docling-project / docling-project/docling
[Feature] Support defining reading order from the layout stage
Open
enhancement
- Dominant language
- Python
- Stars
- 66.4k
- Forks
- 4.8k
- Avg merge
- 2d 21h
- Merged PRs (30d)
- 84
Description
### Requested feature
Some layout models, such as [PP-DocLayoutV3](https://huggingface.co/PaddlePaddle/PP-DocLayoutV3), include reading order in their output. Ideally, Docling could support using that order instead of computing it again later in the reading order stage.
Related: #3224, #3233
Contributor guide
Research direction
Start by reviewing related issues #3224 and #3233, then trace the layout stage and the later reading order stage. Done means a layout model's supplied reading order can be used instead of recomputing it, while existing behavior remains supported.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- ai
- Issue type
- Feature
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Active
- Clarity
- Mostly clear
- Newbie friendliness
- 48/100