* to avoid the extra class for the lambda (and maybe more for memoizingSupplier itself) and the
     * indirection.
     *
     * One thing to *avoid* is a change to make each Logger user use memoizingSupplier directly:
     * That may introduce an extra class for each lambda (currently a dozen).
     */
    Logger local = logger;
    if (local != null) {

  @Override
  @SuppressWarnings("nullness") // should be safe, but not sure if we can avoid the warning
  public <S extends @Nullable T> Ordering<S> reverse() {
    // ordering.reverse() might be optimized, so let it do its thing
    return ordering.<T>reverse().<@NonNull S>nullsLast();
  }

  @SuppressWarnings("unchecked") // still need the right way to explain this
  @Override
  public <S extends @Nullable T> Ordering<@Nullable S> nullsFirst() {

/**
 * Variant of {@link SerializableTester} that does not require the reserialized object's class to be
 * identical to the original.
 *
 * @author Chris Povirk
 */
/*
 * The whole thing is really @GwtIncompatible, but GwtJUnitConvertedTestModule doesn't have a
 * parameter for non-GWT, non-test files, and it didn't seem worth adding one for this unusual case.
 */
@GwtCompatible
@NullUnmarked

    }

    /**
     * In order to reuse the glob matcher from the filesystem, we need
     * to create Path instances.  Those are only used with the toString method.
     * This hack works because the only system-dependent thing is the path
     * separator which should not be part of the groupId or artifactId.
     */
    private static Path createPathProxy(String value) {
        return (Path) Proxy.newProxyInstance(

## Contributing

[You can help to deliver a better GORM, check out things you can do](https://gorm.io/contribute.html)

## Contributors

[Thank you](https://github.com/go-gorm/gorm/graphs/contributors) for contributing to the GORM framework!

## License

© Jinzhu, 2013~time.Now

    assertBasicUrlEscaper(e);

    /*
     * Specified as safe by RFC 2396 but not by java.net.URLEncoder. These tests will start failing
     * when the escaper is made compliant with RFC 2396, but that's a good thing (just change them
     * to assertUnescaped).
     */
    assertEscaping(e, "%21", '!');
    assertEscaping(e, "%28", '(');
    assertEscaping(e, "%29", ')');
    assertEscaping(e, "%7E", '~');

name: Bug Report
description: Something is not working as expected
labels: ["type=defect"]
body:
  - type: markdown
    attributes:
      value: >
        Thank you for filing a bug report. Please help us identify and resolve the bug by filling
        out the following fields. Before we begin, please make sure that the bug is still present in
        the [latest](https://github.com/google/guava/releases/latest) version of Guava available.

        // tricky thing: align what JLine3 detected and Maven thinks:
        // if embedded, we default to context.coloredOutput=false unless overridden (see above)
        // if not embedded, JLine3 may detect redirection and will create dumb terminal.
        // To align Maven with outcomes, we set here color enabled based on these premises.
        // Note: Maven3 suffers from similar thing: if you do `mvn3 foo > log.txt`, the output will

import okio.ByteString.Companion.encodeUtf8
import okio.IOException
import okio.Source
import okio.buffer

class ResponseBodyTest {
  @Test
  fun sourceEmpty() {
    val mediaType = if (null == null) null else "any/thing; charset=${null}".toMediaType()
    val body = "".decodeHex().toResponseBody(mediaType)
    val source = body.source()
    assertThat(source.exhausted()).isTrue()
    assertThat(source.readUtf8()).isEqualTo("")
  }

The main logic with ADRs is to describe (architectural) decisions made:

* To provide best practices and solutions we (as the *build tool* team) want to promote.
* To avoid asking the same thing multiple times during code review.
* To explain *rejected solutions*, for now, and future development, in case they are proposed again.

ADRs can be written by any team.