Skip to content

MLE-22706 Added encode/decode support for vectors #1789

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Jul 10, 2025
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Original file line number Diff line number Diff line change
@@ -0,0 +1,56 @@
/*
* Copyright (c) 2010-2025 Progress Software Corporation and/or its subsidiaries or affiliates. All Rights Reserved.
*/
package com.marklogic.client.util;

import java.nio.ByteBuffer;
import java.nio.ByteOrder;
import java.util.Base64;

/**
* Supports encoding and decoding vectors using the same approach as the vec:base64-encode and vec:base64-decode
* functions supported by the MarkLogic server.
*
* @since 7.2.0
*/
public interface VectorUtil {

/**
* @param vector
* @return a base64-encoded string representing the vector and using the same approach as the vec:base64-encode
* function supported by the MarkLogic server.
*/
static String base64Encode(float... vector) {
final int dimensions = vector.length;
ByteBuffer buffer = ByteBuffer.allocate(8 + 4 * dimensions);
buffer.order(ByteOrder.LITTLE_ENDIAN);
buffer.putInt(0); // version
buffer.putInt(dimensions);
for (float v : vector) {
buffer.putFloat(v);
}
return Base64.getEncoder().encodeToString(buffer.array());
}

/**
* @param encodedVector
* @return a vector represented by the base64-encoded string and using the same approach as the vec:base64-decode
* function supported by the MarkLogic server.
*/
static float[] base64Decode(String encodedVector) {
ByteBuffer buffer = ByteBuffer.wrap(Base64.getDecoder().decode(encodedVector));
buffer.order(ByteOrder.LITTLE_ENDIAN);

final int version = buffer.getInt();
if (version != 0) {
Copy link
Preview

Copilot AI Jul 9, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Use the same named constant instead of the literal 0 when checking the version to maintain consistency with the encoding logic.

Suggested change
if (version != 0) {
if (version != VECTOR_VERSION) {

Copilot uses AI. Check for mistakes.

throw new IllegalArgumentException("Unsupported vector version: " + version);
}

final int dimensions = buffer.getInt();
float[] vector = new float[dimensions];
for (int i = 0; i < dimensions; i++) {
vector[i] = buffer.getFloat();
}
return vector;
}
}
Original file line number Diff line number Diff line change
@@ -0,0 +1,58 @@
/*
* Copyright (c) 2010-2025 Progress Software Corporation and/or its subsidiaries or affiliates. All Rights Reserved.
*/
package com.marklogic.client.util;

import com.fasterxml.jackson.databind.JsonNode;
import com.fasterxml.jackson.databind.node.ArrayNode;
import com.marklogic.client.test.Common;
import org.junit.jupiter.api.Test;

import static org.junit.jupiter.api.Assertions.assertEquals;

class VectorUtilTest {

private final float[] VECTOR = new float[]{3.14f, 1.59f, 2.65f};
private final double ACCEPTABLE_DELTA = 0.0001;

@Test
void encodeAndDecodeWithJavaClient() {
String encoded = VectorUtil.base64Encode(VECTOR);
assertEquals("AAAAAAMAAADD9UhAH4XLP5qZKUA=", encoded);

float[] decoded = VectorUtil.base64Decode(encoded);
assertEquals(VECTOR.length, decoded.length);
for (int i = 0; i < VECTOR.length; i++) {
assertEquals(VECTOR[i], decoded[i], ACCEPTABLE_DELTA);
}
}

@Test
void encodeAndDecodeWithServer() {
String encoded = VectorUtil.base64Encode(VECTOR);
assertEquals("AAAAAAMAAADD9UhAH4XLP5qZKUA=", encoded);

ArrayNode decoded = (ArrayNode) Common.newEvalClient().newServerEval()
Copy link
Preview

Copilot AI Jul 9, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Instead of casting from JsonNode, consider using .evalAs(ArrayNode.class) directly to improve type safety and remove the unchecked cast.

Copilot uses AI. Check for mistakes.

.xquery("vec:base64-decode('%s')".formatted(encoded))
.evalAs(JsonNode.class);

assertEquals(3, decoded.size());
assertEquals(3.14f, decoded.get(0).asDouble(), ACCEPTABLE_DELTA);
assertEquals(1.59f, decoded.get(1).asDouble(), ACCEPTABLE_DELTA);
assertEquals(2.65f, decoded.get(2).asDouble(), ACCEPTABLE_DELTA);
}

@Test
void encodeWithServerAndDecodeWithJavaClient() {
String encoded = Common.newEvalClient().newServerEval()
.xquery("vec:base64-encode(vec:vector((3.14, 1.59, 2.65)))")
.evalAs(String.class);
assertEquals("AAAAAAMAAADD9UhAH4XLP5qZKUA=", encoded);

float[] decoded = VectorUtil.base64Decode(encoded);
assertEquals(VECTOR.length, decoded.length);
for (int i = 0; i < VECTOR.length; i++) {
assertEquals(VECTOR[i], decoded[i], ACCEPTABLE_DELTA);
}
}
}