Skip to content

Add Docling parser adapter (table extraction) #4

Description

@kiyeonjeon21

What

Docling (IBM) has 97.9% table extraction accuracy. The adapter exists at src/parsemux/parsers/docling.py but is untested with real Docling installation.

How

  1. pip install parsemux[docling]
  2. Test with table-heavy PDFs
  3. Verify DoclingParser.is_available() works
  4. Add integration tests

Context

Docling is especially valuable for financial reports, invoices, and academic papers with complex tables.

Metadata

Metadata

Assignees

No one assigned

    Labels

    featureNew feature requesthelp wantedExtra attention is neededparserParser adapter related

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions