/**
     * Time encoding type: 1970 epoch, 32-bit seconds, big-endian.
     */
    public static final int TIME_1970_SEC_32BE = 1;

    /**
     * Time encoding type: 1970 epoch, 32-bit seconds, little-endian.
     */
    public static final int TIME_1970_SEC_32LE = 2;

    /**
     * Time encoding type: 1904 epoch, 32-bit seconds, big-endian.
     */
    public static final int TIME_1904_SEC_32BE = 3;

    /**

    @Test
    @DisplayName("readULong correctly interprets little endian")
    void testReadULong() {
        byte[] a = new byte[] { (byte) 0x01, (byte) 0x02, (byte) 0x03, (byte) 0x04 };
        int result = NtlmMessage.readULong(a, 0);
        assertEquals(0x04030201, result, "ULong should be little‑endian");
    }

    @Test
    @DisplayName("readUShort correctly interprets little endian")
    void testReadUShort() {

    }

    @Test
    public void testToStringContainsCount() {
        // Little-endian: count = 512 (0x0200)
        byte[] buf = new byte[12];
        buf[0] = 0x00;
        buf[1] = 0x02; // count = 512 in little-endian
        resp.readParameterWordsWireFormat(buf, 0);
        String str = resp.toString();

        String msg = "\u00A1\u00A2"; // two Unicode characters
        byte[] encoded = encodeUnicode(msg);
        assertEquals(4, encoded.length, "Encoded Unicode string should be 4 bytes");

        // Verify little-endian encoding
        assertEquals((byte) 0xA1, encoded[0], "First byte of first character");
        assertEquals((byte) 0x00, encoded[1], "Second byte of first character");

   * is big-endian or little-endian depends on the byte order of {@link #asBytes}. This may be
   * surprising for implementations of {@code HashCode} that represent the number in big-endian
   * since everything else in the hashing API uniformly treats multibyte values as little-endian.
   *

* `www_redirect` - Si se establece en True, las requests a versiones sin www de los hosts permitidos serán redirigidas a sus equivalentes con www. Por defecto es `True`.

Si una request entrante no se valida correctamente, se enviará un response `400`.

## `GZipMiddleware` { #gzipmiddleware }

Maneja responses GZip para cualquier request que incluya `"gzip"` en el header `Accept-Encoding`.

            assertEquals(1, buffer[0]); // hash algo count (little endian)
            assertEquals(0, buffer[1]);
            assertEquals(8, buffer[2]); // salt length (little endian)
            assertEquals(0, buffer[3]);
            assertEquals(1, buffer[4]); // SHA512 hash algo (little endian)
            assertEquals(0, buffer[5]);
            assertArrayEquals(testSalt, Arrays.copyOfRange(buffer, 6, 14));

        buf.enc_ndr_small(5); /* RPC version */
        buf.enc_ndr_small(0); /* minor version */
        buf.enc_ndr_small(this.ptype);
        buf.enc_ndr_small(this.flags);
        buf.enc_ndr_long(0x00000010); /* Little-endian / ASCII / IEEE */
        buf.enc_ndr_short(this.length);
        buf.enc_ndr_short(0); /* length of auth_value */
        buf.enc_ndr_long(this.call_id);
    }

### Casos de uso: servicio externo { #use-cases-external-service }

Un ejemplo podría ser que tienes un proveedor de autenticación externo al que necesitas llamar.

Le envías un token y te devuelve un usuario autenticado.

Este proveedor podría estar cobrándote por cada request, y llamarlo podría tomar más tiempo adicional que si tuvieras un usuario de prueba fijo para los tests.

   * is big-endian or little-endian depends on the byte order of {@link #asBytes}. This may be
   * surprising for implementations of {@code HashCode} that represent the number in big-endian
   * since everything else in the hashing API uniformly treats multibyte values as little-endian.
   *