input: []byte(`id,time,num,num2,text
1,2010-01-01T,7867786,4565.908123
2,2017-01-02T03:04Z,-5, 0.765111,Some some
`),
			// Since we are doing offset, no headers are used.
			wantResult: `{"id":"1","time":"2010-01-01T","num":"7867786","num2":"4565.908123"}
{"id":"2","time":"2017-01-02T03:04Z","num":"-5","num2":" 0.765111","text":"Some some"}`,
			wantErr: false,
			requestXML: []byte(`<?xml version="1.0" encoding="UTF-8"?>

        assertTrue(options.directory().isPresent(), "--directory option should be present");
        assertEquals("/some/path", options.directory().get(), "--directory should be /some/path");
    }

    @Test
    void testGoalsParsing() throws ParseException {
        String[] args = {"apply", "--plugins"};
        CommonsCliUpgradeOptions options = CommonsCliUpgradeOptions.parse(args);

   * {@code node}'s outgoing edges in the direction (if any) of the edge.
   *
   * <p>This is <i>not</i> the same as "all nodes reachable from {@code node} by following outgoing
   * edges". For that functionality, see {@link Graphs#reachableNodes(Graph, Object)}.
   *
   * <p>Some algorithms that operate on a {@code SuccessorsFunction} may produce undesired results

Let's start with an example **use case** and then see how to solve it with this.

Let's imagine that you have some **machine learning models** that you want to use to handle requests. 🤖

The same models are shared among requests, so, it's not one model per request, or one per user or something similar.

Let's imagine that loading the model can **take quite some time**, because it has to read a lot of **data from disk**. So you don't want to do it for every request.

        email_file.write(content)


@app.post("/send-notification/{email}")
async def send_notification(email: str, background_tasks: BackgroundTasks):
    background_tasks.add_task(write_notification, email, message="some notification")

</div>

你可以發送如下的請求：

```json
{
    "description": "Some data",
    "data": "aGVsbG8="
}
```

/// tip

`aGVsbG8=` 是 `hello` 的 base64 編碼。

///

接著 Pydantic 會將該 base64 字串解碼，並在模型的 `data` 欄位中提供原始位元組。

你會收到類似以下的回應：

```json
{
  "description": "Some data",
  "content": "hello"
}
```

## Pydantic `bytes` 用於輸出資料 { #pydantic-bytes-for-output-data }

</div>

你可以发送如下请求：

```json
{
    "description": "Some data",
    "data": "aGVsbG8="
}
```

/// tip | 提示

`aGVsbG8=` 是 `hello` 的 base64 编码。

///

随后 Pydantic 会解码该 base64 字符串，并在模型的 `data` 字段中提供原始字节。

你将会收到类似的响应：

```json
{
  "description": "Some data",
  "content": "hello"
}
```

 * Other contributors include Andrew Wright, Jeffrey Hayes,
 * Pat Fisher, Mike Judd.
 */

/*
 * Source:
 * http://gee.cs.oswego.edu/cgi-bin/viewcvs.cgi/jsr166/src/test/tck/JSR166TestCase.java?revision=1.90
 * (We have made some trivial local modifications (commented out
 * uncompilable code).)
 */

package com.google.common.util.concurrent;

import static com.google.common.testing.SerializableTester.reserialize;

# nvidia driver to be installed.
nvidia-driver-580-open
# TODO(b/445248346): The Docker image shouldn't have cuda-compat installed.
# However, hermetic CUDA forward-compatibility mode is still missing some
# libraries.

main channel of communication for contributors.  
There are some guidelines which will make applying PRs easier for us:
+ Create a topic branch from where you want to base your work (this is usually the master branch).
  Push your changes to a topic branch in your fork of the repository.
+ Make commits of logical units.
+ Respect the original code style: by using the same [codestyle][code-style],