Python SDK - Download files from SignatureRequestApi is not working anymore

[Striptik]

Hello,

I tried to download several signed documents from the api apis.SignatureRequestApi, with the method signature_request_files(signature_request_id, file_type="pdf") and it's not working anymore.

I implemented the same code as the exemple without success:

...
with ApiClient(configuration) as api_client:
    signature_request_api = apis.SignatureRequestApi(api_client)

    signature_request_id = "fa5c8a0b0f492d768749333ad6fcc214c111e967"

    try:
        response = signature_request_api.signature_request_files_as_data_uri(signature_request_id)
        pprint(response)
    except ApiException as e:
        print("Exception when calling Dropbox Sign API: %s\n" % e)

The exception thrown is 'utf-8' codec can't decode byte 0xe2 in position 10: continuation byte

The end of the exception traceback and the locals variable values at the exception time can maybe help

/mnt/vendor/lib/python3.12/site-packages/dropbox_sign/api_client.py:313 in response_deserialize                                                         │

   305          elif response_type == "file":
   306             return_data = self.__deserialize_file(response_data)
   307          elif response_type is not None:
   308             match = None
   309             content_type = response_data.getheader("content-type")
   310             if content_type is not None:
   311                match = re.search(r"charset=([a-zA-Z\-\d]+)[\s;]?"
   312             encoding = match.group(1) if match else "utf-8"
 ❱ 313             response_text = response_data.data.decode(encoding)
   314             return_data = self.deserialize(
   315                response_text, response_type, content_type
   316             )
   317       finally:
   318          if not 200 <= response_data.status <= 299:
   319             raise ApiException.from_response(
   320                http_resp=response_data,
   321                body=response_text,

 ── locals variables during the exception

content_type = 'application/pdf'
encoding = 'utf-8'
match = None 
msg = 'RESTResponse.read() must be called before passing it to response_deserialize()'                   
response_data = <dropbox_sign.rest.RESTResponse object at       
0xffff94684a30>                                 
response_text = None                                            
response_type = 'io.IOBase'                                     
response_types_map = {'200': 'io.IOBase', '4XX': 'ErrorResponse'}    
return_data = None                                            
-- 


UnicodeDecodeError: 'utf-8' codec can't decode byte 0xe2 in position 10: invalid continuation byte

It's working with the others download (Data Uri /File Url) methods, but i need the pdf directly so in the meantime I implemented the HTTP request and it's working, but less convenient

[pkatseas]

Hey @Striptik, I just ran into the same issue and after a little bit of debugging I spotted that the issue seems to be down to how response_types_map = {'200': 'io.IOBase', '4XX': 'ErrorResponse'} is not playing nice with the deserialize logic here.

While waiting for an official fix for this, I worked around it with the following snippet (excuse the rather basic copy/paste of the method signature, lacking the type annotations and validation that the original has):

from dropbox_sign import SignatureRequestApi


class PatchedSignatureRequestApi(SignatureRequestApi):
    def signature_request_files(
        self,
        signature_request_id: str,
        file_type: str,
        _request_timeout=None,
        _request_auth=None,
        _content_type=None,
        _headers=None,
        _host_index=0,
    ):
        _param = self._signature_request_files_serialize(
            signature_request_id=signature_request_id,
            file_type=file_type,
            _request_auth=_request_auth,
            _content_type=_content_type,
            _headers=_headers,
            _host_index=_host_index,
        )

        _response_types_map = {
            "200": "bytearray",
            "4XX": "ErrorResponse",
        }
        response_data = self.api_client.call_api(
            *_param, _request_timeout=_request_timeout
        )
        response_data.read()
        return self.api_client.response_deserialize(
            response_data=response_data,
            response_types_map=_response_types_map,
        ).data

Replacing any SignatureRequestApi usage with this should then succeed. 👍

[jtreminio-dropbox]

@Striptik thank you for the bug report, and humble apologies for taking so long to fix this.

@pkatseas thank you for the work-around. To keep backwards compatibility with customers using the SDK, I've changed the underlying SDK code to force use of io.BytesIO instead of bytearray. Accordingly, you'll need to update your usage to follow the examples code:

open("./file_response", "wb").write(response.read())

If you do not upgrade you do not need to apply the above fix, but once you do you'll need to keep this in mind.