## Review PRs

* If a PR doesn't explain what it does or why, if it seems like it could be useful, ask for more information. Otherwise, feel free to close it.

* If a PR seems to be spam, meaningless, only to change statistics (to appear as "contributor") or similar, you can simply mark it as `invalid`, and it will be automatically closed.

Most of these ideas would be more or less **independent**, and in most cases you should only need to study them if they apply directly to **your project**.

If something seems interesting and useful to your project, go ahead and check it, but otherwise, you might probably just skip them.

/// tip

     * Don't bother testing. All non-primitive parameters are used only to construct error messages.
     * We never want to pass null for them, so we haven't annotated them to say that null is
     * allowed. But at the same time, it seems wasteful to bother inserting the checkNotNull calls
     * that NullPointerTester wants.
     *
     * (This empty method disables the automatic null testing provided by PackageSanityTests.)
     */
  }

# History, Design and Future { #history-design-and-future }

Some time ago, [a **FastAPI** user asked](https://github.com/fastapi/fastapi/issues/3#issuecomment-454956920):

> What’s the history of this project? It seems to have come from nowhere to awesome in a few weeks [...]

Here's a little bit of that history.

## Alternatives { #alternatives }

};

// Returns the number of elements in the gradient tensor.
int64_t TapeVSpace::NumElements(AbstractTensorHandle* tensor) const {
  // TODO(srbs): It seems like this is used only for performance optimization
  // and not for correctness. The only downside of keeping this 1 seems to be
  // that the gradient accumulation is unbounded and we will never
  // aggressively aggregate accumulated gradients to recover memory.
  // Revisit and fix.

        "InjectOnConstructorOfAbstractClass", // We use abstract injection as a pattern
        "InlineMeSuggester", // Only suppression seems to actually "fix" this, so make it global
        "JavaUtilDate", // We are fine with using Date
        "JavaxInjectOnAbstractMethod", // We use abstract injection as a pattern

The same example would then look like:

{* ../../docs_src/dependencies/tutorial004_an_py310.py hl[19] *}

...and **FastAPI** will know what to do.

/// tip

If that seems more confusing than helpful, disregard it, you don't *need* it.

It is just a shortcut. Because **FastAPI** cares about helping you minimize code repetition.

        )

        if pr.state == "open" and awaiting_label in label_strs:
            logging.info(
                f"This PR seems to be a language translation and awaiting reviews: #{pr.number}"
            )
            if already_notified_comment:
                logging.info(

      ).build()
  }

  private fun buildServerSslSocketFactory(): SSLSocketFactory {
    // The test uses JDK default SSL Context instead of the Platform provided one
    // as Conscrypt seems to have some differences, we only want to test client side here.
    return try {
      val keyManager =
        newKeyManager(
          null,
          serverCert,
          serverIntermediateCa.certificate,

    // should be reading the byte at index 10.
    // We could use a custom InputStream instead to make the read start at index 10, but since this
    // is a racy situation anyway, this behavior seems reasonable.
    assertEquals(-1, in.read());
  }

  private static class AppendableByteSource extends ByteSource {
    private byte[] bytes;

    AppendableByteSource(byte[] initialBytes) {