siddharthteotia commented on code in PR #12223: URL: https://github.com/apache/pinot/pull/12223#discussion_r1462867575
########## pinot-common/src/test/java/org/apache/pinot/common/utils/FALFInternerTest.java: ########## @@ -0,0 +1,168 @@ +/** + * Licensed to the Apache Software Foundation (ASF) under one + * or more contributor license agreements. See the NOTICE file + * distributed with this work for additional information + * regarding copyright ownership. The ASF licenses this file + * to you under the Apache License, Version 2.0 (the + * "License"); you may not use this file except in compliance + * with the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, + * software distributed under the License is distributed on an + * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY + * KIND, either express or implied. See the License for the + * specific language governing permissions and limitations + * under the License. + */ +package org.apache.pinot.common.utils; + +import com.google.common.collect.Interner; +import com.google.common.collect.Interners; +import java.util.Objects; +import java.util.Random; +import org.apache.pinot.spi.utils.FALFInterner; +import org.testng.Assert; +import org.testng.annotations.Test; + + +public class FALFInternerTest { + @Test + public void testInterningByteBuffers() { + Random random = new Random(1); + + int nUniqueObjs = 1024; + int nTotalObjs = 8 * nUniqueObjs; + + String[] allObjs = new String[nTotalObjs]; + + // Create an array of objects where each object should have ~8 copies + for (int i = 0; i < nTotalObjs; i++) { + int next = random.nextInt(nUniqueObjs); + allObjs[i] = Integer.toString(next); + } + + Interner<String> exactInterner = Interners.newStrongInterner(); + Interner<String> falfInterner = new FALFInterner(nUniqueObjs); + Interner<String> falfInternerCustomHash = new FALFInterner(nUniqueObjs, s -> hashCode((String) s), Objects::equals); + + // Go over all objects and intern them using both exact and FALF interners + int nHits1 = runInterning(allObjs, exactInterner, true); + int nHits2 = runInterning(allObjs, falfInterner, true); + int nHits3 = runInterning(allObjs, falfInternerCustomHash, true); + + System.out.println(nHits1); + System.out.println(nHits2); + System.out.println(nHits3); + + // For the exact interner, we should get a hit for each object except the + // first nUniqueObjs. + Assert.assertEquals(nHits1, nTotalObjs - nUniqueObjs); + + // For the FALF interner, due to its fixed size and thus almost inevitable hash + // collisions, the number of hits is smaller. Let's verify that it's not too small, though. + Assert.assertTrue(nHits2 > (nTotalObjs - nUniqueObjs) * 0.4); + + // With the better hash function, FALF interner should have more hits + Assert.assertTrue(nHits3 > (nTotalObjs - nUniqueObjs) * 0.6); + + // Ad hoc benchmarking code. Disabled to avoid test slowdown. + // In one run the MacBook laptop, FALFInterner below performs nearly twice faster + // (1217 ms vs 2230 ms) With custom hash function, FALFInterner's speed is about the + // same as the Guava interner. +// for (int j = 0; j < 3; j++) { +// long time0 = System.currentTimeMillis(); +// long totNHits = 0; +// for (int i = 0; i < 10000; i++) { +// totNHits += runInterning(allObjs, exactInterner, false); +// } +// long time1 = System.currentTimeMillis(); +// System.out.println("Guava interner. totNHits = " + totNHits + ", time = " + (time1 - time0)); +// +// time0 = System.currentTimeMillis(); +// totNHits = 0; +// for (int i = 0; i < 10000; i++) { +// totNHits += runInterning(allObjs, falfInterner, false); +// } +// time1 = System.currentTimeMillis(); +// System.out.println("FALF interner. totNHits = " + totNHits + ", time = " + (time1 - time0)); +// +// time0 = System.currentTimeMillis(); +// totNHits = 0; +// for (int i = 0; i < 10000; i++) { +// totNHits += runInterning(allObjs, falfInternerCustomHash, false); +// } +// time1 = System.currentTimeMillis(); +// System.out.println("FALF interner Custom Hash. totNHits = " + totNHits + ", time = " + (time1 - time0)); +// } + } + + private int runInterning(String[] objs, Interner<String> interner, boolean performAssert) { + int nHits = 0; + for (String origObj : objs) { + String internedObj = interner.intern(origObj); + if (performAssert) { + Assert.assertEquals(origObj, internedObj); + } + if (origObj != internedObj) { + nHits++; + } + } + return nHits; + } + + // Custom hash code implementation, that gives better distribution than standard hashCode() Review Comment: Same question. Why is all this goodness within a test class ? Seems like it can benefit regular code path as well -- 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: commits-unsubscr...@pinot.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: commits-unsubscr...@pinot.apache.org For additional commands, e-mail: commits-h...@pinot.apache.org
