There's no waiting 🕙 anywhere, just a lot of work to be done, on multiple places of the house.

You could have turns as in the burgers example, first the living room, then the kitchen, but as you are not waiting 🕙 for anything, just cleaning and cleaning, the turns wouldn't affect anything.

It would take the same amount of time to finish with or without turns (concurrency) and you would have done the same amount of work.

/// info

The important point is that your app will be able to produce each line in turn, while the client consumes the previous lines.

///

/// note | Technical Details

            }
        }

        if (firstIncludedTurn < turns.size()) {
            for (int t = firstIncludedTurn; t < turns.size(); t++) {
                final int[] turn = turns.get(t);
                for (int i = turn[0]; i < turn[1]; i++) {
                    request.addMessage(history.get(i));
                }
            }

    }

    @Test
    public void test_addHistoryWithBudget_turnBasedPacking_contiguousRecency() {
        // 3 turns: turn0(Q1+A1=4), turn1(big=200), turn2(Q3+A3=4)
        // Budget=20: turn2 fits (4, remaining=16), turn1 doesn't fit (200>16), stop
        // Even though turn0 (4 chars) would fit individually, contiguity is maintained
        final List<LlmMessage> history = new ArrayList<>();

    }

    private void cleanUpIssueLinks(Document document) {
        String rewritten = document.body().html();
        // Turn Gradle Jira issue numbers into issue links
        rewritten = rewritten.replaceAll("GRADLE-\\d+", "<a href=\"https://issues.gradle.org/browse/$0\">$0</a>");
        // Turn Gradle Github issue numbers into issue links

            project.plugins.withType<JavaLibraryPlugin> {
                // Kotlin-only projects do not hit this, so they don't have nullaway attributes on the outgoing variants.
                // Java project can in turn depend on Kotlin projects even if they have nullaway enabled.
                NullawayAttributes.addToConfiguration(configurations.named(apiElementsConfigurationName), nullawayAttributeValue)

# Testing { #testing }

Thanks to [Starlette](https://www.starlette.dev/testclient/), testing **FastAPI** applications is easy and enjoyable.

It is based on [HTTPX](https://www.python-httpx.org), which in turn is designed based on Requests, so it's very familiar and intuitive.

With it, you can use [pytest](https://docs.pytest.org/) directly with **FastAPI**.

## Using `TestClient` { #using-testclient }

/// info

## A `StreamingResponse` with `yield` { #a-streamingresponse-with-yield }

If you declare a `response_class=StreamingResponse` in your *path operation function*, you can use `yield` to send each chunk of data in turn.

{* ../../docs_src/stream_data/tutorial001_py310.py ln[1:23] hl[20,23] *}

FastAPI will give each chunk of data to the `StreamingResponse` as is, it won't try to convert it to JSON or anything similar.

You can define arbitrarily deeply nested models:

{* ../../docs_src/body_nested_models/tutorial007_py310.py hl[7,12,18,21,25] *}

/// info

Notice how `Offer` has a list of `Item`s, which in turn have an optional list of `Image`s

///

## Bodies of pure lists { #bodies-of-pure-lists }

   *       for every closeable object this step creates in order to capture it for later closing.
   *   <li>Return a {@code ClosingFuture}. To turn a {@link ListenableFuture} into a {@code
   *       ClosingFuture} call {@link #from(ListenableFuture)}.
   *   <li>In case this step doesn't create new closeables, you can adapt an API that returns a