You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@pinot.apache.org by GitBox <gi...@apache.org> on 2020/07/16 16:12:49 UTC
[GitHub] [incubator-pinot] mayankshriv commented on a change in pull request #5709: Store column min/max value into segment metadata
mayankshriv commented on a change in pull request #5709:
URL: https://github.com/apache/incubator-pinot/pull/5709#discussion_r455905335
##########
File path: pinot-core/src/main/java/org/apache/pinot/core/segment/index/loader/columnminmaxvalue/ColumnMinMaxValueGenerator.java
##########
@@ -91,33 +92,46 @@ private void addColumnMinMaxValueForColumn(String columnName) throws Exception {
switch (dataType) {
case INT:
try (IntDictionary intDictionary = new IntDictionary(dictionaryBuffer, length)) {
- SegmentColumnarIndexCreator.addColumnMinMaxValueInfo(_segmentProperties, columnName,
- intDictionary.getStringValue(0), intDictionary.getStringValue(length - 1));
+ SegmentColumnarIndexCreator
Review comment:
How about no-dictionary columns? We do go over all the values, so we should be able to compute that information even without dictionary?
##########
File path: pinot-core/src/main/java/org/apache/pinot/core/segment/creator/impl/SegmentColumnarIndexCreator.java
##########
@@ -509,71 +509,61 @@ public static void addColumnMetadataInfo(PropertiesConfiguration properties, Str
properties.setProperty(getKeyFor(column, HAS_NULL_VALUE), String.valueOf(columnIndexCreationInfo.hasNulls()));
properties.setProperty(getKeyFor(column, HAS_DICTIONARY), String.valueOf(hasDictionary));
properties.setProperty(getKeyFor(column, TEXT_INDEX_TYPE), textIndexType.name());
- properties.setProperty(V1Constants.MetadataKeys.Column.getKeyFor(column, HAS_INVERTED_INDEX),
- String.valueOf(hasInvertedIndex));
- properties.setProperty(V1Constants.MetadataKeys.Column.getKeyFor(column, IS_SINGLE_VALUED),
- String.valueOf(fieldSpec.isSingleValueField()));
- properties.setProperty(V1Constants.MetadataKeys.Column.getKeyFor(column, MAX_MULTI_VALUE_ELEMTS),
+ properties.setProperty(getKeyFor(column, HAS_INVERTED_INDEX), String.valueOf(hasInvertedIndex));
+ properties.setProperty(getKeyFor(column, IS_SINGLE_VALUED), String.valueOf(fieldSpec.isSingleValueField()));
+ properties.setProperty(getKeyFor(column, MAX_MULTI_VALUE_ELEMTS),
String.valueOf(columnIndexCreationInfo.getMaxNumberOfMultiValueElements()));
- properties.setProperty(V1Constants.MetadataKeys.Column.getKeyFor(column, TOTAL_NUMBER_OF_ENTRIES),
+ properties.setProperty(getKeyFor(column, TOTAL_NUMBER_OF_ENTRIES),
String.valueOf(columnIndexCreationInfo.getTotalNumberOfEntries()));
- properties.setProperty(V1Constants.MetadataKeys.Column.getKeyFor(column, IS_AUTO_GENERATED),
- String.valueOf(columnIndexCreationInfo.isAutoGenerated()));
+ properties
+ .setProperty(getKeyFor(column, IS_AUTO_GENERATED), String.valueOf(columnIndexCreationInfo.isAutoGenerated()));
PartitionFunction partitionFunction = columnIndexCreationInfo.getPartitionFunction();
if (partitionFunction != null) {
- properties.setProperty(V1Constants.MetadataKeys.Column.getKeyFor(column, PARTITION_FUNCTION),
- partitionFunction.toString());
- properties.setProperty(V1Constants.MetadataKeys.Column.getKeyFor(column, NUM_PARTITIONS),
- columnIndexCreationInfo.getNumPartitions());
- properties.setProperty(V1Constants.MetadataKeys.Column.getKeyFor(column, PARTITION_VALUES),
- columnIndexCreationInfo.getPartitions());
+ properties.setProperty(getKeyFor(column, PARTITION_FUNCTION), partitionFunction.toString());
+ properties.setProperty(getKeyFor(column, NUM_PARTITIONS), columnIndexCreationInfo.getNumPartitions());
+ properties.setProperty(getKeyFor(column, PARTITION_VALUES), columnIndexCreationInfo.getPartitions());
}
// datetime field
if (fieldSpec.getFieldType().equals(FieldType.DATE_TIME)) {
DateTimeFieldSpec dateTimeFieldSpec = (DateTimeFieldSpec) fieldSpec;
- properties.setProperty(V1Constants.MetadataKeys.Column.getKeyFor(column, DATETIME_FORMAT),
- dateTimeFieldSpec.getFormat());
- properties.setProperty(V1Constants.MetadataKeys.Column.getKeyFor(column, DATETIME_GRANULARITY),
- dateTimeFieldSpec.getGranularity());
+ properties.setProperty(getKeyFor(column, DATETIME_FORMAT), dateTimeFieldSpec.getFormat());
+ properties.setProperty(getKeyFor(column, DATETIME_GRANULARITY), dateTimeFieldSpec.getGranularity());
}
- Object defaultNullValue = columnIndexCreationInfo.getDefaultNullValue();
- if (defaultNullValue instanceof byte[]) {
- String defaultNullValueString = BytesUtils.toHexString((byte[]) defaultNullValue);
- properties
- .setProperty(V1Constants.MetadataKeys.Column.getKeyFor(column, DEFAULT_NULL_VALUE), defaultNullValueString);
+ String minValue = columnIndexCreationInfo.getMin().toString();
+ String maxValue = columnIndexCreationInfo.getMax().toString();
+ String defaultNullValue = columnIndexCreationInfo.getDefaultNullValue().toString();
+ if (dataType == DataType.STRING) {
+ // Escape special character for STRING column
+ properties.setProperty(getKeyFor(column, MIN_VALUE), escapeSpecialCharacter(minValue));
+ properties.setProperty(getKeyFor(column, MAX_VALUE), escapeSpecialCharacter(maxValue));
+ properties.setProperty(getKeyFor(column, DEFAULT_NULL_VALUE), escapeSpecialCharacter(defaultNullValue));
} else {
- properties.setProperty(V1Constants.MetadataKeys.Column.getKeyFor(column, DEFAULT_NULL_VALUE),
- String.valueOf(defaultNullValue));
+ properties.setProperty(getKeyFor(column, MIN_VALUE), minValue);
+ properties.setProperty(getKeyFor(column, MAX_VALUE), maxValue);
+ properties.setProperty(getKeyFor(column, DEFAULT_NULL_VALUE), defaultNullValue);
}
}
public static void addColumnMinMaxValueInfo(PropertiesConfiguration properties, String column, String minValue,
String maxValue) {
- properties.setProperty(getKeyFor(column, MIN_VALUE), minValue);
- properties.setProperty(getKeyFor(column, MAX_VALUE), maxValue);
+ // Escape special character for STRING column
+ properties.setProperty(getKeyFor(column, MIN_VALUE), escapeSpecialCharacter(minValue));
+ properties.setProperty(getKeyFor(column, MAX_VALUE), escapeSpecialCharacter(maxValue));
+ }
+
+ /**
+ * Helper method to escape special character for the property value.
+ */
+ private static String escapeSpecialCharacter(String value) {
Review comment:
Since this is only escaping line character, so may be it should be name `escapeLineCharacter`?
----------------------------------------------------------------
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.
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