///

## Run your code with your debugger { #run-your-code-with-your-debugger }

Because you are running the Uvicorn server directly from your code, you can call your Python program (your FastAPI application) directly from the debugger.

---

For example, in Visual Studio Code, you can:

* Go to the "Debug" panel.
* "Add configuration...".
* Select "Python"

## `Depends()`

Dependencies are handled mainly with the special function `Depends()` that takes a callable.

Here is the reference for it and its parameters.

You can import it directly from `fastapi`:

```python
from fastapi import Depends
```

::: fastapi.Depends

## `Security()`

      result.writeUtf8(string, pos, limit)
      return true
    }

    var pos = pos + 4 // 'xn--'.size.

    // We'd prefer to operate directly on `result` but it doesn't offer insertCodePoint(), only
    // appendCodePoint(). The Punycode algorithm processes code points in increasing code-point
    // order, not in increasing index order.

 * of the elapsed times reported. {@code Stopwatch} is still suitable for logging and metrics where
 * reasonably accurate values are sufficient. If the uncommon case that you need to maximize
 * accuracy, use {@code System.nanoTime()} directly instead.
 *
 * <p>Basic usage:
 *
 * {@snippet :
 * Stopwatch stopwatch = Stopwatch.createStarted();
 * doSomething();
 * stopwatch.stop(); // optional
 *
 * Duration duration = stopwatch.elapsed();
 *

    private NdrBuffer mockNdrBuffer;

    @Mock
    private NdrBuffer mockDeferredBuffer;

    @BeforeEach
    void setUp() {
        MockitoAnnotations.openMocks(this);
        // Set the deferred field directly since it's a public field, not a method
        mockNdrBuffer.deferred = mockDeferredBuffer;
        lenient().when(mockNdrBuffer.derive(anyInt())).thenReturn(mockDeferredBuffer);

    }

    // RAG workflow delegation methods

    /**
     * Detects the intent of a user message using the configured LLM client.
     *
     * @param userMessage the user's message
     * @return the detected intent with extracted keywords
     * @throws LlmException if LLM is not available
     */
    public IntentDetectionResult detectIntent(final String userMessage) {
        if (logger.isDebugEnabled()) {

   * Finalizer to stop.
   *
   * If this library is loaded in the system class loader, FinalizableReferenceQueue can load
   * Finalizer directly with no problems.
   *
   * If this library is loaded in an application class loader, it's important that Finalizer not
   * have a strong reference back to the class loader. Otherwise, you could have a graph like this:
   *

package jcifs.internal.smb2.rdma;

import jcifs.internal.util.SMBUtil;

/**
 * SMB Direct Negotiate Response message.
 *
 * As per MS-SMBD 2.2.2 - SMB_DIRECT_NEGOTIATE_RESPONSE
 * This message is sent in response to negotiate SMB Direct protocol parameters.
 */
public class SmbDirectNegotiateResponse {

    // Protocol constants
    /** SMB Direct negotiate response message type */
    public static final int NEGOTIATE_RESPONSE = 0x02;

 * governing permissions and limitations under the License.
 */
package org.codelibs.fess.llm;

/**
 * Represents the result of intent detection from user input.
 * Contains the detected intent type, Fess query, and other metadata.
 */
public class IntentDetectionResult {

    private final ChatIntent intent;
    private final String query;
    private final String documentUrl;

try:
    import orjson
except ImportError:  # pragma: nocover
    orjson = None  # type: ignore


@deprecated(
    "UJSONResponse is deprecated, FastAPI now serializes data directly to JSON "
    "bytes via Pydantic when a return type or response model is set, which is "
    "faster and doesn't need a custom response class. Read more in the FastAPI "