通过

操作指南：如何在 OneLake 中存储数据

每个 Fabric 项目都可以访问 OneLake 存储，后者为存储与项目关联的文件提供了安全且可扩展的方式。 本指南介绍如何使用FabricPlatformAPIClient和底层的OneLakeClient将文件上传到您的Fabric项目。

了解 Fabric 中的物品存储

Fabric 中的每个项在 OneLake 中都有自己的专用存储区域。 此存储组织到文件夹中，主文件夹为：

  • 文件 - 用于存储常规文件和文档
  • 表 - 用于存储表数据

先决条件

将文件上传到您的项目之前，您需要：

  • 有效的 Fabric 工作区
  • 您要上传文件的现有项目
  • 写入项的适当权限

创建 FabricPlatformAPIClient

首先，创建实例 FabricPlatformAPIClient ：

import { getWorkloadClient } from "../controller/WorkloadClient";
import { FabricPlatformAPIClient } from "../clients/FabricPlatformAPIClient";

// Create client using the current user's context
const fabricClient = FabricPlatformAPIClient.create(getWorkloadClient());

示例：将文本文件上传到项目

下面介绍如何使用 OneLake 客户端将文本文件上传到项目：

async function uploadTextFileToItem(
  workspaceId: string, 
  itemId: string, 
  fileName: string, 
  content: string
) {
  try {
    // Get the FabricPlatformAPIClient
    const fabricClient = FabricPlatformAPIClient.create(getWorkloadClient());
    
    // Access the OneLake client
    const oneLakeClient = fabricClient.oneLake;
    
    // Generate the file path in OneLake for this item
    // This follows the pattern: workspaceId/itemId/Files/fileName
    const filePath = oneLakeClient.constructor.getFilePath(workspaceId, itemId, fileName);
    
    // Write the text content to the file
    await oneLakeClient.writeFileAsText(filePath, content);
    
    console.log(`Successfully uploaded ${fileName} to item ${itemId}`);
    return true;
  } catch (error) {
    console.error("Error uploading file to item:", error);
    throw error;
  }
}

示例：将二进制文件上传到项

对于图像或 PDF 等二进制文件，首先需要将文件转换为 base64：

async function uploadBinaryFileToItem(
  workspaceId: string,
  itemId: string,
  fileName: string,
  fileData: ArrayBuffer // Binary file data
) {
  try {
    const fabricClient = FabricPlatformAPIClient.create(getWorkloadClient());
    const oneLakeClient = fabricClient.oneLake;
    
    // Convert binary data to base64
    const base64Content = arrayBufferToBase64(fileData);
    
    // Generate the file path
    const filePath = oneLakeClient.constructor.getFilePath(workspaceId, itemId, fileName);
    
    // Write the binary content to the file
    await oneLakeClient.writeFileAsBase64(filePath, base64Content);
    
    console.log(`Successfully uploaded binary file ${fileName} to item ${itemId}`);
    return true;
  } catch (error) {
    console.error("Error uploading binary file to item:", error);
    throw error;
  }
}

// Helper function to convert ArrayBuffer to base64
function arrayBufferToBase64(buffer: ArrayBuffer): string {
  let binary = '';
  const bytes = new Uint8Array(buffer);
  const len = bytes.byteLength;
  
  for (let i = 0; i < len; i++) {
    binary += String.fromCharCode(bytes[i]);
  }
  
  return btoa(binary);
}

示例：从浏览器上传文件

如果要生成 Web 界面，可以使用此函数处理文件输入中的文件上传：

async function handleFileUpload(
  workspaceId: string,
  itemId: string,
  fileInputElement: HTMLInputElement
) {
  if (!fileInputElement.files || fileInputElement.files.length === 0) {
    console.warn("No file selected");
    return false;
  }
  
  const file = fileInputElement.files[0];
  const fileName = file.name;
  
  try {
    // Read the file as ArrayBuffer
    const fileBuffer = await readFileAsArrayBuffer(file);
    
    // Upload based on file type
    if (file.type.startsWith('text/')) {
      // For text files, convert to string and upload as text
      const textDecoder = new TextDecoder();
      const textContent = textDecoder.decode(fileBuffer);
      
      return await uploadTextFileToItem(workspaceId, itemId, fileName, textContent);
    } else {
      // For binary files, upload as base64
      return await uploadBinaryFileToItem(workspaceId, itemId, fileName, fileBuffer);
    }
  } catch (error) {
    console.error("Error processing file upload:", error);
    throw error;
  }
}

// Helper function to read file as ArrayBuffer
function readFileAsArrayBuffer(file: File): Promise<ArrayBuffer> {
  return new Promise((resolve, reject) => {
    const reader = new FileReader();
    reader.onload = () => resolve(reader.result as ArrayBuffer);
    reader.onerror = reject;
    reader.readAsArrayBuffer(file);
  });
}

管理项目中的文件

上传文件后，还可以：

检查文件是否存在

async function checkFileExists(workspaceId: string, itemId: string, fileName: string) {
  const fabricClient = FabricPlatformAPIClient.create(getWorkloadClient());
  const filePath = fabricClient.oneLake.constructor.getFilePath(workspaceId, itemId, fileName);
  
  return await fabricClient.oneLake.checkIfFileExists(filePath);
}

读取文件内容

async function readTextFile(workspaceId: string, itemId: string, fileName: string) {
  const fabricClient = FabricPlatformAPIClient.create(getWorkloadClient());
  const filePath = fabricClient.oneLake.constructor.getFilePath(workspaceId, itemId, fileName);
  
  return await fabricClient.oneLake.readFileAsText(filePath);
}

删除文件

async function deleteFile(workspaceId: string, itemId: string, fileName: string) {
  const fabricClient = FabricPlatformAPIClient.create(getWorkloadClient());
  const filePath = fabricClient.oneLake.constructor.getFilePath(workspaceId, itemId, fileName);
  
  await fabricClient.oneLake.deleteFile(filePath);
  console.log(`File ${fileName} deleted successfully`);
}

文件上传的最佳做法

  • 使用适当的文件格式：考虑文件的目的，并使用广受支持的格式。
  • 优雅地处理错误：确保始终包括对网络问题或权限问题的错误处理。
  • 验证文件大小：大型文件上传和处理可能需要更长的时间。
  • 检查权限：在尝试上传之前，请确保用户具有适当的权限。
  • 使用文件前缀或文件夹：对于包含许多文件的复杂项目，请考虑在子文件夹中组织它们。

使用 OneLakeClientItemWrapper

为了简化对项文件的访问，可以使用 OneLakeClientItemWrapper：

async function uploadFileWithItemWrapper(item, fileName, content) {
  const fabricClient = FabricPlatformAPIClient.create(getWorkloadClient());
  
  // Create a wrapper for simpler access to this specific item
  const itemWrapper = fabricClient.oneLake.createItemWrapper({
    workspaceId: item.workspaceId,
    itemId: item.id
  });
  
  // Upload directly to the item (no need to specify paths)
  await itemWrapper.writeFileAsText(fileName, content);
  
  // Read the file back
  const fileContent = await itemWrapper.readFileAsText(fileName);
  
  console.log(`File uploaded and read back: ${fileContent.substring(0, 50)}...`);
}

此包装器通过自动处理完整路径构造来简化文件作。

  • Last updated on