szehon-ho commented on code in PR #6857:
URL: https://github.com/apache/iceberg/pull/6857#discussion_r1110711040
##########
aws/src/main/java/org/apache/iceberg/aws/s3/signer/S3ObjectMapper.java:
##########
@@ -42,26 +43,24 @@
public class S3ObjectMapper {
- private static final JsonFactory FACTORY = new JsonFactory();
- private static final ObjectMapper MAPPER = new ObjectMapper(FACTORY);
- private static volatile boolean isInitialized = false;
+ private static final AtomicReference<ObjectMapper> REFERENCE = new
AtomicReference<>();
private S3ObjectMapper() {}
static ObjectMapper mapper() {
- if (!isInitialized) {
- synchronized (S3ObjectMapper.class) {
- if (!isInitialized) {
- MAPPER.setVisibility(PropertyAccessor.FIELD,
JsonAutoDetect.Visibility.ANY);
- MAPPER.configure(DeserializationFeature.FAIL_ON_UNKNOWN_PROPERTIES,
false);
-
MAPPER.setPropertyNamingStrategy(PropertyNamingStrategies.KebabCaseStrategy.INSTANCE);
- MAPPER.registerModule(initModule());
- isInitialized = true;
- }
- }
+ if (null == REFERENCE.get()) {
Review Comment:
Thanks, the change now looks correct.
I'm a bit on the fence about it, as now different threads init'ing at the
same time will start initializing multiple mappers, which allocates some memory
for nothing, but before they would just wait.
Curious is there some added advantage of atomicReference over double-check
locking, other than code correctness?
--
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]
