pri1712 commented on code in PR #19011:
URL: https://github.com/apache/pinot/pull/19011#discussion_r3699020787
##########
pinot-broker/src/main/java/org/apache/pinot/broker/requesthandler/BaseBrokerRequestHandler.java:
##########
@@ -464,4 +480,107 @@ protected void onQueryFinish(long requestId) {
protected boolean isQueryCancellationEnabled() {
return _enableQueryCancellation;
}
+
+ /// Appends a where clause to the query to filter out of retention data if
SKIP_OUT_OF_RETENTION_VALUES is
+ /// set to True in query Options.
+ /// @param sqlNodeAndOptions
+ private void applySkipOutOfRetentionValuesIfNeeded(SqlNodeAndOptions
sqlNodeAndOptions) {
+ Map<String, String> options = sqlNodeAndOptions.getOptions();
+ if
(!Boolean.parseBoolean(options.get(QueryOptionKey.SKIP_OUT_OF_RETENTION_VALUES)))
{
+ return;
+ }
+ SqlNode sqlNode = sqlNodeAndOptions.getSqlNode();
+ if (sqlNode == null) {
+ return;
+ }
+
+ // Resolve the inner SqlSelect, unwrapping SqlOrderBy if needed.
+ // SqlWith (CTEs) and other statement types are not supported.
+ SqlSelect sqlSelect;
+ if (sqlNode instanceof SqlSelect) {
+ sqlSelect = (SqlSelect) sqlNode;
+ } else if (sqlNode instanceof SqlOrderBy) {
+ SqlNode query = ((SqlOrderBy) sqlNode).query;
+ if (!(query instanceof SqlSelect)) {
+ return;
+ }
+ sqlSelect = (SqlSelect) query;
+ } else {
+ return;
+ }
+
+ TableNameExtractor tableNameExtractor = new TableNameExtractor();
+ tableNameExtractor.extractTableNames(sqlSelect);
+ Set<String> tableNames = tableNameExtractor.getTableNames();
+ if (tableNames.isEmpty()) {
+ return;
+ }
+
+ // Multi-table queries (JOINs) are not supported: it is ambiguous which
table's retention
+ // applies, and injecting a filter with the wrong table's time column
would produce
+ // incorrect results. Callers must issue separate single-table queries.
+ if (tableNames.size() > 1) {
+ LOGGER.debug("skipOutOfRetentionValues is not supported for multi-table
queries; skipping filter injection");
+ return;
+ }
+
+ String rawTableName =
TableNameBuilder.extractRawTableName(tableNames.iterator().next());
+
+ TableConfig tableConfig =
_tableCache.getTableConfig(TableNameBuilder.OFFLINE.tableNameWithType(rawTableName));
+ if (tableConfig == null) {
+ tableConfig =
_tableCache.getTableConfig(TableNameBuilder.REALTIME.tableNameWithType(rawTableName));
+ }
+ Schema schema = _tableCache.getSchema(rawTableName);
+ if (tableConfig == null || schema == null ||
tableConfig.getValidationConfig() == null) {
+ return;
+ }
+
+ //get timestamp column and retention details
+ SegmentsValidationAndRetentionConfig validationConfig =
tableConfig.getValidationConfig();
+ String timeColumnName = validationConfig.getTimeColumnName();
+ if (StringUtils.isEmpty(timeColumnName) ||
!validationConfig.hasRetention()) {
+ return;
+ }
+ try {
+ long retentionTimeMs = validationConfig.getRetentionTimeMillis();
+ long cutoffMs = System.currentTimeMillis() - retentionTimeMs;
+ DateTimeFieldSpec timeFieldSpec =
schema.getSpecForTimeColumn(timeColumnName);
+ if (timeFieldSpec == null) {
+ return;
+ }
+
+ DateTimeFormatSpec formatSpec = new
DateTimeFormatSpec(timeFieldSpec.getFormat());
+ String formattedCutoffTime = formatSpec.fromMillisToFormat(cutoffMs);
+
+ //add where clause to the AST
+ SqlIdentifier timeColNode = new SqlIdentifier(timeColumnName,
SqlParserPos.ZERO);
+ SqlLiteral cutoffNode;
+ if (timeFieldSpec.getDataType().isNumeric()) {
+ cutoffNode = SqlLiteral.createExactNumeric(formattedCutoffTime,
SqlParserPos.ZERO);
+ } else {
+ cutoffNode = SqlLiteral.createCharString(formattedCutoffTime,
SqlParserPos.ZERO);
+ }
+
+ SqlBasicCall rangeFilter = new SqlBasicCall(
+ SqlStdOperatorTable.GREATER_THAN_OR_EQUAL, List.of(timeColNode,
cutoffNode),
+ SqlParserPos.ZERO);
+
+ SqlNode existingWhere = sqlSelect.getWhere();
+ if (existingWhere != null) {
+ //add an AND if the where exists
+ SqlBasicCall andExpr = new SqlBasicCall(
+ SqlStdOperatorTable.AND,
+ List.of(existingWhere, rangeFilter),
+ SqlParserPos.ZERO
+ );
+ sqlSelect.setWhere(andExpr);
+ } else {
+ sqlSelect.setWhere(rangeFilter);
+ }
+ LOGGER.debug("Injected Calcite AST filter for skipOutOfRetentionValues
on table: {}. Cutoff: {}", rawTableName,
+ formattedCutoffTime);
+ } catch (RuntimeException e) {
Review Comment:
Every other invalid-config case in the same method (missing retention,
missing time column, no schema) silently skips filter injection instead. (as
mentioned by @tarun11Mavani )
do we want to be more explicit here?
--
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]
