This is true about `put`, but we still need to choose a key to insert into
`sortedMap`. If I don't declare the `nextKey` variable, I need to have a bunch
of redundant code in the if and else blocks:
```java
final TimeKey<Bytes> previousKey = index.get(key);
// non-resetting semantics:
// if there was a previous version of the same record,
// then insert the new record in the same place in the priority queue
if (previousKey == null) {
final TimeKey<Bytes> nextKey = new TimeKey<>(time, key);
index.put(key, nextKey);
sortedMap.put(nextKey, value);
minTimestamp = Math.min(minTimestamp, nextKey.time());
memBufferSize =
memBufferSize
+ computeRecordSize(key, value);
} else {
final ContextualRecord<byte[]> removedValue =
sortedMap.remove(previousKey);
sortedMap.put(previousKey, value);
memBufferSize =
memBufferSize
+ computeRecordSize(key, value)
- (removedValue == null ? 0 : computeRecordSize(key,
removedValue));
}
```
IMHO, this is less readable than the linear version where we just reuse or
construct the key in line 103.
[ Full content available at: https://github.com/apache/kafka/pull/5693 ]
This message was relayed via gitbox.apache.org for [email protected]
