Azure-Samples / Azure-Samples/document-intelligence-code-samples

Support for Query Fields at Line Item Level (e.g., IGST Rate, IGST Amount in Invoice Items)

Open
#123 0 comments 0 reactions 0 assignees View on GitHub
Dominant language
C#
Stars
201
Forks
115
PR merge metrics
No merged PRs in 30d

Description

Description


Currently, the query_fields parameter in the begin_analyze_document API allows us to specify high-level fields such as Address or InvoiceNumber.


For example:


poller = document_intelligence_client.begin_analyze_document(
"prebuilt-layout",
AnalyzeDocumentRequest(url_source=formUrl),
features=[DocumentAnalysisFeature.QUERY_FIELDS],
query_fields=[
"Address", "InvoiceNumber"],
)

This works well for document-level fields.


However, I would like to specify query fields for each line item (product) in an invoice. For instance, in the attached sample invoice, I need to extract IGST Rate and IGST Amount for each line item in the product table.


Problem


At the moment, query_fields seems to only work for top-level fields and does not allow targeting fields inside a repeating structure like invoice line items.


Request




  • Is there a way to define query fields at the line item (per product) level?




  • If not, can support for item-level query fields (e.g., tax percentages, amounts, discounts per line item) be added in a future release?




Example Use Case


For an invoice line item table:

![Image](https://github.com/user-attachments/assets/df93fdc4-bb9f-4bb2-98e5-4f87aa0f88c0)

I want to query and extract IGST Rate and IGST Amount along with other line item details.

Contributor guide

Open the contributing guide

Research direction

Start with the Python begin_analyze_document example and its query_fields and QUERY_FIELDS parameters. Check whether the Document Intelligence API supports queries inside repeating invoice line items, including IGST Rate and IGST Amount. Done means documenting the supported approach or clearly recording that item-level query fields require future service support.

Written by the indexing model from the issue text.

Assessment

Tech stack
azure, python
Domain
api
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.