stevenzwu commented on code in PR #14500:
URL: https://github.com/apache/iceberg/pull/14500#discussion_r2756136738
##########
api/src/main/java/org/apache/iceberg/expressions/InclusiveMetricsEvaluator.java:
##########
@@ -56,14 +56,25 @@ public class InclusiveMetricsEvaluator {
private static final int IN_PREDICATE_LIMIT = 200;
private final Expression expr;
+ // Expression using signed UUID comparator for backward compatibility with
files written
+ // prior to the introduction of RFC-compliant UUID comparisons.
+ // Null if there are no UUID predicates comparing against bounds in the
expression.
+ private final Expression signedUuidExpr;
Review Comment:
Instead of generating a `signedUuidExpr`, can we just change the
`compartor.compare(...)` for `Literal` and `BoundTerm` to a wrapper method,
which check the literal or field type and performs the additional signed
comparator? I am wondering if this can reduce the size of the changes.
##########
api/src/main/java/org/apache/iceberg/expressions/Literals.java:
##########
@@ -608,9 +608,24 @@ public String toString() {
}
}
- static class UUIDLiteral extends ComparableLiteral<UUID> {
+ static class UUIDLiteral extends BaseLiteral<UUID> {
+ private static final Comparator<UUID> RFC_CMP =
+ Comparators.<UUID>nullsFirst().thenComparing(Comparators.uuids());
+ private static final Comparator<UUID> SIGNED_CMP =
+
Comparators.<UUID>nullsFirst().thenComparing(Comparators.signedUUIDs());
+
+ // Flag to indicate which comparator to use (serializable)
+ private final boolean useSignedComparator;
+ // Transient cached comparator (reconstructed on deserialization)
+ private transient volatile Comparator<UUID> cmp;
Review Comment:
why not make this final and set it in the constructor?
--
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]
