func (e ErasureInfo) ShardSize() int64 {
	return ceilFrac(e.BlockSize, int64(e.DataBlocks))
}

// IsValid - tells if erasure info fields are valid.
func (fi FileInfo) IsValid() bool {
	if fi.Deleted {
		// Delete marker has no data, no need to check
		// for erasure coding information
		return true
	}
	dataBlocks := fi.Erasure.DataBlocks
	parityBlocks := fi.Erasure.ParityBlocks
	correctIndexes := (fi.Erasure.Index > 0 &&

	if !immediatePurge && s.diskAlmostFilled() {
		immediatePurge = true
	}

	// immediately purge the target
	if immediatePurge {
		for _, target := range []string{
			targetPath,
			targetPath2,
		} {
			if target == "" {
				continue
			}
			select {
			case s.immediatePurge <- target:
			default:
				// Too much back pressure, we will perform the delete

		// Part with size larger than 5 MiB.
		{bucketName, objectName, uploadIDs[0], 5, string(validPart), validPartMD5, int64(len(validPart))},
		{bucketName, objectName, uploadIDs[0], 6, string(validPart), validPartMD5, int64(len(validPart))},

		// Part with size larger than 5 MiB.
		// Parts uploaded for anonymous/unsigned API handler test.

	}

	result.UserDefined = cloneMSS(fi.Metadata)
	return result, nil
}

// ListObjectParts - lists all previously uploaded parts for a given
// object and uploadID.  Takes additional input of part-number-marker
// to indicate where the listing should begin from.
//
// Implements S3 compatible ListObjectParts API. The resulting
// ListPartsInfo structure is marshaled directly into XML and
// replied back to the client.

    return new HashMap<>(capacity(expectedSize));
  }

  /**
   * Returns a capacity that is sufficient to keep the map from being resized as long as it grows no
   * larger than expectedSize and the load factor is ≥ its default (0.75).
   */
  static int capacity(int expectedSize) {
    if (expectedSize < 3) {
      checkNonnegative(expectedSize, "expectedSize");

*   <IF RELEASE CONTAINS MULTIPLE FEATURES FROM SAME AREA, GROUP THEM TOGETHER>

*   Add `is_cpu_target_available`, which indicates whether or not TensorFlow was
    built with support for a given CPU target. This can be useful for skipping
    target-specific tests if a target is not supported.

*   `tf.data`
    * Support `data.experimental.distribued_save`. `distribued_save` uses
      tf.data service

    * <a href="https://fastapi.tiangolo.com/tutorial/bigger-applications/" class="external-link" target="_blank">Tutorial - Bigger Applications</a>.