Added Google Caliper microbenchmark suite for every metric and algorithm.

1 files changed, 55 insertions, 0 deletions

diff --git a/core/source/benchmark/scala/org/hashtree/stringmetric/similarity/NGramMetricBenchmark.scala b/core/source/benchmark/scala/org/hashtree/stringmetric/similarity/NGramMetricBenchmark.scala
new file mode 100755
index 0000000..c9c0717
--- /dev/null
+++ b/core/source/benchmark/scala/org/hashtree/stringmetric/similarity/NGramMetricBenchmark.scala
@@ -0,0 +1,55 @@
+package org.hashtree.stringmetric.similarity
+
+import com.google.caliper.Param
+import org.hashtree.stringmetric.{ CaliperBenchmark, CaliperRunner }
+import scala.annotation.tailrec
+import scala.util.Random
+
+final class NGramMetricBenchmark extends CaliperBenchmark {
+	@Param(Array("0", "1", "2", "4", "8", "16"))
+	var length: Int = _
+
+	@Param(Array("2", "3"))
+	var n: Int = _
+
+	var string1: String = _
+	var charArray1: Array[Char] = _
+	var string2: String = _
+	var charArray2: Array[Char] = _
+
+	override protected def setUp() {
+		@tailrec
+		def random(l: Int, ps: String = null): String = {
+			if (l == 0) ""
+			else {
+				val s = Random.alphanumeric.take(l).mkString
+
+				if (ps == null || s != ps) s
+				else random(l, ps)
+			}
+		}
+
+		string1 = random(length)
+		string2 = random(length, string1)
+		charArray1 = string1.toCharArray
+		charArray2 = string2.toCharArray
+	}
+
+	def timeCompareWithDifferentCharArrays(reps: Int) = run(reps) {
+		NGramMetric.compare(charArray1, charArray2)(n)
+	}
+
+	def timeCompareWithDifferentStrings(reps: Int) = run(reps) {
+		NGramMetric.compare(string1, string2)(n)
+	}
+
+	def timeCompareWithIdenticalCharArrays(reps: Int) = run(reps) {
+		NGramMetric.compare(charArray1, charArray1)(n)
+	}
+
+	def timeCompareWithIdenticalStrings(reps: Int) = run(reps) {
+		NGramMetric.compare(string1, string1)(n)
+	}
+}
+
+object NGramMetricBenchmark extends CaliperRunner(classOf[NGramMetricBenchmark])
\ No newline at end of file