* @see org.eclipse.aether.RepositoryEvent
     */
    void onEvent(Object event) throws Exception;

    /**
     * Notifies the spy of Maven's termination, allowing it to free any resources allocated by it.
     */
    void close() throws Exception;

 */
package org.apache.maven.model.locator;

import java.io.File;

/**
 * Locates a POM file within a project base directory.
 *
 * @deprecated use {@code org.apache.maven.api.services.ModelBuilder} instead
 */
@Deprecated(since = "4.0.0")
public interface ModelLocator {

    /**
     * Locates the POM file within the specified project directory. In case the given project directory does not exist

   * errors, for better performance.
   *
   * <p>For a new instance, where the arrays above have not yet been allocated, the value of {@code
   * metadata} is the size that the arrays should be allocated with. Once the arrays have been
   * allocated, the value of {@code metadata} combines the number of bits in the "short hash", in

Obligatory benchmarks performed on files with arrays of objects, where the decoded objects are to be extracted.

Two file sizes are used -- regular (1.6mb, 1000 objects) and large (128mb, 100000 objects)

input size | lib | MB/s | Allocated
--- | --- | --- | ---
regular | standard | 97 | 3.6MB
regular | jstream | 175 | 2.1MB
large | standard | 92 | 305MB
large | jstream | 404 | 69MB

    }

    @Test
    void testWriteULong() {
        // Test writeULong
        byte[] data = new byte[4];
        ByteBuffer expectedBuffer = ByteBuffer.allocate(4).order(ByteOrder.LITTLE_ENDIAN);

        // Test with a positive value
        int value = 0x12345678;
        NtlmMessage.writeULong(data, 0, value);
        expectedBuffer.putInt(0, value);

 *
 * <p>{@link CharStreams#copy} has type specific optimizations for various common Appendable and
 * Reader implementations, this compares the performance of the different options.
 */
// These benchmarks allocate a lot of data so use a large heap
@VmOptions({"-Xms12g", "-Xmx12g", "-d64"})
@NullUnmarked
public class CharStreamsCopyBenchmark {
  enum CopyStrategy {
    OLD {
      @Override

     * {@link NtTransQuerySecurityDesc#writeParametersWireFormat(byte[], int)}.
     */
    private static byte[] expectedParameters(int fid, int securityInformation) {
        ByteBuffer bb = ByteBuffer.allocate(8).order(ByteOrder.LITTLE_ENDIAN);
        bb.putShort((short) fid);
        bb.put((byte) 0x00); // Reserved
        bb.put((byte) 0x00); // Reserved
        bb.putInt(securityInformation);
        return bb.array();

   * errors, for better performance.
   *
   * <p>For a new instance, where the arrays above have not yet been allocated, the value of {@code
   * metadata} is the size that the arrays should be allocated with. Once the arrays have been
   * allocated, the value of {@code metadata} combines the number of bits in the "short hash", in

class PacSignatureTest {

    /**
     * Helper method to write integer in little-endian format.
     */
    private void writeLittleEndianInt(ByteArrayOutputStream baos, int value) {
        ByteBuffer buffer = ByteBuffer.allocate(4);
        buffer.order(ByteOrder.LITTLE_ENDIAN);
        buffer.putInt(value);
        baos.write(buffer.array(), 0, 4);
    }

    /**
     * Test constructor with KERB_CHECKSUM_HMAC_MD5 type.
     *

* IDE warnings become more accurate, reducing warnings fatigue.
* Public API nullability annotations become closer to reality.
* Java compilation of Gradle code suffers from small performance penalty (up to 10%).
* Time has to be allocated to clean up existing projects.
* One has to write NullAway-conformant code even when hacking.
* Some patterns used to satisfy NullAway can be slightly more verbose.