将 PDF ( > 4 MB) 上传到 D365 CRM 中的文件数据类型

Question

I'm trying to upload a PDF file to a CRM record. I've used a File type field in the entity that can hold my uploaded file. I've done this by using this code:

UploadBlockRequest blockRequest = new UploadBlockRequest();
blockRequest.BlockData = Convert.FromBase64String(documentBody);
blockRequest.BlockId = Convert.ToBase64String(Encoding.UTF8.GetBytes(Guid.NewGuid().ToString()));
blockRequest.FileContinuationToken = initResponse.FileContinuationToken;

var blockResponse = (UploadBlockResponse)service.Execute(blockRequest);

It is working fine for PDF files that are under 4 MB. However, If I try to upload a PDF that's more than 4 MB, I get the following error:

Invalid file chunk size: 4 MB. Maximum chunk size supported: 4 MB.

Is there a way to upload large PDF files to CRM record?

using (var stream = new MemoryStream(Convert.FromBase64String(Base64)))
                    {
                        InitializeFileBlocksUploadRequest initializeUploadRequest = new InitializeFileBlocksUploadRequest();
                        initializeUploadRequest.FileAttributeName = "my_fileTypeField";
                        initializeUploadRequest.FileName = "Test.pdf";
                        initializeUploadRequest.Target = new EntityReference("my_entity", new Guid("my_guid"));

                        var initializeUploadResponse = (InitializeFileBlocksUploadResponse)service.Execute(initializeUploadRequest);
                        var uploadRequest = new UploadBlockRequest { FileContinuationToken = initializeUploadResponse.FileContinuationToken };

                        const int blockSize = 4194304; // 4MB
                        int byteCount;
                        var blockList = new List<string>();

                        do
                        {
                            
                            //uploadRequest.BlockData = Convert.FromBase64String(documentBody);
                            byteCount = stream.Read(uploadRequest.BlockData, 0, blockSize);
                            uploadRequest.BlockId = Convert.ToBase64String(Guid.NewGuid().ToByteArray());
                            service.Execute(uploadRequest);
                            blockList.Add(uploadRequest.BlockId);
                            Console.WriteLine(size + " == " + blockSize);
                        } while (size == blockSize);

                        var commitRequest = new CommitFileBlocksUploadRequest
                        {
                            BlockList = blockList.ToArray(),
                            FileContinuationToken = initializeUploadResponse.FileContinuationToken,
                            FileName = initializeUploadRequest.FileName,
                            MimeType = "application/pdf"
                        };
                        var commitResponse = (CommitFileBlocksUploadResponse)service.Execute(commitRequest);
                    }

Answer 1

The UploadBlockRequest is one piece in the file upload procedure. You need 3 distinct requests:

1. InitializeFileBlocksUploadRequest
2. UploadBlockRequest
3. CommitFileBlocksUploadRequest

The UploadBlockRequest can hold a chunk of 4 MB of data at a maximum. Your file can be as large as 128 MB and it can be uploaded using multiple upload requests.

A basic upload method could look like this:

private Guid UploadFile
    (
        FileStream stream,
        string fileName,
        string mimeType,
        EntityReference target,
        string fileAttributeName,
        IOrganizationService organizationService
    )
{
    var initializeUploadRequest = new InitializeFileBlocksUploadRequest
    {
        FileAttributeName = fileAttributeName,
        FileName = fileName,
        Target = target
    };

    var initializeUploadResponse = (InitializeFileBlocksUploadResponse)organizationService.Execute(initializeUploadRequest);
    var uploadRequest = new UploadBlockRequest { FileContinuationToken = initializeUploadResponse.FileContinuationToken };

    const int blockSize = 4194304; // 4 MB
    int byteCount;
    var blockList = new List<string>();

    do
    {
        byteCount = stream.Read(uploadRequest.BlockData, 0, blockSize);
        uploadRequest.BlockId = Convert.ToBase64String(Guid.NewGuid().ToByteArray());
        organizationService.Execute(uploadRequest);
        blockList.Add(uploadRequest.BlockId);
    } while (byteCount == blockSize);

    var commitRequest = new CommitFileBlocksUploadRequest
    {
        BlockList = blockList.ToArray(),
        FileContinuationToken = initializeUploadResponse.FileContinuationToken,
        FileName = initializeUploadRequest.FileName,
        MimeType = mimeType
    };

    var commitResponse = (CommitFileBlocksUploadResponse)organizationService.Execute(commitRequest);
    return commitResponse.FileId;
}

The method uploads the file and returns the ID of the file.

Answer 2

I modified the example from Henk a little because I was having the error:

Buffer cannot be null

However, I believe my approach won't work for big files. I have tested this approach for files up to 5MB sucessfully. Files bigger than that will need to be uploaded using the approach from Henk. I'm not a coding genius myself to figure out the way Henk's example can be improved to avoid the bug.

private Guid UploadFile(byte[] fileBytes, string fileName, string mimeType, EntityReference target, string fileAttributeName, IOrganizationService ctx)
    {
        var initializeUploadRequest = new InitializeFileBlocksUploadRequest
        {
            FileAttributeName = fileAttributeName,
            FileName = fileName,
            Target = target
        };


        var initializeUploadResponse = (InitializeFileBlocksUploadResponse)ctx.Execute(initializeUploadRequest);
        var uploadRequest = new UploadBlockRequest
        {
            FileContinuationToken = initializeUploadResponse.FileContinuationToken,
            BlockData = fileBytes,
            BlockId = Convert.ToBase64String(Guid.NewGuid().ToByteArray())
        };


        var blockList = new List<string>();

        ctx.Execute(uploadRequest);
        blockList.Add(uploadRequest.BlockId);


        var commitRequest = new CommitFileBlocksUploadRequest
        {
            BlockList = blockList.ToArray(),
            FileContinuationToken = initializeUploadResponse.FileContinuationToken,
            FileName = initializeUploadRequest.FileName,
            MimeType = mimeType
        };

        var commitResponse = (CommitFileBlocksUploadResponse)ctx.Execute(commitRequest);
        return commitResponse.FileId;
    }
}

I do work with a Byte Array as the input instead of a stream, as in my Solution I'm downloading a file from 1 entity (which is returned as a byte[]), to upload it in another entity.