Unable to use new API for Document Intelligence

Question

I am trying out the code in this link with a valid key and endpoint Layout model sample code, but I am getting an error message. I am unable to understand what is the reason for this error as I am using the sample document in the sample code to test it out. When I try the same with a previous version as shown in Another sample code but with GA API
it seems to work fine. But the features I want to use are available in the new preview API. Please let me know how I can resolve this error.

Inner error: {

"code": "ParameterMissing",

"message": "The parameter urlSource or base64Source is required."

}

Answer 1

This code worked for me.

from azure.ai.documentintelligence.models import AnalyzeDocumentRequest
from azure.core.credentials import AzureKeyCredential
from azure.ai.documentintelligence import DocumentIntelligenceClient

def recognize_text(file_path):
    client = DocumentIntelligenceClient(
        endpoint=DOCS_ENDPOINT,
        credential=AzureKeyCredential(DOCS_API_KEY)
    )

    with open(file_path, "rb") as file:
        file_bytes = file.read()
        poller = client.begin_analyze_document(
            "prebuilt-read",
            AnalyzeDocumentRequest(bytes_source=file_bytes))
        result = poller.result()

    resulted_text = ""
    for page in result.pages:
        for line in page.lines:
            resulted_text += line.content + "\n"

    return resulted_text.strip()