* thought of as overkill; however, it's difficult to determine which proper subset of this massive
 * set would be sufficient to expose any possible bug. Brute force is simpler.
 *
 * <p>To use this class the concrete subclass must implement the {@link
 * IteratorTester#newTargetIterator()} method. This is because it's impossible to test an Iterator

          public <T extends @Nullable Object> T[] toArray(T[] array) {
            T[] result = super.toArray(array);
            if (size() < result.length) {
              // It works around a GWT bug where elements after last is not
              // properly null'ed.
              @Nullable Object[] unsoundlyCovariantArray = result;
              unsoundlyCovariantArray[size()] = null;
            }

    }

    @Override
    public @Nullable Object[] toArray() {
      return standardToArray();
    }

    @Override
    @SuppressWarnings("nullness") // bug in our checker's handling of toArray signatures
    public <T extends @Nullable Object> T[] toArray(T[] array) {
      return standardToArray(array);
    }

    @Override
    public String toString() {

    ImmutableTypeToInstanceMap<Iterable<?>[]> map =
        ImmutableTypeToInstanceMap.<Iterable<?>[]>builder().put(type, array).build();
    assertEquals(1, map.size());
    // Redundant cast works around a javac bug.
    assertThat((Iterable<?>[]) map.getInstance(type)).asList().containsExactly(array[0]);
  }

  public void testWildcardType() {
    TypeToken<ImmutableList<?>> type = new TypeToken<ImmutableList<?>>() {};

    connectTimeout: Int,
  ) {
    try {
      socket.connect(address, connectTimeout)
    } catch (e: ClassCastException) {
      // On android 8.0, socket.connect throws a ClassCastException due to a bug
      // see https://issuetracker.google.com/issues/63649622
      if (Build.VERSION.SDK_INT == 26) {
        throw IOException("Exception in connect", e)
      } else {
        throw e
      }
    }
  }

## 可用版本

你可以在[发行说明](../release-notes.md){.internal-link target=_blank}中查看可用版本（例如查看当前最新版本）。

## 关于版本

遵循语义版本控制约定，任何低于`1.0.0`的版本都可能会添加 breaking changes。

FastAPI 还遵循这样的约定：任何`PATCH`版本更改都是为了bug修复和non-breaking changes。

/// tip

"PATCH"是最后一个数字，例如，在`0.2.3`中，PATCH版本是`3`。

///

因此，你应该能够固定到如下版本：

```txt
fastapi>=0.45.0,<0.46.0
```

"MINOR"版本中会添加breaking changes和新功能。

/// tip

				return
			}
		case "ic":
			z.InitCapacity, err = dc.ReadUint64()
			if err != nil {
				err = msgp.WrapError(err, "InitCapacity")
				return
			}
		case "bus":
			var zb0002 uint32
			zb0002, err = dc.ReadArrayHeader()
			if err != nil {
				err = msgp.WrapError(err, "Buckets")
				return
			}
			if cap(z.Buckets) >= int(zb0002) {

    "EEE dd-MMM-yyyy HH-mm-ss z",
    "EEE dd-MMM-yy HH:mm:ss z",
    "EEE dd MMM yy HH:mm:ss z",
    "EEE,dd-MMM-yy HH:mm:ss z",
    "EEE,dd-MMM-yyyy HH:mm:ss z",
    "EEE, dd-MM-yyyy HH:mm:ss z",
    // RI bug 6641315 claims a cookie of this format was once served by www.yahoo.com:
    "EEE MMM d yyyy HH:mm:ss z",
  )

private val BROWSER_COMPATIBLE_DATE_FORMATS =

    }
    return dummy;
  }

  /**
   * Similar to the above, but keeps a boolean flag rather than checking for the string accumulated
   * so far being empty. As a result, it does not have the above-mentioned bug.
   */
  @Benchmark
  int booleanIfFirst(int reps) {
    int dummy = 0;
    for (int i = 0; i < reps; i++) {
      StringBuilder sb = new StringBuilder();
      boolean append = false;

 * thought of as overkill; however, it's difficult to determine which proper subset of this massive
 * set would be sufficient to expose any possible bug. Brute force is simpler.
 *
 * <p>To use this class the concrete subclass must implement the {@link
 * IteratorTester#newTargetIterator()} method. This is because it's impossible to test an Iterator