rohityadav1993 commented on code in PR #16727: URL: https://github.com/apache/pinot/pull/16727#discussion_r2430147579
########## pinot-segment-local/src/main/java/org/apache/pinot/segment/local/segment/creator/impl/stats/ColumnarSegmentPreIndexStatsContainer.java: ########## @@ -0,0 +1,217 @@ +/** + * 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.pinot.segment.local.segment.creator.impl.stats; + +import java.io.IOException; +import java.util.HashMap; +import java.util.Map; +import org.apache.pinot.segment.spi.creator.ColumnStatistics; +import org.apache.pinot.segment.spi.creator.SegmentPreIndexStatsCollector; +import org.apache.pinot.segment.spi.creator.StatsCollectorConfig; +import org.apache.pinot.spi.data.FieldSpec; +import org.apache.pinot.spi.data.Schema; +import org.apache.pinot.spi.data.readers.ColumnReader; +import org.apache.pinot.spi.data.readers.GenericRow; +import org.slf4j.Logger; +import org.slf4j.LoggerFactory; + + +/** + * Stats container that efficiently collects statistics from columnar data using ColumnReader instances. + * + * <p>This implementation collects statistics by iterating column-wise instead of row-wise, + * which is more efficient for columnar data sources. It supports: + * <ul> + * <li>Column-wise statistics collection</li> + * <li>Existing columns from source data</li> + * <li>New columns with default values</li> + * <li>Data type conversions during schema evolution</li> + * </ul> + * + * <p>The statistics are collected using the same underlying collectors as the row-based approach + * (SegmentPreIndexStatsCollectorImpl) but with more efficient column-wise iteration. + */ +public class ColumnarSegmentPreIndexStatsContainer implements SegmentPreIndexStatsCollector { + private static final Logger LOGGER = LoggerFactory.getLogger(ColumnarSegmentPreIndexStatsContainer.class); + + private final Map<String, ColumnReader> _columnReaders; + private final StatsCollectorConfig _statsCollectorConfig; + private final Schema _targetSchema; + private final Map<String, AbstractColumnStatisticsCollector> _columnStatsCollectorMap; + private int _totalDocCount; + + /** + * Create a ColumnarSegmentPreIndexStatsContainer. + * + * @param columnReaders Map of column name to ColumnReader instances + * @param statsCollectorConfig Configuration for statistics collection + */ + public ColumnarSegmentPreIndexStatsContainer(Map<String, ColumnReader> columnReaders, + StatsCollectorConfig statsCollectorConfig) { + _columnReaders = columnReaders; + _statsCollectorConfig = statsCollectorConfig; + _targetSchema = statsCollectorConfig.getSchema(); + _columnStatsCollectorMap = new HashMap<>(); + _totalDocCount = -1; // indicates unset + + initializeStatsCollectors(); + collectColumnStats(); + } + + /** + * Initialize stats collectors for all columns in the target schema. + */ + private void initializeStatsCollectors() { + for (FieldSpec fieldSpec : _targetSchema.getAllFieldSpecs()) { + if (fieldSpec.isVirtualColumn()) { + continue; + } + + String columnName = fieldSpec.getName(); + switch (fieldSpec.getDataType().getStoredType()) { + case INT: + _columnStatsCollectorMap.put(columnName, + new IntColumnPreIndexStatsCollector(columnName, _statsCollectorConfig)); + break; + case LONG: + _columnStatsCollectorMap.put(columnName, + new LongColumnPreIndexStatsCollector(columnName, _statsCollectorConfig)); + break; + case FLOAT: + _columnStatsCollectorMap.put(columnName, + new FloatColumnPreIndexStatsCollector(columnName, _statsCollectorConfig)); + break; + case DOUBLE: + _columnStatsCollectorMap.put(columnName, + new DoubleColumnPreIndexStatsCollector(columnName, _statsCollectorConfig)); + break; + case BIG_DECIMAL: + _columnStatsCollectorMap.put(columnName, + new BigDecimalColumnPreIndexStatsCollector(columnName, _statsCollectorConfig)); + break; + case STRING: + _columnStatsCollectorMap.put(columnName, + new StringColumnPreIndexStatsCollector(columnName, _statsCollectorConfig)); + break; + case BYTES: + _columnStatsCollectorMap.put(columnName, + new BytesColumnPredIndexStatsCollector(columnName, _statsCollectorConfig)); + break; + case MAP: + _columnStatsCollectorMap.put(columnName, + new MapColumnPreIndexStatsCollector(columnName, _statsCollectorConfig)); + break; + default: + throw new IllegalStateException("Unsupported data type: " + fieldSpec.getDataType()); + } + } + } + + /** + * Collect stats by iterating column-wise using the provided ColumnReader instances. + */ + private void collectColumnStats() { + LOGGER.info("Collecting stats for {} columns using column-wise iteration", _columnReaders.size()); + + for (FieldSpec fieldSpec : _targetSchema.getAllFieldSpecs()) { + if (fieldSpec.isVirtualColumn()) { + continue; + } + + String columnName = fieldSpec.getName(); + AbstractColumnStatisticsCollector statsCollector = _columnStatsCollectorMap.get(columnName); + ColumnReader columnReader = _columnReaders.get(columnName); + + if (columnReader == null) { + throw new RuntimeException("Column reader for column " + columnName + " not found"); + } + + LOGGER.debug("Collecting stats for column: {}", columnName); + collectStatsFromColumnReader(columnName, columnReader, statsCollector); + + // Seal the stats collector + statsCollector.seal(); + } + } + + /** + * Collect stats from a column reader by iterating over all values using the iterator pattern. + */ + private void collectStatsFromColumnReader(String columnName, ColumnReader columnReader, + AbstractColumnStatisticsCollector statsCollector) { + try { + // Reset the column reader to start from the beginning + columnReader.rewind(); + + int docCount = 0; + while (columnReader.hasNext()) { + Object value = columnReader.next(); + statsCollector.collect(value); + docCount++; + } + + // if totalDocCount is unset then set total doc count from the first column + if (_totalDocCount == -1) { + _totalDocCount = docCount; + } else if (_totalDocCount != docCount) { + // all columns should have same count + LOGGER.warn("Column {} has {} documents, but expected {} documents", Review Comment: Putting assumptions with max might be confusing and we need to handle missing data for smaller columns. Can throw an error here to fail early. ########## pinot-segment-local/src/main/java/org/apache/pinot/segment/local/segment/creator/impl/stats/ColumnarSegmentPreIndexStatsContainer.java: ########## @@ -0,0 +1,217 @@ +/** + * 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.pinot.segment.local.segment.creator.impl.stats; + +import java.io.IOException; +import java.util.HashMap; +import java.util.Map; +import org.apache.pinot.segment.spi.creator.ColumnStatistics; +import org.apache.pinot.segment.spi.creator.SegmentPreIndexStatsCollector; +import org.apache.pinot.segment.spi.creator.StatsCollectorConfig; +import org.apache.pinot.spi.data.FieldSpec; +import org.apache.pinot.spi.data.Schema; +import org.apache.pinot.spi.data.readers.ColumnReader; +import org.apache.pinot.spi.data.readers.GenericRow; +import org.slf4j.Logger; +import org.slf4j.LoggerFactory; + + +/** + * Stats container that efficiently collects statistics from columnar data using ColumnReader instances. + * + * <p>This implementation collects statistics by iterating column-wise instead of row-wise, + * which is more efficient for columnar data sources. It supports: + * <ul> + * <li>Column-wise statistics collection</li> + * <li>Existing columns from source data</li> + * <li>New columns with default values</li> + * <li>Data type conversions during schema evolution</li> + * </ul> + * + * <p>The statistics are collected using the same underlying collectors as the row-based approach + * (SegmentPreIndexStatsCollectorImpl) but with more efficient column-wise iteration. + */ +public class ColumnarSegmentPreIndexStatsContainer implements SegmentPreIndexStatsCollector { + private static final Logger LOGGER = LoggerFactory.getLogger(ColumnarSegmentPreIndexStatsContainer.class); + + private final Map<String, ColumnReader> _columnReaders; + private final StatsCollectorConfig _statsCollectorConfig; + private final Schema _targetSchema; + private final Map<String, AbstractColumnStatisticsCollector> _columnStatsCollectorMap; + private int _totalDocCount; + + /** + * Create a ColumnarSegmentPreIndexStatsContainer. + * + * @param columnReaders Map of column name to ColumnReader instances + * @param statsCollectorConfig Configuration for statistics collection + */ + public ColumnarSegmentPreIndexStatsContainer(Map<String, ColumnReader> columnReaders, + StatsCollectorConfig statsCollectorConfig) { + _columnReaders = columnReaders; + _statsCollectorConfig = statsCollectorConfig; + _targetSchema = statsCollectorConfig.getSchema(); + _columnStatsCollectorMap = new HashMap<>(); + _totalDocCount = -1; // indicates unset + + initializeStatsCollectors(); + collectColumnStats(); + } + + /** + * Initialize stats collectors for all columns in the target schema. + */ + private void initializeStatsCollectors() { + for (FieldSpec fieldSpec : _targetSchema.getAllFieldSpecs()) { + if (fieldSpec.isVirtualColumn()) { + continue; + } + + String columnName = fieldSpec.getName(); + switch (fieldSpec.getDataType().getStoredType()) { + case INT: + _columnStatsCollectorMap.put(columnName, + new IntColumnPreIndexStatsCollector(columnName, _statsCollectorConfig)); + break; + case LONG: + _columnStatsCollectorMap.put(columnName, + new LongColumnPreIndexStatsCollector(columnName, _statsCollectorConfig)); + break; + case FLOAT: + _columnStatsCollectorMap.put(columnName, + new FloatColumnPreIndexStatsCollector(columnName, _statsCollectorConfig)); + break; + case DOUBLE: + _columnStatsCollectorMap.put(columnName, + new DoubleColumnPreIndexStatsCollector(columnName, _statsCollectorConfig)); + break; + case BIG_DECIMAL: + _columnStatsCollectorMap.put(columnName, + new BigDecimalColumnPreIndexStatsCollector(columnName, _statsCollectorConfig)); + break; + case STRING: + _columnStatsCollectorMap.put(columnName, + new StringColumnPreIndexStatsCollector(columnName, _statsCollectorConfig)); + break; + case BYTES: + _columnStatsCollectorMap.put(columnName, + new BytesColumnPredIndexStatsCollector(columnName, _statsCollectorConfig)); + break; + case MAP: + _columnStatsCollectorMap.put(columnName, + new MapColumnPreIndexStatsCollector(columnName, _statsCollectorConfig)); + break; + default: + throw new IllegalStateException("Unsupported data type: " + fieldSpec.getDataType()); + } + } + } + + /** + * Collect stats by iterating column-wise using the provided ColumnReader instances. + */ + private void collectColumnStats() { + LOGGER.info("Collecting stats for {} columns using column-wise iteration", _columnReaders.size()); + + for (FieldSpec fieldSpec : _targetSchema.getAllFieldSpecs()) { + if (fieldSpec.isVirtualColumn()) { + continue; + } + + String columnName = fieldSpec.getName(); + AbstractColumnStatisticsCollector statsCollector = _columnStatsCollectorMap.get(columnName); + ColumnReader columnReader = _columnReaders.get(columnName); + + if (columnReader == null) { + throw new RuntimeException("Column reader for column " + columnName + " not found"); + } + + LOGGER.debug("Collecting stats for column: {}", columnName); + collectStatsFromColumnReader(columnName, columnReader, statsCollector); + + // Seal the stats collector + statsCollector.seal(); + } + } + + /** + * Collect stats from a column reader by iterating over all values using the iterator pattern. + */ + private void collectStatsFromColumnReader(String columnName, ColumnReader columnReader, + AbstractColumnStatisticsCollector statsCollector) { + try { + // Reset the column reader to start from the beginning + columnReader.rewind(); + + int docCount = 0; + while (columnReader.hasNext()) { + Object value = columnReader.next(); + statsCollector.collect(value); + docCount++; + } + + // if totalDocCount is unset then set total doc count from the first column + if (_totalDocCount == -1) { + _totalDocCount = docCount; + } else if (_totalDocCount != docCount) { + // all columns should have same count + LOGGER.warn("Column {} has {} documents, but expected {} documents", + columnName, docCount, _totalDocCount); + } + } catch (IOException e) { + LOGGER.error("Failed to collect stats for column: {}", columnName, e); + throw new RuntimeException("Failed to collect stats for column: " + columnName, e); + } + } + + @Override + public ColumnStatistics getColumnProfileFor(String column) { + return _columnStatsCollectorMap.get(column); + } + + @Override + public int getTotalDocCount() { + return _totalDocCount; + } + + @Override + public void init() { + // Already initialized in constructor + } + + @Override + public void build() { + // Stats are already collected in constructor + } + + @Override + public void logStats() { + LOGGER.info("Columnar segment stats collection completed for {} columns with {} documents", Review Comment: column stats get persisted as segment metadata in segment file. -- 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: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
