieldInfos().fieldInfo(field);
+ if (fi == null || fi.getVectorDimension() == 0) {
+ // The field does not exist or does not index vectors
+ return NO_RESULTS;
+ }
+
+ VectorScorer vectorScorer = VectorScorer.create(context, fi, target);
+ HitQueue queue = new HitQueue(k, true);
+ ScoreDoc topDoc = queue.top();
+ int doc;
+ while ((doc = acceptIterator.nextDoc()) != DocIdSetIterator.NO_MORE_DOCS) {
+ boolean advanced = vectorScorer.advanceExact(doc);
+ assert advanced;
+
+ float score = vectorScorer.score();
+ if (score >= topDoc.score) {
Review Comment:
That seems right, I updated this and pushed a test covering the tie-breaking case. As a note, we don't guarantee we'll always return the lowest matching doc IDs (since approximate HNSW search can't do this efficiently).
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] jtibshirani merged pull request #11756: LUCENE-10577: Remove LeafReader#searchNearestVectorsExhaustively
Posted by GitBox <gi...@apache.org>.
jtibshirani merged PR #11756:
URL: https://github.com/apache/lucene/pull/11756
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] jtibshirani commented on pull request #11756: LUCENE-10577: Remove LeafReader#searchNearestVectorsExhaustively
Posted by GitBox <gi...@apache.org>.
jtibshirani commented on PR #11756:
URL: https://github.com/apache/lucene/pull/11756#issuecomment-1241095296
Thanks for the reviews, I'll merge and backport to 9.4.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] jtibshirani commented on a diff in pull request #11756: LUCENE-10577: Remove LeafReader#searchNearestVectorsExhaustively
Posted by GitBox <gi...@apache.org>.
jtibshirani commented on code in PR #11756:
URL: https://github.com/apache/lucene/pull/11756#discussion_r965171139
##########
lucene/core/src/java/org/apache/lucene/search/VectorScorer.java:
##########
@@ -0,0 +1,107 @@
+/*
+ * 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.
+ */
+package org.apache.lucene.search;
+
+import java.io.IOException;
+import org.apache.lucene.index.FieldInfo;
+import org.apache.lucene.index.LeafReaderContext;
+import org.apache.lucene.index.VectorSimilarityFunction;
+import org.apache.lucene.index.VectorValues;
+import org.apache.lucene.util.BytesRef;
+import org.apache.lucene.util.VectorUtil;
+
+/**
+ * Computes the similarity score between a given query vector and different document vectors. This
+ * is primarily used by {@link org.apache.lucene.search.KnnVectorQuery} to run an exact, exhaustive
+ * search over the vectors.
+ */
+interface VectorScorer {
+
+ /**
+ * Create a new vector scorer instance.
+ *
+ * @param context the reader context
+ * @param fieldInfo the field to computeinformation on the
+ * @param query the query vector to compute the similarity for
+ */
+ static VectorScorer create(LeafReaderContext context, FieldInfo fieldInfo, float[] query)
+ throws IOException {
+ VectorValues values = context.reader().getVectorValues(fieldInfo.name);
+ VectorSimilarityFunction similarity = fieldInfo.getVectorSimilarityFunction();
+ return switch (fieldInfo.getVectorEncoding()) {
+ case BYTE -> new ByteVectorScorer(values, query, similarity);
+ case FLOAT32 -> new FloatVectorScorer(values, query, similarity);
+ };
+ }
+
+ /**
+ * Advance the instance to the given document ID and return true if there is a value for that
+ * document.
+ */
+ boolean advanceExact(int doc) throws IOException;
+
+ /** Compute the similarity score for the current document. */
+ float score() throws IOException;
+
+ class ByteVectorScorer implements VectorScorer {
+ private final VectorValues values;
+ private final BytesRef query;
+ private final VectorSimilarityFunction similarity;
+
+ protected ByteVectorScorer(
+ VectorValues values, float[] query, VectorSimilarityFunction similarity) {
+ this.values = values;
+ this.similarity = similarity;
+ this.query = VectorUtil.toBytesRef(query);
+ }
+
+ @Override
+ public boolean advanceExact(int doc) throws IOException {
+ int vectorDoc = values.advance(doc);
Review Comment:
Right! I overlooked this because VectorScorer is only used in KnnVectorQuery, which asserts `advanceExact` always returns `true`. But this is a more general abstraction so we should fix it.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] msokolov commented on a diff in pull request #11756: LUCENE-10577: Remove LeafReader#searchNearestVectorsExhaustively
Posted by GitBox <gi...@apache.org>.
msokolov commented on code in PR #11756:
URL: https://github.com/apache/lucene/pull/11756#discussion_r965934198
##########
lucene/core/src/java/org/apache/lucene/search/KnnVectorQuery.java:
##########
@@ -175,9 +176,42 @@ private TopDocs approximateSearch(LeafReaderContext context, Bits acceptDocs, in
}
// We allow this to be overridden so that tests can check what search strategy is used
- protected TopDocs exactSearch(LeafReaderContext context, DocIdSetIterator acceptDocs)
+ protected TopDocs exactSearch(LeafReaderContext context, DocIdSetIterator acceptIterator)
throws IOException {
- return context.reader().searchNearestVectorsExhaustively(field, target, k, acceptDocs);
+ FieldInfo fi = context.reader().getFieldInfos().fieldInfo(field);
+ if (fi == null || fi.getVectorDimension() == 0) {
+ // The field does not exist or does not index vectors
+ return NO_RESULTS;
+ }
+
+ VectorScorer vectorScorer = VectorScorer.create(context, fi, target);
+ HitQueue queue = new HitQueue(k, true);
+ ScoreDoc topDoc = queue.top();
+ int doc;
+ while ((doc = acceptIterator.nextDoc()) != DocIdSetIterator.NO_MORE_DOCS) {
+ boolean advanced = vectorScorer.advanceExact(doc);
+ assert advanced;
+
+ float score = vectorScorer.score();
+ if (score >= topDoc.score) {
Review Comment:
Yes I think this was done with the docid-tie-breaking in mind, but that is not guaranteed, so we might as well relax this too.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] jtibshirani commented on pull request #11756: LUCENE-10577: Remove LeafReader#searchNearestVectorsExhaustively
Posted by GitBox <gi...@apache.org>.
jtibshirani commented on PR #11756:
URL: https://github.com/apache/lucene/pull/11756#issuecomment-1238843697
This cuts down on the API surface area for LeafReader results in less logic. I think what we tried to do initially was to push the exhaustive search down to the codec level, so that KnnVectorQuery doesn't need to know about the encoding. However, the VectorEncoding is already public, so it doesn't seem too bad for KnnVectorQuery to consider it (especially since it's abstracted away into VectorScorer).
If we want to limit knowledge of the encoding outside each codec, I think we should focus on removing VectorValues#binaryValue altogether, which would require a different refactoring approach.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] jpountz commented on a diff in pull request #11756: LUCENE-10577: Remove LeafReader#searchNearestVectorsExhaustively
Posted by GitBox <gi...@apache.org>.
jpountz commented on code in PR #11756:
URL: https://github.com/apache/lucene/pull/11756#discussion_r965022161
##########
lucene/core/src/java/org/apache/lucene/search/KnnVectorQuery.java:
##########
@@ -175,9 +176,42 @@ private TopDocs approximateSearch(LeafReaderContext context, Bits acceptDocs, in
}
// We allow this to be overridden so that tests can check what search strategy is used
- protected TopDocs exactSearch(LeafReaderContext context, DocIdSetIterator acceptDocs)
+ protected TopDocs exactSearch(LeafReaderContext context, DocIdSetIterator acceptIterator)
throws IOException {
- return context.reader().searchNearestVectorsExhaustively(field, target, k, acceptDocs);
+ FieldInfo fi = context.reader().getFieldInfos().fieldInfo(field);
+ if (fi == null || fi.getVectorDimension() == 0) {
+ // The field does not exist or does not index vectors
+ return NO_RESULTS;
+ }
+
+ VectorScorer vectorScorer = VectorScorer.create(context, fi, target);
+ HitQueue queue = new HitQueue(k, true);
+ ScoreDoc topDoc = queue.top();
+ int doc;
+ while ((doc = acceptIterator.nextDoc()) != DocIdSetIterator.NO_MORE_DOCS) {
+ boolean advanced = vectorScorer.advanceExact(doc);
+ assert advanced;
+
+ float score = vectorScorer.score();
+ if (score >= topDoc.score) {
Review Comment:
This is pre-existing code but I would assume that we only need to update the priority queue when we find a score that is strictly greater than the current top entry of the priority queue.
##########
lucene/core/src/java/org/apache/lucene/search/VectorScorer.java:
##########
@@ -0,0 +1,107 @@
+/*
+ * 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.
+ */
+package org.apache.lucene.search;
+
+import java.io.IOException;
+import org.apache.lucene.index.FieldInfo;
+import org.apache.lucene.index.LeafReaderContext;
+import org.apache.lucene.index.VectorSimilarityFunction;
+import org.apache.lucene.index.VectorValues;
+import org.apache.lucene.util.BytesRef;
+import org.apache.lucene.util.VectorUtil;
+
+/**
+ * Computes the similarity score between a given query vector and different document vectors. This
+ * is primarily used by {@link org.apache.lucene.search.KnnVectorQuery} to run an exact, exhaustive
+ * search over the vectors.
+ */
+interface VectorScorer {
+
+ /**
+ * Create a new vector scorer instance.
+ *
+ * @param context the reader context
+ * @param fieldInfo the field to computeinformation on the
+ * @param query the query vector to compute the similarity for
+ */
+ static VectorScorer create(LeafReaderContext context, FieldInfo fieldInfo, float[] query)
+ throws IOException {
+ VectorValues values = context.reader().getVectorValues(fieldInfo.name);
+ VectorSimilarityFunction similarity = fieldInfo.getVectorSimilarityFunction();
+ return switch (fieldInfo.getVectorEncoding()) {
+ case BYTE -> new ByteVectorScorer(values, query, similarity);
+ case FLOAT32 -> new FloatVectorScorer(values, query, similarity);
+ };
+ }
+
+ /**
+ * Advance the instance to the given document ID and return true if there is a value for that
+ * document.
+ */
+ boolean advanceExact(int doc) throws IOException;
+
+ /** Compute the similarity score for the current document. */
+ float score() throws IOException;
+
+ class ByteVectorScorer implements VectorScorer {
+ private final VectorValues values;
+ private final BytesRef query;
+ private final VectorSimilarityFunction similarity;
+
+ protected ByteVectorScorer(
+ VectorValues values, float[] query, VectorSimilarityFunction similarity) {
+ this.values = values;
+ this.similarity = similarity;
+ this.query = VectorUtil.toBytesRef(query);
+ }
+
+ @Override
+ public boolean advanceExact(int doc) throws IOException {
+ int vectorDoc = values.advance(doc);
Review Comment:
Given that `advance` requires that the target doc ID is strictly beyond the current doc ID, I think we need to check the current doc ID and only advance if we're not on or beyond `doc` already.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] jtibshirani commented on a diff in pull request #11756: LUCENE-10577: Remove LeafReader#searchNearestVectorsExhaustively
Posted by GitBox <gi...@apache.org>.
jtibshirani commented on code in PR #11756:
URL: https://github.com/apache/lucene/pull/11756#discussion_r966310331
##########
lucene/core/src/java/org/apache/lucene/search/VectorScorer.java:
##########
@@ -0,0 +1,104 @@
+/*
+ * 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.
+ */
+package org.apache.lucene.search;
+
+import java.io.IOException;
+import org.apache.lucene.index.FieldInfo;
+import org.apache.lucene.index.LeafReaderContext;
+import org.apache.lucene.index.VectorSimilarityFunction;
+import org.apache.lucene.index.VectorValues;
+import org.apache.lucene.util.BytesRef;
+import org.apache.lucene.util.VectorUtil;
+
+/**
+ * Computes the similarity score between a given query vector and different document vectors. This
+ * is primarily used by {@link org.apache.lucene.search.KnnVectorQuery} to run an exact, exhaustive
+ * search over the vectors.
+ */
+abstract class VectorScorer {
+ protected final VectorValues values;
+
+ /**
+ * Create a new vector scorer instance.
+ *
+ * @param context the reader context
+ * @param fi the FieldInfo for the field containing document vectors
+ * @param query the query vector to compute the similarity for
+ */
+ static VectorScorer create(LeafReaderContext context, FieldInfo fi, float[] query)
+ throws IOException {
+ VectorValues values = context.reader().getVectorValues(fi.name);
+ VectorSimilarityFunction similarity = fi.getVectorSimilarityFunction();
+ return switch (fi.getVectorEncoding()) {
+ case BYTE -> new ByteVectorScorer(values, query, similarity);
+ case FLOAT32 -> new FloatVectorScorer(values, query, similarity);
+ };
+ }
+
+ VectorScorer(VectorValues values) {
+ this.values = values;
+ }
+
+ /**
+ * Advance the instance to the given document ID and return true if there is a value for that
+ * document.
+ */
+ public boolean advanceExact(int doc) throws IOException {
+ int vectorDoc = values.docID();
+ if (vectorDoc < doc) {
+ vectorDoc = values.advance(doc);
+ }
+ return vectorDoc == doc;
+ }
+
+ /** Compute the similarity score for the current document. */
+ abstract float score() throws IOException;
+
+ private static class ByteVectorScorer extends VectorScorer {
+ private final BytesRef query;
+ private final VectorSimilarityFunction similarity;
Review Comment:
👍
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org
[GitHub] [lucene] msokolov commented on a diff in pull request #11756: LUCENE-10577: Remove LeafReader#searchNearestVectorsExhaustively
Posted by GitBox <gi...@apache.org>.
msokolov commented on code in PR #11756:
URL: https://github.com/apache/lucene/pull/11756#discussion_r965936802
##########
lucene/core/src/java/org/apache/lucene/search/VectorScorer.java:
##########
@@ -0,0 +1,104 @@
+/*
+ * 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.
+ */
+package org.apache.lucene.search;
+
+import java.io.IOException;
+import org.apache.lucene.index.FieldInfo;
+import org.apache.lucene.index.LeafReaderContext;
+import org.apache.lucene.index.VectorSimilarityFunction;
+import org.apache.lucene.index.VectorValues;
+import org.apache.lucene.util.BytesRef;
+import org.apache.lucene.util.VectorUtil;
+
+/**
+ * Computes the similarity score between a given query vector and different document vectors. This
+ * is primarily used by {@link org.apache.lucene.search.KnnVectorQuery} to run an exact, exhaustive
+ * search over the vectors.
+ */
+abstract class VectorScorer {
+ protected final VectorValues values;
+
+ /**
+ * Create a new vector scorer instance.
+ *
+ * @param context the reader context
+ * @param fi the FieldInfo for the field containing document vectors
+ * @param query the query vector to compute the similarity for
+ */
+ static VectorScorer create(LeafReaderContext context, FieldInfo fi, float[] query)
+ throws IOException {
+ VectorValues values = context.reader().getVectorValues(fi.name);
+ VectorSimilarityFunction similarity = fi.getVectorSimilarityFunction();
+ return switch (fi.getVectorEncoding()) {
+ case BYTE -> new ByteVectorScorer(values, query, similarity);
+ case FLOAT32 -> new FloatVectorScorer(values, query, similarity);
+ };
+ }
+
+ VectorScorer(VectorValues values) {
+ this.values = values;
+ }
+
+ /**
+ * Advance the instance to the given document ID and return true if there is a value for that
+ * document.
+ */
+ public boolean advanceExact(int doc) throws IOException {
+ int vectorDoc = values.docID();
+ if (vectorDoc < doc) {
+ vectorDoc = values.advance(doc);
+ }
+ return vectorDoc == doc;
+ }
+
+ /** Compute the similarity score for the current document. */
+ abstract float score() throws IOException;
+
+ private static class ByteVectorScorer extends VectorScorer {
+ private final BytesRef query;
+ private final VectorSimilarityFunction similarity;
Review Comment:
nit: we could move the similarity to the base class
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org