import java.util.HashMap;

import jcifs.smb1.util.Encdec;

/**
 * NDR (Network Data Representation) buffer for encoding and decoding DCE/RPC messages.
 * This class provides methods for reading and writing primitive types and strings
 * in NDR format for DCE/RPC communication.
 */
public class NdrBuffer {
    int referent;
    HashMap referents;

    static class Entry {
        int referent;

Par exemple, si vous aviez une session de base de données dans une dépendance avec `yield`, la `StreamingResponse` ne pourrait pas utiliser cette session pendant le streaming des données, car la session aurait déjà été fermée dans le code d’arrêt après `yield`.

Ce comportement a été annulé en 0.118.0, afin que le code d’arrêt après `yield` s’exécute après l’envoi de la réponse.

/// info

/// tip

Instead of returning a `StreamingResponse` directly, you should probably follow the style in [Stream Data](./stream-data.md), it's much more convenient and handles cancellation behind the scenes for you.

If you are streaming JSON Lines, follow the [Stream JSON Lines](../tutorial/stream-json-lines.md) tutorial.

///

### `FileResponse` { #fileresponse }

Asynchronously streams a file as the response.

/**
 * An object that accurately measures <i>elapsed time</i>: the measured duration between two
 * successive readings of "now" in the same process.
 *
 * <p>In contrast, <i>wall time</i> is a reading of "now" as given by a method like
 * {@link System#currentTimeMillis()}, best represented as an {@link java.time.Instant}. Such values
 * <i>can</i> be subtracted to obtain a {@link Duration} (such as by {@link Duration#between}), but

 * of the access check algorithm used by Windows is required. The following
 * is a basic description of the algorithm. For a more complete description
 * we recommend reading the section on Access Control in Keith Brown's
 * "The .NET Developer's Guide to Windows Security" (which is also
 * available online).
 * <p>
 * Direct ACEs are evaluated first in order. The SID of the user performing

        // Test zero time separately - it's handled specially
        SMBUtil.writeTime(0L, buffer, 0);
        long readTime = SMBUtil.readTime(buffer, 0);
        // When writing 0, it writes 0 directly, and when reading 0, it returns negative offset
        assertEquals(-SmbConstants.MILLISECONDS_BETWEEN_1970_AND_1601, readTime);
    }

    @Test
    void testUTimeConversionRoundTrip() {

org.apache.lucene.search.TimeLimitingCollector#getGlobalCounter()

@defaultMessage Don't interrupt threads use FutureUtils#cancel(Future<T>) instead
java.util.concurrent.Future#cancel(boolean)

@defaultMessage Don't try reading from paths that are not configured in Environment, resolve from Environment instead
org.elasticsearch.core.PathUtils#get(java.lang.String, java.lang.String[])
org.elasticsearch.core.PathUtils#get(java.net.URI)

        void testDecodeWithBufferOverflow() {
            assertThrows(SMBProtocolDecodingException.class, () -> testResponse.decode(testBuffer, 250, 10),
                    "Should throw exception when reading beyond buffer bounds");
        }

        @Test
        @DisplayName("Should handle zero-length decode")
        void testDecodeWithZeroLength() throws SMBProtocolDecodingException {

}

// Target - Kafka target.
type Target struct {
	status int32

	totalMessages  int64
	failedMessages int64

	wg sync.WaitGroup

	// Channel of log entries.
	// Reading logCh must hold read lock on logChMu (to avoid read race)
	// Sending a value on logCh must hold read lock on logChMu (to avoid closing)
	logCh   chan any
	logChMu sync.RWMutex

            }
            maxLineBuffer = maxOutputLineBuffer;
        }

        /**
         * Runs the input stream thread, reading lines and buffering them.
         */
        @Override
        public void run() {
            for (;;) {
                try {
                    final String line = br.readLine();