怎样在PHP项目中实现GPT集成？

本文目录导读：

1. 使用OpenAI官方API（最推荐）
2. 使用cURL直接调用API
3. 安全性最佳实践
4. Laravel项目集成示例
5. 配置文件示例
6. 环境变量配置
7. 注意事项

在PHP项目中集成GPT（如OpenAI的GPT-3.5或GPT-4）主要有以下几种方式，以下是完整的实现指南：

使用OpenAI官方API（最推荐）

安装依赖

composer require openai-php/client

基本实现代码

<?php
require 'vendor/autoload.php';
use OpenAI\Client;
class GPTIntegration {
    private $client;
    public function __construct(string $apiKey) {
        $this->client = OpenAI::client($apiKey);
    }
    /**
     * 发送聊天请求
     */
    public function chat(array $messages, array $options = []): string {
        try {
            $response = $this->client->chat()->create([
                'model' => $options['model'] ?? 'gpt-3.5-turbo',
                'messages' => $messages,
                'temperature' => $options['temperature'] ?? 0.7,
                'max_tokens' => $options['max_tokens'] ?? 1500,
                'top_p' => $options['top_p'] ?? 1.0,
                'frequency_penalty' => $options['frequency_penalty'] ?? 0.0,
                'presence_penalty' => $options['presence_penalty'] ?? 0.0,
            ]);
            return $response->choices[0]->message->content;
        } catch (Exception $e) {
            return "错误: " . $e->getMessage();
        }
    }
    /**
     * 简单的问答
     */
    public function ask(string $prompt): string {
        return $this->chat([
            ['role' => 'user', 'content' => $prompt]
        ]);
    }
    /**
     * 流式响应（实时输出）
     */
    public function streamChat(array $messages, callable $callback): void {
        $stream = $this->client->chat()->createStreamed([
            'model' => 'gpt-3.5-turbo',
            'messages' => $messages,
        ]);
        foreach ($stream as $response) {
            $text = $response->choices[0]->delta->content ?? '';
            if ($text) {
                call_user_func($callback, $text);
            }
        }
    }
}

使用示例

// 初始化
$gpt = new GPTIntegration('你的OpenAI API密钥');
// 简单问答
$answer = $gpt->ask("什么是PHP？");
echo $answer;
// 多轮对话
$messages = [
    ['role' => 'system', 'content' => '你是一个PHP专家助手'],
    ['role' => 'user', 'content' => '如何在PHP中处理文件上传？'],
    ['role' => 'assistant', 'content' => '使用$_FILES超全局变量...'],
    ['role' => 'user', 'content' => '请给出具体代码示例']
];
$response = $gpt->chat($messages, [
    'model' => 'gpt-4',
    'temperature' => 0.3
]);
// 流式输出
echo "思考过程：";
$gpt->streamChat($messages, function($text) {
    echo $text;
    ob_flush();
    flush();
});

使用cURL直接调用API

<?php
class DirectGPTClient {
    private $apiKey;
    private $apiUrl = 'https://api.openai.com/v1/chat/completions';
    public function __construct(string $apiKey) {
        $this->apiKey = $apiKey;
    }
    public function ask(string $prompt, array $options = []): string {
        $data = [
            'model' => $options['model'] ?? 'gpt-3.5-turbo',
            'messages' => [
                ['role' => 'user', 'content' => $prompt]
            ],
            'temperature' => $options['temperature'] ?? 0.7,
            'max_tokens' => $options['max_tokens'] ?? 1500
        ];
        $ch = curl_init($this->apiUrl);
        curl_setopt_array($ch, [
            CURLOPT_RETURNTRANSFER => true,
            CURLOPT_POST => true,
            CURLOPT_HTTPHEADER => [
                'Content-Type: application/json',
                'Authorization: Bearer ' . $this->apiKey
            ],
            CURLOPT_POSTFIELDS => json_encode($data),
            CURLOPT_TIMEOUT => 60
        ]);
        $response = curl_exec($ch);
        $httpCode = curl_getinfo($ch, CURLINFO_HTTP_CODE);
        curl_close($ch);
        if ($httpCode !== 200) {
            throw new Exception("API请求失败: HTTP $httpCode");
        }
        $result = json_decode($response, true);
        return $result['choices'][0]['message']['content'] ?? '';
    }
}

安全性最佳实践

<?php
class SecureGPTIntegration {
    private $apiKey;
    private $rateLimiter;
    public function __construct() {
        // 从环境变量获取API密钥
        $this->apiKey = getenv('OPENAI_API_KEY');
        // 初始化速率限制器（需要先安装）
        // composer require predis/predis
        $this->rateLimiter = new \Predis\Client();
    }
    /**
     * 安全地发送请求
     */
    public function safeChat(string $userInput): string {
        // 1. 输入验证和清理
        $cleanedInput = $this->sanitizeInput($userInput);
        // 2. 速率限制检查
        if (!$this->checkRateLimit()) {
            return "请求过于频繁，请稍后再试。";
        }
        // 3. 内容过滤
        if ($this->containsSensitiveContent($cleanedInput)) {
            return "输入包含不允许的内容。";
        }
        // 4. Token预算控制
        if (strlen($cleanedInput) > 4000) {
            return "输入太长了，请缩短。";
        }
        // 5. 构建安全的消息
        $messages = [
            [
                'role' => 'system',
                'content' => '你是一个安全的AI助手，不能回答任何违法、有害或不道德的问题。'
            ],
            [
                'role' => 'user',
                'content' => $cleanedInput
            ]
        ];
        return $this->callAPI($messages);
    }
    private function sanitizeInput(string $input): string {
        // 移除危险字符
        $input = strip_tags($input);
        $input = htmlspecialchars($input, ENT_QUOTES, 'UTF-8');
        // 限制长度
        return mb_substr($input, 0, 5000);
    }
    private function checkRateLimit(): bool {
        $ip = $_SERVER['REMOTE_ADDR'];
        $key = "gpt_rate_limit:$ip";
        $current = $this->rateLimiter->incr($key);
        if ($current === 1) {
            $this->rateLimiter->expire($key, 60); // 60秒过期
        }
        return $current <= 20; // 每分钟最多20次请求
    }
    private function containsSensitiveContent(string $input): bool {
        $blockedPatterns = [
            '/黑客/i',
            '/攻击方法/i',
            '/制造武器/i',
            '/违法/i'
        ];
        foreach ($blockedPatterns as $pattern) {
            if (preg_match($pattern, $input)) {
                return true;
            }
        }
        return false;
    }
    private function callAPI(array $messages): string {
        // 实际API调用代码...
        return "API响应";
    }
}

Laravel项目集成示例

创建服务提供者

// app/Providers/OpenAIServiceProvider.php
<?php
namespace App\Providers;
use Illuminate\Support\ServiceProvider;
use OpenAI\Client;
class OpenAIServiceProvider extends ServiceProvider
{
    public function register(): void
    {
        $this->app->singleton(Client::class, function ($app) {
            return OpenAI::client(config('services.openai.api_key'));
        });
    }
}

创建Facade

// app/Facades/GPT.php
<?php
namespace App\Facades;
use Illuminate\Support\Facades\Facade;
class GPT extends Facade
{
    protected static function getFacadeAccessor(): string
    {
        return 'gpt';
    }
}

控制器中使用

<?php
namespace App\Http\Controllers;
use OpenAI\Client;
use Illuminate\Http\Request;
class ChatController extends Controller
{
    private $client;
    public function __construct(Client $client)
    {
        $this->client = $client;
    }
    public function chat(Request $request)
    {
        $request->validate([
            'message' => 'required|string|max:1000'
        ]);
        $response = $this->client->chat()->create([
            'model' => 'gpt-3.5-turbo',
            'messages' => [
                ['role' => 'user', 'content' => $request->message]
            ]
        ]);
        return response()->json([
            'reply' => $response->choices[0]->message->content,
            'usage' => $response->usage->toArray()
        ]);
    }
}

配置文件示例

// config/services.php
return [
    'openai' => [
        'api_key' => env('OPENAI_API_KEY'),
        'organization' => env('OPENAI_ORGANIZATION'), // 可选
        'default_model' => env('OPENAI_MODEL', 'gpt-3.5-turbo'),
        'max_tokens' => env('OPENAI_MAX_TOKENS', 1500),
        'temperature' => env('OPENAI_TEMPERATURE', 0.7),
    ]
];

环境变量配置

OPENAI_API_KEY=sk-your-api-key-here
OPENAI_ORGANIZATION=org-your-org-id
OPENAI_MODEL=gpt-3.5-turbo
OPENAI_MAX_TOKENS=2000
OPENAI_TEMPERATURE=0.7

注意事项

1. API密钥安全：永远不要将API密钥硬编码到代码中，使用环境变量
2. 成本控制：设置合适的token限制和速率限制
3. 错误处理：实现完善的异常捕获和回退机制
4. 缓存策略：对常见问题实现缓存以减少API调用
5. 隐私合规：不要将用户敏感数据发送到外部API

集成GPT到PHP项目时,建议先从小功能开始，逐步扩展，并密切监控API使用成本。