Thanks to visit codestin.com
Credit goes to github.com

Skip to content

Include metadata in ingested document chunks #3192

@dmartinol

Description

@dmartinol

Is your feature request related to a problem? Please describe.
Current ingestion pipeline (ilab rag ingest ...) calculate document chunks to be embedded and indexed in the target document store: the stored documents do not include any metadata.

Describe the solution you'd like
Replicate the behavior of the docling-haystack package to include the metadata defining the chunk hierarchy and the basic information calculated by the Docling HybridChunker:
converter.py

Metadata

Metadata

Assignees

Labels

RAGRAG specific issuesenhancementNew feature or request

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions