jmalkin commented on code in PR #644:
URL: https://github.com/apache/datasketches-java/pull/644#discussion_r1931227360
##########
src/main/java/org/apache/datasketches/theta/UnionImpl.java:
##########
@@ -328,38 +327,13 @@ public void union(final Sketch sketchIn) {
unionThetaLong_ = min(min(unionThetaLong_, sketchIn.getThetaLong()),
gadget_.getThetaLong()); //Theta rule
unionEmpty_ = false;
- final int curCountIn = sketchIn.getRetainedEntries(true);
- if (curCountIn > 0) {
- if (sketchIn.isOrdered() && (sketchIn instanceof CompactSketch)) { //Use
early stop
- //Ordered, thus compact
- if (sketchIn.hasMemory()) {
- final Memory skMem = sketchIn.getMemory();
- final int preambleLongs = skMem.getByte(PREAMBLE_LONGS_BYTE) & 0X3F;
- for (int i = 0; i < curCountIn; i++ ) {
- final int offsetBytes = preambleLongs + i << 3;
- final long hashIn = skMem.getLong(offsetBytes);
- if (hashIn >= unionThetaLong_) { break; } // "early stop"
- gadget_.hashUpdate(hashIn); //backdoor update, hash function is
bypassed
- }
- }
- else { //sketchIn is on the Java Heap or has array
- final long[] cacheIn = sketchIn.getCache(); //not a copy!
- for (int i = 0; i < curCountIn; i++ ) {
- final long hashIn = cacheIn[i];
- if (hashIn >= unionThetaLong_) { break; } // "early stop"
- gadget_.hashUpdate(hashIn); //backdoor update, hash function is
bypassed
- }
- }
- } //End ordered, compact
- else { //either not-ordered compact or Hash Table form. A HT may have
dirty values.
- final long[] cacheIn = sketchIn.getCache(); //if off-heap this will be
a copy
- final int arrLongs = cacheIn.length;
- for (int i = 0, c = 0; i < arrLongs && c < curCountIn; i++ ) {
- final long hashIn = cacheIn[i];
- if (hashIn <= 0L || hashIn >= unionThetaLong_) { continue; }
//rejects dirty values
- gadget_.hashUpdate(hashIn); //backdoor update, hash function is
bypassed
- c++; //ensures against invalid state inside the incoming sketch
- }
+ HashIterator it = sketchIn.iterator();
+ while (it.next()) {
+ final long hash = it.get();
+ if (hash < unionThetaLong_ && hash < gadget_.getThetaLong()) {
+ gadget_.hashUpdate(hash); // backdoor update, hash function is bypassed
+ } else {
+ if (sketchIn.isOrdered()) { break; }
Review Comment:
The comment on intersection appeared higher in my review. So the longer
version of the relevant comment is that one.
--
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]
