it may already exist, even as a closed issue.
-->

<!--
Describe the feature.

Please give us as much context as possible about the feature. For example,
you could include a story about a time when you wanted to use the feature,
and also tell us what you had to do instead. The last part is helpful
because it gives us an idea of how much harder your life is without the
feature.

	}{Version: healthInfo.Version}

	enc := json.NewEncoder(gzWriter)
	if e := enc.Encode(header); e != nil {
		internalLogIf(ctx, fmt.Errorf("Could not encode health info header: %w", e))
		return nil
	}

	if e := enc.Encode(healthInfo); e != nil {
		internalLogIf(ctx, fmt.Errorf("Could not encode health info: %w", e))
		return nil
	}

	gzWriter.Flush()
	gzWriter.Close()

	return b.Bytes()

</blockquote>

## Investigation { #investigation }

By using all the previous alternatives I had the chance to learn from all of them, take ideas, and combine them in the best way I could find for myself and the teams of developers I have worked with.

For example, it was clear that ideally it should be based on standard Python type hints.

Also, the best approach was to use already existing standards.

```python
from app.main import app
```

That way the `fastapi` command will know where to find your app.

/// Note

You could also pass the path to the command, like:

```console
$ fastapi dev app/main.py
```

But you would have to remember to pass the correct path every time you call the `fastapi` command.

            /*
             * When Gradle walks the file tree, it will follow symbolic links. This means that if there is a symbolic link to a directory
             * in the source file tree, we could otherwise end up duplicating the entries below that directory in the resulting tar archive.
             * To avoid this, we track which symbolic links we have visited, and skip files that are children of symbolic links that we have

  //   system scheduling and as such we could either miss our deadline, or unpark() could be delayed
  //   so that it looks like we timed out even though we didn't. For comparison FutureTask respects
  //   completion preferably and AQS is non-deterministic (depends on where in the queue the waiter
  //   is). If we wanted to be strict about it, we could store the unpark() time in the Waiter node

    /*
     * In principle, a Multimap implementation could add e3 first before failing on the null. But
     * that seems unlikely enough to be worth complicating the test over, especially if there's any
     * chance that a permissive test could mask a bug.
     */
    expectUnchanged();
    // Be extra thorough in case internal state was corrupted by the expected null.

                () -> newModelResolver().resolveModel(parent),
                "Expected 'UnresolvableModelException' not thrown.");
        assertNotNull(e.getMessage());
        assertTrue(e.getMessage().contains("Could not find artifact ut.simple:artifact:pom:0 in repo"));
    }

    @Test
    public void testResolveParentThrowsUnresolvableModelExceptionWhenNoMatchingVersionFound() throws Exception {

    /**
     * Return the sAMAccountName of this SID unless it could not
     * be resolved in which case the numeric RID is returned. If this
     * SID is a domain SID, this method will return an empty String.
     *
     * @return the account name
     */
    String getAccountName();

    /**
     * Return the domain name of this SID unless it could not be

For example, you could want to **return a dictionary** or a database object, but **declare it as a Pydantic model**. This way the Pydantic model would do all the data documentation, validation, etc. for the object that you returned (e.g. a dictionary or database object).