v2.139.0

docling-project/doclingv2.139.0Oct 14, 2026by docling-ops[bot]

Upgrading?

AI Summary

This release introduces conversion progress bars and event callbacks for better user feedback, enhances parsing capabilities for Numbers and PPTX files, and includes numerous performance optimizations. It also addresses various bug fixes related to image handling, HTML decoding, and VLM streaming reliability.

Key Highlights

  • Added conversion progress bars and an event callback to track processing status.
  • Enhanced support for reading pictures in Numbers table cells and plots in PPTX combination charts.
  • Improved VLM streaming request handling, including stop reason reporting and failure error handling.
  • Optimized performance by caching OpenDocument style lookups and moving detections to the CPU.

New Features

  • Expose max_concurrent_results on ThreadedDoclingParseBackendOptions
  • Expose per-page confidence breakdown in ConfidenceScores
  • Allow docling-tools to download whisper artifacts from artifactory
  • Read pictures in Numbers table cells
  • Read every plot of a combination chart in PPTX
  • Add conversion progress bars and an event callback

Full Release Notes

### Feature

* Expose max_concurrent_results on ThreadedDoclingParseBackendOptions ([#4804](https://github.com/docling-project/docling/issues/4804)) ([`ce5621b`](https://github.com/docling-project/docling/commit/ce5621b5e5569cfda1a010566247c93faa7d7032))
* Expose per-page confidence breakdown in ConfidenceScores ([#4233](https://github.com/docling-project/docling/issues/4233)) ([`5bdabb9`](https://github.com/docling-project/docling/commit/5bdabb917617d0ff434f676d537bda80cd19615f))
* **progress:** Add conversion progress bars and an event callback ([#4560](https://github.com/docling-project/docling/issues/4560)) ([`fdb6321`](https://github.com/docling-project/docling/commit/fdb6321046b4cdd75cde8ed804bf76dd1968b7cf))
* **iwork:** Read pictures in Numbers table cells ([#4736](https://github.com/docling-project/docling/issues/4736)) ([`dd477a9`](https://github.com/docling-project/docling/commit/dd477a9f5feb3c505caeb6e73931bf93cecc79f4))
* Allow docling-tools to download whisper artifacts and for docling to run ASR from an artifactory directory ([#4760](https://github.com/docling-project/docling/issues/4760)) ([`765dc86`](https://github.com/docling-project/docling/commit/765dc86182b56ac3ad5baba1a68bd330d11108cf))
* **pptx:** Read every plot of a combination chart ([#4773](https://github.com/docling-project/docling/issues/4773)) ([`070a6d3`](https://github.com/docling-project/docling/commit/070a6d390e5c94e279937b69babb02f23e0af62f))

### Fix

* **md:** Keep pictures through HTML conversion ([#4777](https://github.com/docling-project/docling/issues/4777)) ([`774387e`](https://github.com/docling-project/docling/commit/774387e7507e9be80b6f081bace13f2c5418719e))
* **vlm:** Report the stop reason of streaming requests ([#4778](https://github.com/docling-project/docling/issues/4778)) ([`fd346c7`](https://github.com/docling-project/docling/commit/fd346c7fc8884c59ee4f8e846fd6ffb588d34720))
* **docx:** Read the active branch of a body-level mc:AlternateContent ([#4615](https://github.com/docling-project/docling/issues/4615)) ([`b35ad57`](https://github.com/docling-project/docling/commit/b35ad57af3a9311d5da8c20ede17865c40b87870))
* **docx:** Keep table rows after a list resumes in a cell ([#4674](https://github.com/docling-project/docling/issues/4674)) ([`0f23355`](https://github.com/docling-project/docling/commit/0f233559b364d779d2ce945b32a44267755ee904))
* **odf:** Keep ODS pictures anchored to the sheet ([#4678](https://github.com/docling-project/docling/issues/4678)) ([`f3eda5f`](https://github.com/docling-project/docling/commit/f3eda5f3e9a32a3850c2d7c1f5b2796980b05171))
* **pptx:** Move footer placeholders to the furniture layer ([#4746](https://github.com/docling-project/docling/issues/4746)) ([`3acd5ca`](https://github.com/docling-project/docling/commit/3acd5ca3be5fb34f525e365acc1426baf140a780))
* **html:** Decode a rendered stream with the encoding the document declares ([#4468](https://github.com/docling-project/docling/issues/4468)) ([`85dcebc`](https://github.com/docling-project/docling/commit/85dcebc7726487193add4dd419e5b05619370774))
* **pptx:** Keep a hyperlink attached to a whole shape ([#4787](https://github.com/docling-project/docling/issues/4787)) ([`8508141`](https://github.com/docling-project/docling/commit/85081410b9f53d65eec244fce9e8ad738ba8d66f))
* **md:** Keep the hyperlink of autolinks outside table cells ([#4791](https://github.com/docling-project/docling/issues/4791)) ([`4757220`](https://github.com/docling-project/docling/commit/475722097f45f69c9c7fbde3ef5b30ef6479d157))
* **pypdfium2:** Keep letters of adjacent lines out of text cells ([#4802](https://github.com/docling-project/docling/issues/4802)) ([`006c098`](https://github.com/docling-project/docling/commit/006c098d322a1fd709ec53fd7e798ca6fe7fbcd1))
* **email:** Route an Outlook .msg that filetype reports as XLS to the email backend ([#4798](https://github.com/docling-project/docling/issues/4798)) ([`385ddd4`](https://github.com/docling-project/docling/commit/385ddd4fc2e448f88904c5f189ba24d478224fdd))
* **markdown:** Resolve relative images for local sources ([#4816](https://github.com/docling-project/docling/issues/4816)) ([`8119cdc`](https://github.com/docling-project/docling/commit/8119cdceef7494dc361b9d30b40458d287a06a34))
* **asciidoc:** Nest lists with different bullet markers ([#4822](https://github.com/docling-project/docling/issues/4822)) ([`f1d1725`](https://github.com/docling-project/docling/commit/f1d17255b649938c5f9d68ea2c6c5f64cca42b3d))
* **asciidoc:** Stop an unclosed block delimiter from hanging the parser ([#4662](https://github.com/docling-project/docling/issues/4662)) ([`2104df5`](https://github.com/docling-project/docling/commit/2104df53e9c49c59db16d2404e54432e4e0ff8dc))
* **odf:** Keep image references out of list item text ([#4693](https://github.com/docling-project/docling/issues/4693)) ([`4b8231f`](https://github.com/docling-project/docling/commit/4b8231f10cfb7747e28b8d3e64a66e4ebff02589))
* **vlm:** Report streaming request failures as per-page inference errors ([#4561](https://github.com/docling-project/docling/issues/4561)) ([`ab4419b`](https://github.com/docling-project/docling/commit/ab4419b8129ffd3dd90f5da85138e5d356f20d5b))
* **asciidoc:** Strengthen the parsing of nested lists ([#4721](https://github.com/docling-project/docling/issues/4721)) ([`c0be875`](https://github.com/docling-project/docling/commit/c0be875615f1c2fadb10bf20f760c7d3dc4b1059))
* **html:** Resolve the line break sentinel outside paragraphs ([#4675](https://github.com/docling-project/docling/issues/4675)) ([`7be1913`](https://github.com/docling-project/docling/commit/7be1913a2c0b93f8b756644fe682d0515d0b667f))
* **docx:** Resolve linked style numbering ([#4727](https://github.com/docling-project/docling/issues/4727)) ([`ffce191`](https://github.com/docling-project/docling/commit/ffce191795f9c576d73dfd544ed8587a42285bc2))
* **pdf:** Run threaded pipeline models on long-lived threads ([#4707](https://github.com/docling-project/docling/issues/4707)) ([`cb1df98`](https://github.com/docling-project/docling/commit/cb1df981e0b49906d638c652a27bfeeeb13ec5d1))
* **cli:** Keep convert-remote working on docling-client installs ([#4803](https://github.com/docling-project/docling/issues/4803)) ([`711fc47`](https://github.com/docling-project/docling/commit/711fc47c451d28597fdb445a593bf34a5231f5e1))
* **csv:** Rewind content before repeated dialect detection ([#4754](https://github.com/docling-project/docling/issues/4754)) ([`f0d9098`](https://github.com/docling-project/docling/commit/f0d90989b461656102da5d449db98e6be9285f93))
* **email:** Decode the HTML body once when a meta tag declares a charset ([#4756](https://github.com/docling-project/docling/issues/4756)) ([`ff2342c`](https://github.com/docling-project/docling/commit/ff2342c551aeccd6aedd37bdc608e39386b6904c))
* **asciidoc:** Keep a table whole across blank and wrapped lines ([#4761](https://github.com/docling-project/docling/issues/4761)) ([`78d1251`](https://github.com/docling-project/docling/commit/78d12518e21706a797e4db9b6ba7ee588fb6015e))
* **html:** Don't spread one run's formatting or link to a whole heading ([#4670](https://github.com/docling-project/docling/issues/4670)) ([`930bf20`](https://github.com/docling-project/docling/commit/930bf2041651ac12b3ebc89fb7e8093f87cbc20f))
* **html:** Parse the start of an ordered list by the HTML rules for integers ([#4713](https://github.com/docling-project/docling/issues/4713)) ([`a057e8d`](https://github.com/docling-project/docling/commit/a057e8d846ca27fac981c0188fc84cac33067b28))
* **ebcdic:** Preserve scaled numbers without context rounding ([#4759](https://github.com/docling-project/docling/issues/4759)) ([`3807f7f`](https://github.com/docling-project/docling/commit/3807f7f8a2e4fada02f0165cb112ad05eca91e56))

### Documentation

* **pdf:** Document parser thread configuration ([#4303](https://github.com/docling-project/docling/issues/4303)) ([`f951b83`](https://github.com/docling-project/docling/commit/f951b832d29a6f9e38d454f31e4743b9c56e7b1a))

### Performance

* **odf:** Cache get_style() lookups in the OpenDocument backend ([#4763](https://github.com/docling-project/docling/issues/4763)) ([`6329eea`](https://github.com/docling-project/docling/commit/6329eeaaf73064123415148a9cdb4d233211bdeb))
* **layout:** Move detections to the CPU once before building the output ([#4764](https://github.com/docling-project/docling/issues/4764)) ([`7e16edc`](https://github.com/docling-project/docling/commit/7e16edc966fb0e048d7ef6976ea7907d6b377a97))

Related Repositories