[Python] Azure documen intelligence python SDK doesn't separate pages

When trying to extract content from a MS Word .docx file using Azure Document Intelligence, I expected the returned response to contain a page element for each page in the document and for each of those page elements to contain multiple lines in line with the documentation.

Instead, I always receive as a single page with no (None) lines and the entire document's contents as a list of words.

Sample document:

Minimal reproducible example:

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

def main():
client = DocumentIntelligenceClient(
'MY ENDPOINT',
AzureKeyCredential('MY KEY')
)

document = 'small_test_document.docx'

with open(document, "rb") as f:
poller = client.begin_analyze_document(
"prebuilt-layout",
analyze_request=f,
content_type="application/octet-stream"
)
result = poller.result()

print(f'Found {len(result.pages)} page(s)')
for page in result.pages:
print(f'Page #{page.page_number}')
print(f' {page.lines=}')
print(f' {len(page.words)=}')

if __name__ == '__main__':
main()


Expected output:

Found 2 page(s)
Page #1
page.lines=6
len(page.words)=58
Page #2
page.lines=1
len(page.words)=8


Actual output:

Found 1 page(s)
Page #1
page.lines=None
len(page.words)=66


My question is: Why, and what should I do differently to get the expected output?

Continue reading...