* {@linkplain Cache#invalidate invalidations}.
   */
  public long evictionCount() {
    return evictionCount;
  }

  /**
   * Returns a new {@code CacheStats} representing the difference between this {@code CacheStats}
   * and {@code other}. Negative values, which aren't supported by {@code CacheStats} will be
   * rounded up to zero.
   */
  public CacheStats minus(CacheStats other) {

   * {@linkplain Cache#invalidate invalidations}.
   */
  public long evictionCount() {
    return evictionCount;
  }

  /**
   * Returns a new {@code CacheStats} representing the difference between this {@code CacheStats}
   * and {@code other}. Negative values, which aren't supported by {@code CacheStats} will be
   * rounded up to zero.
   */
  public CacheStats minus(CacheStats other) {

    collected using form data (instead of JSON) and that it should have the specific
    fields `username` and `password`.

    All the initialization parameters are extracted from the request.

    The only difference between `OAuth2PasswordRequestFormStrict` and
    `OAuth2PasswordRequestForm` is that `OAuth2PasswordRequestFormStrict` requires the
    client to send the form field `grant_type` with the value `"password"`, which

					common = i
					break
				}
			}
			a, b = a[common:], b[common:]
			if len(a) > 64 {
				a = a[:64]
			}
			if len(b) > 64 {
				b = b[:64]
			}
			t.Errorf("after %d common bytes, difference \nread: %x\nwrote:%x", common, a, b)
		}
		t.Fatalf("expect read %d bytes but got %d", wroteBytes, readBytes)
	}
	if readHash.Sum32() != wroteHash.Sum32() {

 * resulting XML document is parsed, the parser API will return this text as the original literal
 * string {@code "Foo<Bar>"}.
 *
 * <p><b>Note:</b> This class is similar to {@link CharEscaper} but with one very important
 * difference. A CharEscaper can only process Java <a
 * href="http://en.wikipedia.org/wiki/UTF-16">UTF16</a> characters in isolation and may not cope
 * when it encounters surrogate pairs. This class facilitates the correct escaping of all Unicode

Vou lhe contar um pouco mais sobre esses **conceitos** aqui, e espero que isso lhe dê a **intuição** necessária para decidir como implantar sua API em ambientes muito diferentes, possivelmente até mesmo em **futuros** ambientes que ainda não existem.

Ao considerar esses conceitos, você será capaz de **avaliar e projetar** a melhor maneira de implantar **suas próprias APIs**.

**Concurrencia** y **paralelismo** ambos se relacionan con "cosas diferentes que suceden más o menos al mismo tiempo".

Pero los detalles entre *concurrencia* y *paralelismo* son bastante diferentes.

Para entender las diferencias, imagina la siguiente historia sobre hamburguesas:

### Hamburguesas Concurrentes

Por exemplo, vamos supor que você possua 4 endpoints na sua API (*operações de rota*):

* `/items/public/`
* `/items/private/`
* `/users/{user_id}/activate`
* `/items/pro/`

Você poderia adicionar diferentes requisitos de permissão para cada um deles utilizando apenas dependências e sub-dependências:

```mermaid
graph TB

current_user(["current_user"])
active_user(["active_user"])
admin_user(["admin_user"])

    // array of indexes to be calculated is modified between the calls to indexes and compute: since
    // the contract is that it is snapshotted, this shouldn't make any difference to the result.
    int[] indexes = {0, 10, 5, 1, 8, 10};
    ScaleAndIndexes intermediate = Quantiles.scale(10).indexes(indexes);
    indexes[0] = 3;
    assertThat(intermediate.compute(SIXTEEN_SQUARES_DOUBLES))

   * "an unused RateLimiter" is also defined by that notion: when we observe that the
   * "expected arrival time of the next request" is actually in the past, then the difference (now -
   * past) is the amount of time that the RateLimiter was formally unused, and it is that amount of
   * time which we translate to storedPermits. (We increase storedPermits with the amount of permits