The result of calling it is something that can be encoded with the Python standard <a href="https://docs.python.org/3/library/json.html#json.dumps" class="external-link" target="_blank">`json.dumps()`</a>.

It doesn't return a large `str` containing the data in JSON format (as a string). It returns a Python standard data structure (e.g. a `dict`) with values and sub-values that are all compatible with JSON.

!!! note

        axesSizeProduct[i] = IntMath.checkedMultiply(axesSizeProduct[i + 1], axes.get(i).size());
      }
    } catch (ArithmeticException e) {
      throw new IllegalArgumentException(
          "Cartesian product too large; must have size at most Integer.MAX_VALUE");
    }
    this.axesSizeProduct = axesSizeProduct;
  }

  private int getAxisIndexForProductIndex(int index, int axis) {

      }
    }
  }

  /**
   * Used to delay string formatting until actually required, as it otherwise shows up in the test
   * execution profile when running an extremely large numbers of tests.
   */
  private String reportContext(List<E> expected) {
    return Platform.format(
        "expected collection %s; actual collection %s", expected, this.collection);
  }

	"Gone":                            http.StatusGone,
	"Length Required":                 http.StatusLengthRequired,
	"Precondition Failed":             http.StatusPreconditionFailed,
	"Request Entity Too Large":        http.StatusRequestEntityTooLarge,
	"Request URI Too Long":            http.StatusRequestURITooLong,
	"Unsupported Media Type":          http.StatusUnsupportedMediaType,

      }
    }
  }

  /**
   * Used to delay string formatting until actually required, as it otherwise shows up in the test
   * execution profile when running an extremely large numbers of tests.
   */
  private String reportContext(List<E> expected) {
    return Platform.format(
        "expected collection %s; actual collection %s", expected, this.collection);
  }

* What test cases should it have? What could go wrong?
* How will it roughly be implemented? We'll happily provide code pointers to save you time.

We may ask you to answer these questions directly in the GitHub issue or (for large changes) in a shared Google Doc.

pkg debug/pe, const IMAGE_FILE_DLL ideal-int
pkg debug/pe, const IMAGE_FILE_EXECUTABLE_IMAGE = 2
pkg debug/pe, const IMAGE_FILE_EXECUTABLE_IMAGE ideal-int
pkg debug/pe, const IMAGE_FILE_LARGE_ADDRESS_AWARE = 32
pkg debug/pe, const IMAGE_FILE_LARGE_ADDRESS_AWARE ideal-int
pkg debug/pe, const IMAGE_FILE_LINE_NUMS_STRIPPED = 4
pkg debug/pe, const IMAGE_FILE_LINE_NUMS_STRIPPED ideal-int
pkg debug/pe, const IMAGE_FILE_LOCAL_SYMS_STRIPPED = 8

	r := &netperfReader{eof: make(chan struct{})}
	r.buf = make([]byte, 128*humanize.KiByte)
	rand.Read(r.buf)

	connectionsPerPeer := 16

	if len(globalNotificationSys.peerClients) > 16 {
		// For a large cluster it's enough to have 1 connection per peer to saturate the network.
		connectionsPerPeer = 1
	}

	errStr := ""
	var wg sync.WaitGroup
	for index := range globalNotificationSys.peerClients {

 * 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"})
public class CharStreamsCopyBenchmark {
  enum CopyStrategy {
    OLD {
      @Override
      long copy(Readable from, Appendable to) throws IOException {

 * 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"})
public class CharStreamsCopyBenchmark {
  enum CopyStrategy {
    OLD {
      @Override
      long copy(Readable from, Appendable to) throws IOException {