SDK 与代码示例

pip install requests

import requests
import time

class DocumentParser:
    def __init__(self, api_key):
        self.api_key = api_key
        self.base_url = "http://8.148.69.123:8088/api"

    def create_task(self, file_url, **options):
        """创建解析任务"""
        headers = {
            "Authorization": f"Bearer {self.api_key}",
            "Content-Type": "application/json"
        }

        data = {
            "url": file_url,
            "is_ocr": options.get("is_ocr", True),
            "enable_formula": options.get("enable_formula", True),
            "enable_table": options.get("enable_table", True),
        }

        response = requests.post(
            f"{self.base_url}/v4/extract/task",
            headers=headers,
            json=data
        )

        result = response.json()
        if result["code"] == 0:
            return result["data"]["task_id"]
        else:
            raise Exception(f"创建任务失败: {result['msg']}")

    def get_result(self, task_id):
        """获取任务结果"""
        headers = {
            "Authorization": f"Bearer {self.api_key}"
        }

        response = requests.get(
            f"{self.base_url}/v4/extract/task/{task_id}",
            headers=headers
        )

        return response.json()["data"]

    def wait_for_completion(self, task_id, max_wait=300):
        """等待任务完成"""
        start_time = time.time()

        while time.time() - start_time < max_wait:
            result = self.get_result(task_id)

            if result["state"] == "done":
                return result
            elif result["state"] == "failed":
                raise Exception(f"解析失败: {result.get('err_msg')}")

            time.sleep(5)

        raise Exception("任务超时")

# 使用示例
parser = DocumentParser("YOUR_API_KEY")

# 创建任务
task_id = parser.create_task(
    "https://example.com/document.pdf",
    is_ocr=True,
    enable_formula=True
)

print(f"任务 ID: {task_id}")

# 等待完成
result = parser.wait_for_completion(task_id)
print(f"解析完成！下载链接: {result['full_zip_url']}")

npm install axios

const axios = require('axios');

class DocumentParser {
  constructor(apiKey) {
    this.apiKey = apiKey;
    this.baseURL = 'http://8.148.69.123:8088/api';
  }

  async createTask(fileUrl, options = {}) {
    try {
      const response = await axios.post(
        `${this.baseURL}/v4/extract/task`,
        {
          url: fileUrl,
          is_ocr: options.isOcr || true,
          enable_formula: options.enableFormula || true,
          enable_table: options.enableTable || true,
        },
        {
          headers: {
            'Authorization': `Bearer ${this.apiKey}`,
            'Content-Type': 'application/json',
          },
        }
      );

      if (response.data.code === 0) {
        return response.data.data.task_id;
      } else {
        throw new Error(`创建任务失败: ${response.data.msg}`);
      }
    } catch (error) {
      throw new Error(`API 请求失败: ${error.message}`);
    }
  }

  async getResult(taskId) {
    const response = await axios.get(
      `${this.baseURL}/v4/extract/task/${taskId}`,
      {
        headers: {
          'Authorization': `Bearer ${this.apiKey}`,
        },
      }
    );

    return response.data.data;
  }

  async waitForCompletion(taskId, maxWait = 300000) {
    const startTime = Date.now();

    while (Date.now() - startTime < maxWait) {
      const result = await this.getResult(taskId);

      if (result.state === 'done') {
        return result;
      } else if (result.state === 'failed') {
        throw new Error(`解析失败: ${result.err_msg}`);
      }

      await new Promise(resolve => setTimeout(resolve, 5000));
    }

    throw new Error('任务超时');
  }
}

// 使用示例
(async () => {
  const parser = new DocumentParser('YOUR_API_KEY');

  try {
    // 创建任务
    const taskId = await parser.createTask(
      'https://example.com/document.pdf',
      {
        isOcr: true,
        enableFormula: true,
      }
    );

    console.log(`任务 ID: ${taskId}`);

    // 等待完成
    const result = await parser.waitForCompletion(taskId);
    console.log(`解析完成！下载链接: ${result.full_zip_url}`);
  } catch (error) {
    console.error('错误:', error.message);
  }
})();

import java.net.http.*;
import java.net.URI;
import com.google.gson.Gson;
import com.google.gson.JsonObject;

public class DocumentParser {
    private String apiKey;
    private String baseUrl = "http://8.148.69.123:8088/api";
    private HttpClient client;
    private Gson gson;

    public DocumentParser(String apiKey) {
        this.apiKey = apiKey;
        this.client = HttpClient.newHttpClient();
        this.gson = new Gson();
    }

    public String createTask(String fileUrl, boolean isOcr, boolean enableFormula) throws Exception {
        JsonObject requestBody = new JsonObject();
        requestBody.addProperty("url", fileUrl);
        requestBody.addProperty("is_ocr", isOcr);
        requestBody.addProperty("enable_formula", enableFormula);

        HttpRequest request = HttpRequest.newBuilder()
            .uri(URI.create(baseUrl + "/v4/extract/task"))
            .header("Authorization", "Bearer " + apiKey)
            .header("Content-Type", "application/json")
            .POST(HttpRequest.BodyPublishers.ofString(gson.toJson(requestBody)))
            .build();

        HttpResponse<String> response = client.send(request, HttpResponse.BodyHandlers.ofString());
        JsonObject result = gson.fromJson(response.body(), JsonObject.class);

        if (result.get("code").getAsInt() == 0) {
            return result.getAsJsonObject("data").get("task_id").getAsString();
        } else {
            throw new Exception("创建任务失败: " + result.get("msg").getAsString());
        }
    }

    // 使用示例
    public static void main(String[] args) {
        try {
            DocumentParser parser = new DocumentParser("YOUR_API_KEY");
            String taskId = parser.createTask("https://example.com/document.pdf", true, true);
            System.out.println("任务 ID: " + taskId);
        } catch (Exception e) {
            e.printStackTrace();
        }
    }
}