Add ProbabilitySampler in SDK. (#337)

* Add ProbabilitySampler in SDK.

* Add more tests.

* Fix merging error.

sdk/src/main/java/io/opentelemetry/sdk/trace/samplers/ProbabilitySampler.java

@@ -0,0 +1,135 @@
+/*
+ * Copyright 2019, OpenTelemetry Authors
+ *
+ * Licensed 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 io.opentelemetry.sdk.trace.samplers;
+
+import com.google.auto.value.AutoValue;
+import com.google.common.base.Preconditions;
+import io.opentelemetry.trace.AttributeValue;
+import io.opentelemetry.trace.Sampler;
+import io.opentelemetry.trace.Span;
+import io.opentelemetry.trace.SpanContext;
+import io.opentelemetry.trace.SpanId;
+import io.opentelemetry.trace.TraceId;
+import java.util.Collections;
+import java.util.List;
+import java.util.Map;
+import javax.annotation.Nullable;
+import javax.annotation.concurrent.Immutable;
+
+/**
+ * We assume the lower 64 bits of the traceId's are randomly distributed around the whole (long)
+ * range. We convert an incoming probability into an upper bound on that value, such that we can
+ * just compare the absolute value of the id and the bound to see if we are within the desired
+ * probability range. Using the low bits of the traceId also ensures that systems that only use 64
+ * bit ID's will also work with this sampler.
+ */
+@AutoValue
+@Immutable
+public abstract class ProbabilitySampler implements Sampler {
+
+  ProbabilitySampler() {}
+
+  abstract double getProbability();
+
+  abstract long getIdUpperBound();
+
+  /**
+   * Returns a new {@link ProbabilitySampler}. The probability of sampling a trace is equal to that
+   * of the specified probability.
+   *
+   * @param probability The desired probability of sampling. Must be within [0.0, 1.0].
+   * @return a new {@link ProbabilitySampler}.
+   * @throws IllegalArgumentException if {@code probability} is out of range
+   */
+  public static ProbabilitySampler create(double probability) {
+    Preconditions.checkArgument(
+        probability >= 0.0 && probability <= 1.0, "probability must be in range [0.0, 1.0]");
+    long idUpperBound;
+    // Special case the limits, to avoid any possible issues with lack of precision across
+    // double/long boundaries. For probability == 0.0, we use Long.MIN_VALUE as this guarantees
+    // that we will never sample a trace, even in the case where the id == Long.MIN_VALUE, since
+    // Math.Abs(Long.MIN_VALUE) == Long.MIN_VALUE.
+    if (probability == 0.0) {
+      idUpperBound = Long.MIN_VALUE;
+    } else if (probability == 1.0) {
+      idUpperBound = Long.MAX_VALUE;
+    } else {
+      idUpperBound = (long) (probability * Long.MAX_VALUE);
+    }
+    return new AutoValue_ProbabilitySampler(probability, idUpperBound);
+  }
+
+  @Override
+  public final Decision shouldSample(
+      @Nullable SpanContext parentContext,
+      @Nullable Boolean hasRemoteParent,
+      TraceId traceId,
+      SpanId spanId,
+      String name,
+      @Nullable List<Span> parentLinks) {
+    // If the parent is sampled keep the sampling decision.
+    if (parentContext != null && parentContext.getTraceOptions().isSampled()) {
+      return new SimpleDecision(true);
+    }
+    if (parentLinks != null) {
+      // If any parent link is sampled keep the sampling decision.
+      for (Span parentLink : parentLinks) {
+        if (parentLink.getContext().getTraceOptions().isSampled()) {
+          return new SimpleDecision(true);
+        }
+      }
+    }
+    // Always sample if we are within probability range. This is true even for child spans (that
+    // may have had a different sampling decision made) to allow for different sampling policies,
+    // and dynamic increases to sampling probabilities for debugging purposes.
+    // Note use of '<' for comparison. This ensures that we never sample for probability == 0.0,
+    // while allowing for a (very) small chance of *not* sampling if the id == Long.MAX_VALUE.
+    // This is considered a reasonable tradeoff for the simplicity/performance requirements (this
+    // code is executed in-line for every Span creation).
+    return new SimpleDecision(Math.abs(traceId.getLowerLong()) < getIdUpperBound());
+  }
+
+  @Override
+  public final String getDescription() {
+    return String.format("ProbabilitySampler{%.6f}", getProbability());
+  }
+
+  /** Sampling decision without attributes. */
+  private static final class SimpleDecision implements Decision {
+
+    private final boolean decision;
+
+    /**
+     * Creates sampling decision without attributes.
+     *
+     * @param decision sampling decision
+     */
+    SimpleDecision(boolean decision) {
+      this.decision = decision;
+    }
+
+    @Override
+    public boolean isSampled() {
+      return decision;
+    }
+
+    @Override
+    public Map<String, AttributeValue> attributes() {
+      return Collections.emptyMap();
+    }
+  }
+}