✨ add API for RAG operations

[ianardee]

Description

Types of changes

• Docs change / refactoring / dependency upgrade
• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to change)
• Requires a change to the official Guide documentation.

[Copilot]

Copilot review overview

🟡 Changes recommended

Polling validation and HTTP-client cleanup are broken, compatibility APIs are removed, and integration cleanup is unreliable.

Review effort: Balanced
Findings: 5

Open (5)

• Validate polling settings before the initial sleep · New
• Update finalizers to close the renamed _http_client · New
• Retain get_result_slug as a compatibility alias · New
• Preserve compatibility exports for StringDataClass · New
• Always delete test documents in a finally block · New

What changed in this PR

Adds V2 APIs for uploading, retrieving, annotating, polling, and deleting extraction RAG documents.

Changes:

• Adds RAG document models, parameters, HTTP operations, and lifecycle methods.
• Introduces registry-based inference and annotation field parsing.
• Updates serialization, response slugs, tests, and CI coverage.

File	Description
.github/​workflows/​_test-units.yml	Reduces lite-test Python matrix.
tests/​v2/​test_client.py	Updates private HTTP client access.
tests/​v2/​test_client_integration.py	Uses typed field retrieval.
tests/​v2/​test_base_classes.py	Tests renamed product slug API.
tests/​v2/​product/​extraction/​test_rag_documents.py	Tests RAG parameters and parsing.
tests/​v2/​product/​extraction/​test_rag_documents_integration.py	Tests the RAG document lifecycle.
tests/​v2/​product/​extraction/​test_extraction_response.py	Updates public imports.
tests/​v2/​input/​test_inference_parameters.py	Expects compact JSON serialization.
mindee/​v2/​product/​extraction/​rag_documents/​rag_annotation.py	Defines RAG annotations.
mindee/​v2/​product/​extraction/​rag_documents/​params/​rag_document_upload_parameters.py	Defines upload parameters.
mindee/​v2/​product/​extraction/​rag_documents/​params/​rag_document_annotation_parameters.py	Defines annotation updates.
mindee/​v2/​product/​extraction/​rag_documents/​params/​__init__.py	Exports RAG parameters.
mindee/​v2/​product/​extraction/​rag_documents/​extraction_rag_annotation_response.py	Parses extraction RAG responses.
mindee/​v2/​product/​extraction/​rag_documents/​annotated_simple_field.py	Adds annotated simple fields.
mindee/​v2/​product/​extraction/​rag_documents/​annotated_object_field.py	Adds annotated object fields.
mindee/​v2/​product/​extraction/​rag_documents/​annotated_list_field.py	Adds annotated list fields.
mindee/​v2/​product/​extraction/​rag_documents/​annotated_fields.py	Provides typed annotation lookup.
mindee/​v2/​product/​extraction/​rag_documents/​annotated_base_field.py	Adds annotation field registry.
mindee/​v2/​product/​extraction/​rag_documents/​__init__.py	Exports RAG document types.
mindee/​v2/​product/​extraction/​params/​data_schema.py	Uses shared serialization types.
mindee/​v2/​product/​extraction/​params/​data_schema_replace.py	Updates serializer import.
mindee/​v2/​product/​extraction/​params/​data_schema_field.py	Updates serializer import.
mindee/​v2/​product/​extraction/​params/​__init__.py	Removes serializer export.
mindee/​v2/​product/​extraction/​extraction_result.py	Uses registry-based field parsing.
mindee/​v2/​product/​extraction/​extraction_response.py	Removes redundant slug method.
mindee/​v2/​product/​extraction/​__init__.py	Removes serializer export.
mindee/​v2/​parsing/​string_data_class.py	Produces compact JSON.
mindee/​v2/​parsing/​search/​base_search_response.py	Centralizes search response typing.
mindee/​v2/​parsing/​inference/​field/​simple_field.py	Registers simple fields.
mindee/​v2/​parsing/​inference/​field/​object_field.py	Registers and recursively parses objects.
mindee/​v2/​parsing/​inference/​field/​list_field.py	Registers and parses list items.
mindee/​v2/​parsing/​inference/​field/​inference_fields.py	Uses the field registry.
mindee/​v2/​parsing/​inference/​field/​factory.py	Removes the legacy field factory.
mindee/​v2/​parsing/​inference/​field/​base_field.py	Implements field registration and construction.
mindee/​v2/​parsing/​inference/​base_inference_response.py	Renames slug access and adds typing.
mindee/​v2/​parsing/​base_rag_annotation_response.py	Adds the base RAG response model.
mindee/​v2/​mindee_http/​mindee_api_v2.py	Adds RAG HTTP operations and caller helpers.
mindee/​v2/​error/​mindee_http_error_v2.py	Logs structured HTTP errors.
mindee/​v2/​client.py	Exposes RAG lifecycle and polling APIs.
mindee/​v2/​client_options/​base_search_parameters.py	Reuses shared search response typing.
mindee/​v2/​client_options/​base_rag_document_upload_parameters.py	Adds base upload options.
mindee/​v2/​client_options/​base_product_parameters.py	Clarifies file-closing documentation.
mindee/​v2/​client_options/​base_annotation_parameters.py	Adds base annotation options.

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.