} else {
        // if our write lost the race, it must have lost to a nonzero value, so we can stop
        return countMap.putIfAbsent(element, new AtomicInteger(newCount)) == null;
      }
    }
    int oldValue = existingCounter.get();
    if (oldValue == expectedOldCount) {
      if (oldValue == 0) {
        if (newCount == 0) {
          // Just observed a 0; try to remove the entry to clean up the map

   *
   * <p>This method is useful primarily when we make a reflective call to a method with no {@code
   * throws} clause: Java forces us to handle an arbitrary {@link Throwable} from that method,
   * rather than just the {@link RuntimeException} or {@link Error} that should be possible. (And in
   * fact the static type of {@link Throwable} is occasionally justified even for a method with no

this is just a marker file....

The important thing to remember is that **the first *type parameter*** you pass to `Annotated` is the **actual type**. The rest, is just metadata for other tools.

For now, you just need to know that `Annotated` exists, and that it's standard Python. 😎

Later you will see how **powerful** it can be.

/// tip

That doesn't add any extra security to your API, the *path operations* will still be available where they are.

If there's a security flaw in your code, it will still exist.

You can create production applications with **FastAPI** right now (and you have probably been doing it for some time), you just have to make sure that you use a version that works correctly with the rest of your code.

## Pin your `fastapi` version { #pin-your-fastapi-version }

You could create a first dependency ("dependable") like:

{* ../../docs_src/dependencies/tutorial005_an_py310.py hl[8:9] *}

It declares an optional query parameter `q` as a `str`, and then it just returns it.

This is quite simple (not very useful), but will help us focus on how the sub-dependencies work.

## Second dependency, "dependable" and "dependant" { #second-dependency-dependable-and-dependant }

from typing import Annotated


def say_hello(name: Annotated[str, "this is just metadata"]) -> str:

         * unique within a single POM. Upon multiple declarations, 2.x just kept the last one but retained the order of
         * the first occurrence. So when we're in lenient/compat mode, we have to deal with such broken POMs and mimic
         * the way 2.x works. When we're in strict mode, the removal of duplicates just saves other merging steps from
         * aftereffects and bogus error messages.
         */

    full_name = user_dict["full_name"],
    hashed_password = hashed_password,
)
```

/// warning

The supporting additional functions `fake_password_hasher` and `fake_save_user` are just to demo a possible flow of the data, but they of course are not providing any real security.

///

## Reduce duplication { #reduce-duplication }

Reducing code duplication is one of the core ideas in **FastAPI**.