Uploading curation from tsv

[ofersabo]

Hi,

Trying to upload an annotated TSV file as a curation document to a project.

I possess the curator file id as well as the document id.
The following code returns an error:

        with open(os.path.join(path_to_curation_dir, _name_document, "CURATION_USER.tsv")) as annotation:
            try:
                client.api.create_curation(project_id, document_id, curation_format=InceptionFormat.TSV, content=annotation)
            except InceptionBadResponse as e:
                print(e.bad_response.content)
                raise e

error:
b'{"messages":[{"level":"ERROR","message":"Internal server error: The document appears to be empty. Unable to detect any tokens or sentences. Empty documents cannot be imported."}]}'

Both the annotated TSV file on my file system contains annotations, as well as the curation file in the tool.

Any idea why this error occurs?

[ofersabo]

When trying to load the same TSV file from the swagger API, I get a different type of error.
error:

"message": "Internal server error: upload11954770680320360405.bin This is not a valid TSV File"

See the image with information about the request and the response.

[FrogFeather]

Following up on this, I'm experiencing the same problem even just for uploading regular annotations for a document.

    try:
        client.api.create_annotation(project_id, document_id, annotation_format=InceptionFormat.TSV, content=annotation)
    except InceptionBadResponse as e:
        print(e.bad_response.content)
        raise e

Yet I still receive the following error:
b'{"messages":[{"level":"ERROR","message":"Internal server error: The document appears to be empty. Unable to detect any tokens or sentences. Empty documents cannot be imported."}]}'

I've tried uploading annotations in various different formats (CONLL, XMI, TCF, NIF, etc.) but nothing works.

[ofersabo]

Hi @JavierLuna

Comment here so you'd see this issue.

[JavierLuna]

Thank you for the valuable feedback, I'm working on it in inception-project/pycaprio#1

[reckart]

I have tried reproducing this by exporting a document as WebAnno TSV 3.2 from (format identifier ctsv3) from the curation page and then using the Swagger API again to upload the curation document (again using format identifier ctsv3). It worked for me.

I'll comment on inception-project/pycaprio#1