deltaToCeiling++;
          }

          int diff = Long.compare(deltaToFloor, deltaToCeiling);
          if (diff < 0) { // closer to floor
            return roundFloorAsDouble;
          } else if (diff > 0) { // closer to ceiling
            return roundCeilingAsDouble;
          }
          // halfway between the representable values; do the half-whatever logic

	defer closeValidYAMLFile.Close()

	validFilenameJSON, closeValidJSONFile := createTestFile(t, validJSON)
	defer closeValidJSONFile.Close()

	invalidFilenameYAML, closeInvalidYAMLFile := createTestFile(t, invalidYAML)
	defer closeInvalidYAMLFile.Close()

	warningFilename, closeWarningFile := createTestFile(t, warningsYAML)
	defer closeWarningFile.Close()

    /**
     * Constructs a new iterator over all values for the specified key starting at the specified
     * index. This constructor is optimized so that it starts at either the head or the tail,
     * depending on which is closer to the specified index. This allows adds to the tail to be done
     * in constant time.
     *
     * @throws IndexOutOfBoundsException if index is invalid
     */

```
cat ${HADOOP_CONF_DIR}/core-site.xml | kv-pairify | grep "mapred"

mapred.maxthreads.generate.mapoutput=2 # Num threads to write map outputs
mapred.maxthreads.partition.closer=0 # Asynchronous map flushers
mapreduce.fileoutputcommitter.algorithm.version=2 # Use the latest committer version
mapreduce.job.reduce.slowstart.completedmaps=0.99 # 99% map, then reduce

Routes are declared in a single place, using functions declared in other places (instead of using decorators that can be placed right on top of the function that handles the endpoint). This is closer to how Django does it than to how Flask (and Starlette) does it. It separates in the code things that are relatively tightly coupled.

/// check | "Inspired **FastAPI** to"

   * {@code upper}. Assumes that {@code lower <= upper}. Correctly handles infinities (but not
   * {@code NaN}).
   */
  private static double interpolate(double lower, double upper, double remainder, double scale) {
    if (lower == NEGATIVE_INFINITY) {
      if (upper == POSITIVE_INFINITY) {
        // Return NaN when lower == NEGATIVE_INFINITY and upper == POSITIVE_INFINITY:

And as you can have parallelism and asynchronicity at the same time, you get higher performance than most of the tested NodeJS frameworks and on par with Go, which is a compiled language closer to C <a href="https://www.techempower.com/benchmarks/#section=data-r17&hw=ph&test=query&l=zijmkf-1" class="external-link" target="_blank">(all thanks to Starlette)</a>.

### Is concurrency better than parallelism?

		}
		w.Close()
		return errMoreData
	}

	// Only interested in flushing the size_t not mtime/atime
	if err = Fdatasync(w); err != nil {
		w.Close()
		return err
	}

	// Dealing with error returns from close() - 'man 2 close'
	//
	// A careful programmer will check the return value of close(), since it is quite possible that

)

// DrainBody close non nil response with any response Body.
// convenient wrapper to drain any remaining data on response body.
//
// Subsequently this allows golang http RoundTripper
// to reuse the same connection for future requests.
func DrainBody(respBody io.ReadCloser) {
	// Callers should close resp.Body when done reading from it.
	// If resp.Body is not closed, the Client's underlying RoundTripper

		// causing error.
		n, err := reader.Read(shard)
		w.Write(shard[:n])
		if err == nil {
			continue
		}
		if err == io.EOF {
			break
		}
		t.Fatal(err)
	}
	w.(io.Closer).Close()
	if err := storage.storage.bitrotVerify(context.Background(), pathJoin(path, volName, fileName), size, algo, nil, shardSize); err != nil {
		t.Fatal(err)
	}

	// 4) Streaming bitrot check on corrupted file