You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@pinot.apache.org by "abhioncbr (via GitHub)" <gi...@apache.org> on 2023/06/23 12:59:45 UTC
[GitHub] [pinot] abhioncbr commented on a diff in pull request #10891: Ensure min/max value generation in the segment metadata.
abhioncbr commented on code in PR #10891:
URL: https://github.com/apache/pinot/pull/10891#discussion_r1239788013
##########
pinot-segment-local/src/main/java/org/apache/pinot/segment/local/segment/index/loader/columnminmaxvalue/ColumnMinMaxValueGenerator.java:
##########
@@ -119,65 +121,126 @@ private List<String> getColumnsToAddMinMaxValue() {
private boolean needAddColumnMinMaxValueForColumn(String columnName) {
ColumnMetadata columnMetadata = _segmentMetadata.getColumnMetadataFor(columnName);
- return columnMetadata.hasDictionary() && columnMetadata.getMinValue() == null
+ return columnMetadata.getMinValue() == null
&& columnMetadata.getMaxValue() == null && !columnMetadata.isMinMaxValueInvalid();
}
private void addColumnMinMaxValueForColumn(String columnName)
throws Exception {
// Skip column without dictionary or with min/max value already set
ColumnMetadata columnMetadata = _segmentMetadata.getColumnMetadataFor(columnName);
- if (!columnMetadata.hasDictionary() || columnMetadata.getMinValue() != null
- || columnMetadata.getMaxValue() != null) {
+ if (columnMetadata.getMinValue() != null || columnMetadata.getMaxValue() != null) {
return;
}
- PinotDataBuffer dictionaryBuffer = _segmentWriter.getIndexFor(columnName, StandardIndexes.dictionary());
DataType dataType = columnMetadata.getDataType().getStoredType();
- int length = columnMetadata.getCardinality();
- switch (dataType) {
- case INT:
- try (IntDictionary intDictionary = new IntDictionary(dictionaryBuffer, length)) {
- SegmentColumnarIndexCreator.addColumnMinMaxValueInfo(_segmentProperties, columnName,
- intDictionary.getStringValue(0), intDictionary.getStringValue(length - 1));
- }
- break;
- case LONG:
- try (LongDictionary longDictionary = new LongDictionary(dictionaryBuffer, length)) {
- SegmentColumnarIndexCreator.addColumnMinMaxValueInfo(_segmentProperties, columnName,
- longDictionary.getStringValue(0), longDictionary.getStringValue(length - 1));
- }
- break;
- case FLOAT:
- try (FloatDictionary floatDictionary = new FloatDictionary(dictionaryBuffer, length)) {
- SegmentColumnarIndexCreator.addColumnMinMaxValueInfo(_segmentProperties, columnName,
- floatDictionary.getStringValue(0), floatDictionary.getStringValue(length - 1));
- }
- break;
- case DOUBLE:
- try (DoubleDictionary doubleDictionary = new DoubleDictionary(dictionaryBuffer, length)) {
- SegmentColumnarIndexCreator.addColumnMinMaxValueInfo(_segmentProperties, columnName,
- doubleDictionary.getStringValue(0), doubleDictionary.getStringValue(length - 1));
- }
- break;
- case STRING:
- try (StringDictionary stringDictionary = new StringDictionary(dictionaryBuffer, length,
- columnMetadata.getColumnMaxLength())) {
- SegmentColumnarIndexCreator.addColumnMinMaxValueInfo(_segmentProperties, columnName,
- stringDictionary.getStringValue(0), stringDictionary.getStringValue(length - 1));
- }
- break;
- case BYTES:
- try (BytesDictionary bytesDictionary = new BytesDictionary(dictionaryBuffer, length,
- columnMetadata.getColumnMaxLength())) {
- SegmentColumnarIndexCreator.addColumnMinMaxValueInfo(_segmentProperties, columnName,
- bytesDictionary.getStringValue(0), bytesDictionary.getStringValue(length - 1));
- }
- break;
- default:
- throw new IllegalStateException("Unsupported data type: " + dataType + " for column: " + columnName);
+ if (columnMetadata.hasDictionary()) {
+ PinotDataBuffer dictionaryBuffer = _segmentWriter.getIndexFor(columnName, StandardIndexes.dictionary());
+ int length = columnMetadata.getCardinality();
+ switch (dataType) {
+ case INT:
+ try (IntDictionary intDictionary = new IntDictionary(dictionaryBuffer, length)) {
+ SegmentColumnarIndexCreator.addColumnMinMaxValueInfo(_segmentProperties, columnName,
+ intDictionary.getStringValue(0), intDictionary.getStringValue(length - 1));
+ }
+ break;
+ case LONG:
+ try (LongDictionary longDictionary = new LongDictionary(dictionaryBuffer, length)) {
+ SegmentColumnarIndexCreator.addColumnMinMaxValueInfo(_segmentProperties, columnName,
+ longDictionary.getStringValue(0), longDictionary.getStringValue(length - 1));
+ }
+ break;
+ case FLOAT:
+ try (FloatDictionary floatDictionary = new FloatDictionary(dictionaryBuffer, length)) {
+ SegmentColumnarIndexCreator.addColumnMinMaxValueInfo(_segmentProperties, columnName,
+ floatDictionary.getStringValue(0), floatDictionary.getStringValue(length - 1));
+ }
+ break;
+ case DOUBLE:
+ try (DoubleDictionary doubleDictionary = new DoubleDictionary(dictionaryBuffer, length)) {
+ SegmentColumnarIndexCreator.addColumnMinMaxValueInfo(_segmentProperties, columnName,
+ doubleDictionary.getStringValue(0), doubleDictionary.getStringValue(length - 1));
+ }
+ break;
+ case STRING:
+ try (StringDictionary stringDictionary = new StringDictionary(dictionaryBuffer, length,
+ columnMetadata.getColumnMaxLength())) {
+ SegmentColumnarIndexCreator.addColumnMinMaxValueInfo(_segmentProperties, columnName,
+ stringDictionary.getStringValue(0), stringDictionary.getStringValue(length - 1));
+ }
+ break;
+ case BYTES:
+ try (BytesDictionary bytesDictionary = new BytesDictionary(dictionaryBuffer, length,
+ columnMetadata.getColumnMaxLength())) {
+ SegmentColumnarIndexCreator.addColumnMinMaxValueInfo(_segmentProperties, columnName,
+ bytesDictionary.getStringValue(0), bytesDictionary.getStringValue(length - 1));
+ }
+ break;
+ default:
+ throw new IllegalStateException("Unsupported data type: " + dataType + " for column: " + columnName);
+ }
+ } else if (_segmentMetadata.getSchema().getFieldSpecFor(columnName).getFieldType() == FieldSpec.FieldType.METRIC) {
Review Comment:
I have made the change. However, do we also need to support MV fields? If yes, the max and min will be the max and min values seen in any array or something else? Thanks
--
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: commits-unsubscribe@pinot.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@pinot.apache.org
For additional commands, e-mail: commits-help@pinot.apache.org