clintropolis commented on code in PR #18524:
URL: https://github.com/apache/druid/pull/18524#discussion_r2353781587
##########
server/src/main/java/org/apache/druid/segment/indexing/DataSchema.java:
##########
@@ -415,9 +415,73 @@ private static Set<String>
computeAndValidateOutputFieldNames(
}
}
+ return getFieldsOrThrowIfErrors(fields);
+ }
+
+ public static void validateProjections(
+ @Nullable List<AggregateProjectionSpec> projections,
+ @Nullable Granularity segmentGranularity
+ )
+ {
+ if (projections != null) {
+ final Set<String> names =
Sets.newHashSetWithExpectedSize(projections.size());
+ for (AggregateProjectionSpec projection : projections) {
+ if (names.contains(projection.getName())) {
+ throw InvalidInput.exception("projection[%s] is already defined,
projection names must be unique", projection.getName());
+ }
+ names.add(projection.getName());
+ final AggregateProjectionMetadata.Schema schema =
projection.toMetadataSchema();
+
+ if (schema.getTimeColumnName() != null) {
+ final Granularity projectionGranularity =
schema.getEffectiveGranularity();
+ if (segmentGranularity != null) {
+ if (segmentGranularity.isFinerThan(projectionGranularity)) {
+ throw InvalidInput.exception(
+ "projection[%s] has granularity[%s] which must be finer than
or equal to segment granularity[%s]",
+ projection.getName(),
+ projectionGranularity,
+ segmentGranularity
+ );
+ }
+ }
+ }
+
+ final Map<String, Multiset<String>> fields = new TreeMap<>();
+ int position = 0;
+ for (DimensionSchema grouping : projection.getGroupingColumns()) {
+ final String field = grouping.getName();
+ if (Strings.isNullOrEmpty(field)) {
+ throw DruidException
+ .forPersona(DruidException.Persona.USER)
+ .ofCategory(DruidException.Category.INVALID_INPUT)
+ .build("Encountered grouping column with null or empty name at
position[%d]", position);
+ }
+ fields.computeIfAbsent(field, k ->
TreeMultiset.create()).add("projection[" + projection.getName() + "] grouping
column list");
Review Comment:
this is to detect if there are grouping columns and aggs with the same name,
see other comment about following the pattern of
`computeAndValidateOutputFieldNames` both of which are making that the output
schema is valid
--
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]
