Exception in deserilization of post message

[chandraarora]

Hi

We are getting the error when we try a method, please see the definition below. We are sending base64 string for byte[] and it is getting encoded again?

Could you please suggest what is wrong here?

Thanks
Chandra

System.Text.DecoderFallbackException: Unable to translate bytes [8B] at index 1 from specified code page to Unicode.
at System.Text.DecoderExceptionFallbackBuffer.Throw(Byte[] bytesUnknown, Int32 index)
at System.Text.DecoderExceptionFallbackBuffer.Fallback(Byte[] bytesUnknown, Int32 index)
at System.Text.Encoding.GetCharsWithFallback(ReadOnlySpan1 bytes, Int32 originalBytesLength, Span1 chars, Int32 originalCharsLength, DecoderNLS decoder)
at System.Text.UTF8Encoding.GetCharsWithFallback(ReadOnlySpan1 bytes, Int32 originalBytesLength, Span1 chars, Int32 originalCharsLength, DecoderNLS decoder)
at System.Text.Encoding.GetCharsWithFallback(Byte* pOriginalBytes, Int32 originalByteCount, Char* pOriginalChars, Int32 originalCharCount, Int32 bytesConsumedSoFar, Int32 charsWrittenSoFar, DecoderNLS decoder)
at System.Text.Encoding.GetChars(Byte* pBytes, Int32 byteCount, Char* pChars, Int32 charCount, DecoderNLS decoder)
at System.Text.DecoderNLS.GetChars(Byte* bytes, Int32 byteCount, Char* chars, Int32 charCount, Boolean flush)
at System.Text.DecoderNLS.GetChars(Byte[] bytes, Int32 byteIndex, Int32 byteCount, Char[] chars, Int32 charIndex, Boolean flush)
at System.Text.DecoderNLS.GetChars(Byte[] bytes, Int32 byteIndex, Int32 byteCount, Char[] chars, Int32 charIndex)
at Microsoft.AspNetCore.WebUtilities.HttpRequestStreamReader.ReadIntoBuffer()

Open API method

/api/v2/ConvertToPdf:
    post:
      consumes:
        - application/json
      tags:
        - convert
      description: >-
        Convert any type of office documents or images into PDFs
      operationId: ConvertToPdf_V1
      parameters:
        - in: body
          name: body
          required: false
          schema:
            properties:
              docContent:
                description: Map source File Content from the source action
                **format: byte
                title: File Content
                type: string**
              document:
                description: document
                properties:
                  Name:
                    description: Source file name with file extension
                    title: File Name
                    type: string
                required:
                  - Name
                type: object
            required:
              - docContent
              - document
            type: object
      responses:
        '200':
          description: File Converted successfully!
          headers:
            FileName:
              description: File Name
              type: string
            PageCount:
              description: Page Count
              type: string
          schema:
            description: Output file content from the API response
            format: binary
            title: File Content
            type: string
      summary: Convert to PDF

[baywet]

Hi @chandraarora
Thank you for using kiota and for reaching out.

Can you confirm in which language your client application is please?

[chandraarora]

Hi It's in typescript

[baywet]

Thank you for the additional information.

Can you share the code snippet where you're initializing the request?

Also why are you sharing a dotnet stack trace in the initial post?

[chandraarora]

Stack strace trace message is from server side and unable to deserialize in controller.
it happens like this

input - >{ "docName": "test.docx", "docContent": "asdf" }

output -> "\u001f�\b\u0000\u0000\u0000\u0000\u0000\u0000\n�VJ�Ov��+I�+Q�RJ,NIS�\u0001��%�*Y)��\u0016���'W(�\u0002\u0000F�Ô+\u0000\u0000\u0000",

Code sample

import { ApiKeyAuthenticationProvider, ApiKeyLocation, RequestConfiguration  } from "@microsoft/kiota-abstractions";
import { FetchRequestAdapter } from "@microsoft/kiota-http-fetchlibrary";
import { createPdf4meClient } from "../src/client/pdf4meClient.js";
import { ConvertToPdfPostRequestBody, ConvertToPdfPostRequestBody_document } from "../src/client/api/v2/convertToPdf/index.js";
import { WebhookSubscribePostRequestBody} from "../src/client/api/v2/webhookSubscribe/index.js";

async function  convertToPdf():  Promise<void> {
  try {
// API requires no authentication, so use the anonymous
// authentication provider
const authProvider = new ApiKeyAuthenticationProvider("Basic ****", "Authorization", ApiKeyLocation.Header);
// Create request adapter using the fetch-based implementation

const adapter = new FetchRequestAdapter(authProvider);
adapter.baseUrl= "http://localhost:5000";
// Create the API client
const client = createPdf4meClient(adapter);

const document :  ConvertToPdfPostRequestBody_document = 
{
  name: "test.png"
}

const body : ConvertToPdfPostRequestBody = 
{ 
  docContent : "base64 of png",
  document : document
}

const postReq =  client.api.v2.convertToPdf.post(body);

}catch (err) {
 console.log(err);
  }

}

convertToPdf() 

[baywet]

Thank you for the additional information.

When setting the doc content, are you base64 encoding the byte array? The client is already doing that for you.
Removing the base64 encoding from your application should solve your issue.

Let us know if you have any additional comments or questions.

[chandraarora]

Thank you, we will try and give you a feedback

[chandraarora]

DocContent is generated as string type from openapi spec, i can not assign byte []. May be i misunderstood.

[baywet]

Thank you for the additional information.

If you try generating a csharp client instead, what's the type of the property?

[chandraarora]

That is also string and I send base64 which works.

[baywet]

Thank you for the additional information.

In TypeScript, here is the serialization code that's generated.

export function serializeConvertToPdfPostRequestBody(writer: SerializationWriter, convertToPdfPostRequestBody: Partial<ConvertToPdfPostRequestBody> | undefined | null = {}) : void {
    if (convertToPdfPostRequestBody) {
        writer.writeStringValue("docContent", convertToPdfPostRequestBody.docContent);
        writer.writeObjectValue<ConvertToPdfPostRequestBody_document>("document", convertToPdfPostRequestBody.document, serializeConvertToPdfPostRequestBody_document);
        writer.writeAdditionalData(convertToPdfPostRequestBody.additionalData);
    }
}

And the equivalent in CSharp

public virtual void Serialize(ISerializationWriter writer)
        {
            _ = writer ?? throw new ArgumentNullException(nameof(writer));
            writer.WriteStringValue("docContent", DocContent);
            writer.WriteObjectValue<global::Graphdotnetv4.Api.V2.ConvertToPdf.ConvertToPdfPostRequestBody_document>("document", Document);
            writer.WriteAdditionalData(AdditionalData);
        }

As you can see, both languages are calling "writeStringValue" which in JSON ends up writing a JSON string without further transformations being applied.

I encourage you to compare the application code between the dotnet and javascript applications to understand whether the pre-processing you're applying results in different strings being set on the object for the same initial document.