exitReader: make(chan struct{}),
	}
	r.onReaderExit = func() {
		close(r.decoded)
		readCloser.Close()
		for range r.input {
			// Read until EOF trickles through.
			// Otherwise, we risk the decoder hanging.
		}
		r.readerWg.Done()
	}

	// We cannot reuse as we are sending parsed objects elsewhere.
	simdjson.ParseNDStream(readCloser, r.input, nil)
	r.readerWg.Add(1)
	go r.startReader()
	return &r
}

```Python hl_lines="10"
{!> ../../docs_src/query_params_str_validations/tutorial002_an.py!}
```

////

Notice that the default value is still `None`, so the parameter is still optional.

But now, having `Query(max_length=50)` inside of `Annotated`, we are telling FastAPI that we want it to have **additional validation** for this value, we want it to have maximum 50 characters. 😎

/// tip

    @SuppressWarnings({
      "unused",
      /*
       * Since reflection can't tell the difference between <T> and <T extends Object>, it doesn't
       * make a ton of sense to have a separate tests for each. But having tests for each doesn't
       * really hurt anything, and maybe it will serve a purpose in a future in which Java has a
       * built-in nullness feature?
       */
      "ExtendsObject",
    })

    @SuppressWarnings({
      "unused",
      /*
       * Since reflection can't tell the difference between <T> and <T extends Object>, it doesn't
       * make a ton of sense to have a separate tests for each. But having tests for each doesn't
       * really hurt anything, and maybe it will serve a purpose in a future in which Java has a
       * built-in nullness feature?
       */
      "ExtendsObject",
    })

All the same process that applied for path parameters also applies for query parameters:

* Editor support (obviously)
* Data <abbr title="converting the string that comes from an HTTP request into Python data">"parsing"</abbr>
* Data validation
* Automatic documentation

## Defaults

As query parameters are not a fixed part of a path, they can be optional and can have default values.

 * of the <i>next</i> request. I.e., if an expensive task arrives at an idle RateLimiter, it will be
 * granted immediately, but it is the <i>next</i> request that will experience extra throttling,
 * thus paying for the cost of the expensive task.
 *
 * @author Dimitris Andreou
 * @since 13.0
 */
// TODO(user): switch to nano precision. A natural unit of cost is "bytes", and a micro precision

   * this initiates service shutdown and returns immediately. If the service is {@linkplain
   * State#NEW new}, it is {@linkplain State#TERMINATED terminated} without having been started nor
   * stopped. If the service has already been stopped, this method returns immediately without
   * taking action.
   *
   * @return this
   * @since 15.0
   */
  @CanIgnoreReturnValue
  Service stopAsync();

  /**

Readme.md would be the same file on Windows. In this test, you would get a
warning for "readme.md" because it makes everything lowercase. There are
repeats of these filename(s) with different casing:
EOF
    find . | tr '[A-Z]' '[a-z]' | sort | uniq -d | tee $BATS_FILE_TMPDIR/repeats
    [[ ! -s $BATS_FILE_TMPDIR/repeats ]]
}

# It's unclear why, but running this on //tensorflow/... is faster than running

## Conflict Resolution

Conflicts in an open source project can take many forms, from someone having a bad day and using harsh and hurtful language in the issue queue, to more serious instances such as sexist/racist statements or threats of violence, and everything in between.

 *       }
 *       return endOfData();
 *     }
 *   };
 * }
 * }</pre>
 *
 * <p>This class supports iterators that include null elements.
 *
 * @author Kevin Bourrillion
 * @since 2.0
 */
// When making changes to this class, please also update the copy at
// com.google.common.base.AbstractIterator
@GwtCompatible
@ElementTypesAreNonnullByDefault