This decoupling of parts, and being a "microframework" that could be extended to cover exactly what is needed was a key feature that I wanted to keep.

Given the simplicity of Flask, it seemed like a good match for building APIs. The next thing to find was a "Django REST Framework" for Flask.

    sampleInstances.putAll(checkNotNull(type), checkNotNull(instances));
  }

  /**
   * Returns a fresh instance for {@code type} if possible. The returned instance could be:
   *
   * <ul>
   *   <li>exactly of the given type, including generic type parameters, such as {@code
   *       ImmutableList<String>};
   *   <li>of the raw type;
   *   <li>null if no value can be generated.
   * </ul>
   */

    return (c >= 'A') && (c <= 'Z');
  }

  /**
   * Truncates the given character sequence to the given maximum length. If the length of the
   * sequence is greater than {@code maxLength}, the returned string will be exactly {@code
   * maxLength} chars in length and will end with the given {@code truncationIndicator}. Otherwise,
   * the sequence will be returned as a string with no changes to the content.
   *
   * <p>Examples:
   *

            assertTrue(result.errorCount() > 0, "Downgrade should result in errors");
            assertFalse(result.success(), "Downgrade should not be successful");
            assertEquals(1, result.errorCount(), "Should have exactly one error");
        }

        @Test
        @DisplayName("should succeed when upgrading from 4.0.0 to 4.1.0")
        void shouldSucceedWhenUpgrading() throws Exception {
            String pomXml = """

 * <p>Usage example:
 *
 * {@snippet :
 * MinMaxPriorityQueue<User> users = MinMaxPriorityQueue.orderedBy(userComparator)
 *     .maximumSize(1000)
 *     .create();
 * }
 *
 * <p>As a {@link Queue} it functions exactly as a {@link PriorityQueue}: its head element -- the
 * implicit target of the methods {@link #peek()}, {@link #poll()} and {@link #remove()} -- is

      // Many spliterators will have trySplits that are SUBSIZED even if they are not themselves
      // SUBSIZED.
      if (spliterator.hasCharacteristics(Spliterator.SUBSIZED)) {
        // we can drill down to exactly the smallest nonempty spliterator
        while (true) {
          Spliterator<T> prefix = spliterator.trySplit();
          if (prefix == null || prefix.getExactSizeIfKnown() == 0) {
            break;

     * <pre>maven.relocations.entries = org.foo:*:*>, \\<br/>    org.here:*:*>org.there:*:*, \\<br/>    javax.inject:javax.inject:1>>jakarta.inject:jakarta.inject:1.0.5</pre>
     * means: 3 entries, ban <code>org.foo group</code> (exactly, so <code>org.foo.bar</code> is allowed),
     * relocate <code>org.here</code> to <code>org.there</code> and finally globally relocate (see <code>&gt;&gt;</code> above)

   * ByteArrayDataOutput} is created, the contract for {@link ByteArrayDataOutput#toByteArray} will
   * not be honored (the bytes returned in the byte array may not be exactly what was written via
   * calls to {@code ByteArrayDataOutput}).
   *
   * @since 17.0
   */
  @J2ktIncompatible
  public static ByteArrayDataOutput newDataOutput(ByteArrayOutputStream byteArrayOutputStream) {

        name.srcHashCode = svr != null ? svr.hashCode() : 0;
        addr = getCachedAddress(name);

        if (addr == null) {
            addr = (NbtAddress) checkLookupTable(name);
            /* This is almost exactly like InetAddress.java. See the
             * comments there for a description of how the LOOKUP_TABLE prevents
             * redundant queries from going out on the wire.
             */
            if (addr == null) {

	// for ListObjects, ListObjectsV2
	requestedVersions int

	bucket string // Name of the bucket. Used for generating cached fileinfo.
	strict bool   // Versions must match exactly, including all metadata.

	// Reusable slice for resolution
	candidates [][]xlMetaV2ShallowVersion
}

// resolve multiple entries.
// entries are resolved by majority, then if tied by mod-time and versions.