kbendick commented on a change in pull request #751:
URL: https://github.com/apache/orc/pull/751#discussion_r671756394
##########
File path: java/core/src/java/org/apache/orc/impl/StringHashTableDictionary.java
##########
@@ -75,18 +75,19 @@ public StringHashTableDictionary(int initialCapacity, float
loadFactor) {
this.capacity = initialCapacity;
this.loadFactor = loadFactor;
this.keyOffsets = new DynamicIntArray(initialCapacity);
- initHashBuckets(initialCapacity);
+ initHashBuckets(false);
this.threshold = (int)Math.min(initialCapacity * loadFactor,
MAX_ARRAY_SIZE + 1);
}
- private void initHashBuckets(int capacity) {
- DynamicIntArray[] buckets = new DynamicIntArray[capacity];
- for (int i = 0; i < capacity; i++) {
+ private void initHashBuckets(final boolean reinit) {
+ final DynamicIntArray[] newBuckets =
+ (reinit) ? this.hashBuckets : new DynamicIntArray[this.capacity];
Review comment:
Would there be any benefit in `clear`ing out or nulling out the current
elements to ensure they're able to be GC'd as soon as possible? I've noticed
that this new `hash` dictionary uses more memory than the `rbtree` dictionary
in some tests I'm working on.
--
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]
