Help with supported Document Types Azure Documents Intelligence SDK Client

Derek Crain 0

Hello.

I am currently trying to use the Azure AI Document Intelligence SDK to upload and process files using the Read model. I am using the latest version of the sdk and the latest api version for the service.

I can successfully upload PDF documents, but when I attempt upload any of the supported Microsoft Office documents (Doc, Excel, Powerpoint) as suggested to be supported by the documentation I receive an error that the file type is unsupported.

I have previously opened a github issue that pointed me towards here for further answers. It contains the code used and the specific error response.

1 answer

Saideep Anchuri 330 Reputation points Microsoft Vendor

2024-11-14T06:50:28.54+00:00
Hi Derek Crain

Welcome to Microsoft Q&A Forum, thank you for posting your query here!

It appears that you're facing a challenge with file type compatibility while utilizing the Azure AI Document Intelligence SDK. The SDK is designed to handle multiple document formats, there may be specific conditions or limitations that need to be considered.

Ensure that the documents are saved in the appropriate formats (DOCX, XLSX, PPTX).

make sure the file size does not exceed the service's allowable limits.

Ensure you're using the appropriate API version that supports Office documents.

Confirm you have the required permissions to access and upload the files.

Refer to the official Azure documentation for any specific requirements or restrictions related to document types.

Kindly refer Below Document:

https://learn.microsoft.com/en-us/python/api/overview/azure/ai-documentintelligence-readme?view=azure-python-preview

Hope this helps. Do let us know if you any further queries.

If this answers your query, do click Accept Answer and Yes for was this answer helpful. And, if you have any further query do let us know.

Thank You.
Please sign in to rate this answer.
Derek Crain 0 Reputation points

2024-11-14T21:47:50.2133333+00:00

sorry duplicate

Derek Crain 0 Reputation points

2024-11-14T22:01:32.2433333+00:00

I have done all of these and so far have not had success.

I am uploading a DOCX file of a small size for testing purposes.

I have the permissions, as I can successfully upload PDF and Image files.

I am using the nuget package Azure.AI.DocumentIntelligence v 1.0.0-beta.3, which supports version 2024-07-31-preview of the API for the service.

This is for my .NET project so the python documentation is not applicable. I have linked the documentation that I have used to reach this point.

The Document Intelligence resource is provisioned in West US 2 which is one of the regions that supports that api version according to the read model documentation I linked.

Please advise.

Saideep Anchuri 330 Reputation points Microsoft Vendor

2024-11-15T01:49:34.8+00:00

Hi Derek Crain

Sorry for the trouble you are facing.

Make sure the DOCX file is not corrupted and is in the proper format. Even minor discrepancies can lead to problems.

When uploading Office documents, ensure the Content-Type header in your request is correctly set. For Word files, it should be application/vnd.openxmlformats-officedocument.wordprocessingml.document; for Excel files, use application/vnd.openxmlformats-officedocument.spreadsheetml.sheet; and for PowerPoint files, use application/vnd.openxmlformats-officedocument.presentationml.presentation. If the Content-Type is incorrect, the server may reject the request.

In .NET, make sure to include this header when making the request, especially if you're directly using the REST API. However, the SDK usually takes care of this automatically.

If you're uploading a document from local storage, you might use methods like AnalyzeDocumentFromStream or AnalyzeDocumentFromFile. Ensure the file stream is correctly passed and that the SDK accurately identifies the file type.

Could you please take a moment to retake the survey on the above response? Your feedback is greatly appreciated.

Thank You.

Deleted

This comment has been deleted due to a violation of our Code of Conduct. The comment was manually reported or identified through automated detection before action was taken. Please refer to our Code of Conduct for more information.

Saideep Anchuri 330 Reputation points Microsoft Vendor

2024-11-16T02:34:19.6933333+00:00

Hi Derek Crain,

I understand that you're working with .NET and facing challenges with processing DOCX files. I want to reassure you that the Azure AI Document Intelligence SDK for .NET supports the processing of various document types, including DOCX files, using the prebuilt-read model, as outlined in the documentation. I successfully worked with Python for similar tasks, and the concept remains consistent across languages—whether you're using Python or C#, you can utilize the SDK to process DOCX files.

Here’s the C# code that demonstrates how to upload and process a DOCX document using the DocumentAnalysisClient:

using Azure; using Azure.AI.FormRecognizer.DocumentAnalysis; string endpoint = "YOUR_FORM_RECOGNIZER_ENDPOINT"; string key = "YOUR_FORM_RECOGNIZER_KEY"; AzureKeyCredential credential = new AzureKeyCredential(key); DocumentAnalysisClient client = new DocumentAnalysisClient(new Uri(endpoint), credential); // Sample document (can replace with local file path) Uri fileUri = new Uri( AnalyzeDocumentOperation operation = await client.AnalyzeDocumentFromUriAsync(WaitUntil.Completed, "prebuilt-read", fileUri); AnalyzeResult result = operation.Value; foreach (DocumentPage page in result.Pages) { Console.WriteLine($"Document Page {page.PageNumber} has {page.Lines.Count} line(s), {page.Words.Count} word(s),"); for (int i = 0; i < page.Lines.Count; i++) { DocumentLine line = page.Lines[i]; Console.WriteLine($" Line {i} has content: '{line.Content}'."); Console.WriteLine($" Its bounding box is:"); Console.WriteLine($" Upper left => X: {line.BoundingPolygon[0].X}, Y= {line.BoundingPolygon[0].Y}"); Console.WriteLine($" Upper right => X: {line.BoundingPolygon[1].X}, Y= {line.BoundingPolygon[1].Y}"); Console.WriteLine($" Lower right => X: {line.BoundingPolygon[2].X}, Y= {line.BoundingPolygon[2].Y}"); Console.WriteLine($" Lower left => X: {line.BoundingPolygon[3].X}, Y= {line.BoundingPolygon[3].Y}"); } } foreach (DocumentStyle style in result.Styles) { bool isHandwritten = style.IsHandwritten.HasValue && style.IsHandwritten == true; if (isHandwritten && style.Confidence > 0.8) { Console.WriteLine($"Handwritten content found:"); foreach (DocumentSpan span in style.Spans) { Console.WriteLine($" Content: {result.Content.Substring(span.Index, span.Length)}"); } } } foreach (DocumentLanguage language in result.Languages) { Console.WriteLine($" Found language '{language.Locale}' with confidence {language.Confidence}."); }

Here is the output I received using the python

In your case, processing DOCX files should work seamlessly with this code by simply providing the correct file path or URI. The SDK's capabilities are robust and cross-platform, so the implementation in .NET should offer the same benefits you would expect from Python. If you’re facing issues, it might be helpful to check the specific API version, permissions, and document formatting, as the API's ability to process Office documents depends on these factors.

Let me know, if you face any issues and will try to resolve the issue.

Could you please take a moment to retake the survey on the above response? Your feedback is greatly appreciated

Saideep Anchuri 330 Reputation points Microsoft Vendor

2024-11-18T00:31:53.07+00:00

Hi Derek Crain,

Following up to see if the above answer was helpful.

Thank You.

Saideep Anchuri 330 Reputation points Microsoft Vendor

2024-11-19T00:50:28.7666667+00:00

Hi Derek Crain,

We haven’t heard from you on the last response and was just checking back to see if the give response was helpful.

Thank you.

Derek Crain 0 Reputation points

2024-11-19T18:10:03.9733333+00:00

I understand. This is the same code that I have used to analyze PDF documents with success. I now suspect that my organization's use of security labeling on microsoft office documents to be part of the issue, can you investigate to see if that might be the cause?

Saideep Anchuri 330 Reputation points Microsoft Vendor

2024-11-20T06:09:55.0966667+00:00

Hi Derek Crain,

It seems likely that security labeling could be affecting the document analysis. Security labels in Microsoft 365, like those from Microsoft Purview Information Protection, can limit access to sensitive data. If the Document Intelligence service lacks the appropriate permissions to access labeled content, it might not be able to analyze the document correctly

Thank You.

Derek Crain 0 Reputation points

2024-11-20T06:30:34.7233333+00:00

How would I give the document intelligence service the appropriate permissions for this use?

Saideep Anchuri 330 Reputation points Microsoft Vendor

2024-11-20T15:11:37.7133333+00:00

Hi Derek Crain,

Thank you for your follow-up query.

To grant Azure Document Intelligence access to security-labeled documents, assign it the necessary Azure AD roles (e.g., Reader). Ensure the service has permissions to access the information, possibly by adding the service identity to an appropriate security group.

Please check: https://learn.microsoft.com/en-us/azure/role-based-access-control/overview

Could you please take a moment to retake the survey on my initial response? Your feedback is greatly appreciated.

Thank you.
Sign in to comment

Use comments to ask for clarification, additional information, or improvements to the question.

Share via

Help with supported Document Types Azure Documents Intelligence SDK Client

1 answer

Your answer