zipkin2.reporter.BytesMessageEncoder Maven / Gradle / Ivy
/*
* Copyright 2016-2024 The OpenZipkin 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 zipkin2.reporter;
import java.util.List;
/**
* Senders like Kafka use byte[] message encoding. This provides helpers to concatenate spans into a
* list.
*/
public enum BytesMessageEncoder {
JSON {
@Override public byte[] encode(List values) {
int sizeOfArray = 2;
int length = values.size();
for (int i = 0; i < length; ) {
sizeOfArray += values.get(i++).length;
if (i < length) sizeOfArray++;
}
byte[] buf = new byte[sizeOfArray];
int pos = 0;
buf[pos++] = '[';
for (int i = 0; i < length; ) {
byte[] v = values.get(i++);
System.arraycopy(v, 0, buf, pos, v.length);
pos += v.length;
if (i < length) buf[pos++] = ',';
}
buf[pos] = ']';
return buf;
}
},
/**
* The first format of Zipkin was TBinaryProtocol, big-endian thrift. It is no longer used, but
* defined here to allow legacy code to migrate to the current reporter library.
*
* This writes the list header followed by a concatenation of the input.
*
* @see Encoding#THRIFT
* @deprecated this format is deprecated in favor of json or proto3
*/
@Deprecated
THRIFT {
@Override public byte[] encode(List values) {
int sizeOfArray = 5;
int length = values.size();
for (int i = 0; i < length; i++) {
sizeOfArray += values.get(i).length;
}
byte[] buf = new byte[sizeOfArray];
int pos = 0;
// TBinaryProtocol List header is element type followed by count
buf[pos++] = 12; // TYPE_STRUCT
buf[pos++] = (byte) ((length >>> 24L) & 0xff);
buf[pos++] = (byte) ((length >>> 16L) & 0xff);
buf[pos++] = (byte) ((length >>> 8L) & 0xff);
buf[pos++] = (byte) (length & 0xff);
// Then each struct is written one-at-a-time with no delimiters until done.
for (int i = 0; i < length; ) {
byte[] v = values.get(i++);
System.arraycopy(v, 0, buf, pos, v.length);
pos += v.length;
}
return buf;
}
},
/**
* This function simply concatenates the byte arrays.
*
* The list of byte arrays represents a repeated (type 2) field. As such, each byte array is
* expected to have a prefix of the field number, followed by the encoded length of the span,
* finally, the actual span bytes.
*
* @see Encoding#PROTO3
*/
PROTO3 {
@Override public byte[] encode(List values) {
int sizeOfArray = 0;
int length = values.size();
for (int i = 0; i < length; ) {
sizeOfArray += values.get(i++).length;
}
byte[] buf = new byte[sizeOfArray];
int pos = 0;
for (int i = 0; i < length; ) {
byte[] v = values.get(i++);
System.arraycopy(v, 0, buf, pos, v.length);
pos += v.length;
}
return buf;
}
};
/**
* Combines a list of encoded spans into an encoded list. For example, in thrift, this would be
* length-prefixed, whereas in json, this would be comma-separated and enclosed by brackets.
*
* The primary use of this is batch reporting spans. For example, spans are {@link
* BytesEncoder#encode(Object) encoded} one-by-one into a queue. This queue is drained up to a byte
* threshold. Then, the list is encoded with this function and reported out-of-process.
*/
public abstract byte[] encode(List encodedSpans);
public static BytesMessageEncoder forEncoding(Encoding encoding) {
if (encoding == null) throw new NullPointerException("encoding == null");
switch (encoding) {
case JSON:
return JSON;
case PROTO3:
return PROTO3;
case THRIFT:
return THRIFT;
default:
throw new UnsupportedOperationException(encoding.name());
}
}
}