v26.6.2

flydocs 26.6.2 — Apache 2.0 open-source release

flydocs is now open source under the Apache License 2.0 (© 2024-2026 Firefly Software Foundation); the repository is public.

Changed

• Added the full Apache 2.0 LICENSE (root + Python/Java SDKs) and an Apache 2.0 header to every source file.
• Declared org.opencontainers.image.licenses=Apache-2.0 on the published image; surfaced the license in the README badge and OpenAPI metadata.
• Aligned all version references to 26.6.2.

Removed

• Removed the bundled flydocs-whitepaper.pdf from the repository and its git history.

SDKs

• Python: wheel + sdist attached below.
• Java: com.firefly:flydocs-sdk:26.6.2 on GitHub Packages.

v26.6.1

Python SDK

Install the wheel for this release with uv:

uv add https://github.com/firefly-operationOS/flydocs/releases/download/v26.6.1/flydocs_sdk-26.6.1-py3-none-any.whl

Or pin it in your pyproject.toml:

[project]
dependencies = ["flydocs-sdk"]

[tool.uv.sources]
flydocs-sdk = { url = "https://github.com/firefly-operationOS/flydocs/releases/download/v26.6.1/flydocs_sdk-26.6.1-py3-none-any.whl" }

Full Changelog: v26.6.0...v26.6.1

v26.6.0

Full v1 contract redesign — snake_case end-to-end, single recursive Field, unified Extraction lifecycle + EventEnvelope, RFC 7807 errors. See CHANGELOG.md and docs/migration-v0-to-v1.md for the full rename table and worked examples.

Python SDK

Install the wheel for this release with uv:

uv add https://github.com/firefly-operationOS/flydocs/releases/download/v26.6.0/flydocs_sdk-26.6.0-py3-none-any.whl

Or pin it in your pyproject.toml:

[project]
dependencies = ["flydocs-sdk"]

[tool.uv.sources]
flydocs-sdk = { url = "https://github.com/firefly-operationOS/flydocs/releases/download/v26.6.0/flydocs_sdk-26.6.0-py3-none-any.whl" }

What's Changed

• feat(api): v1 contract redesign + Python/Java SDKs + docs by @ancongui in #28

Full Changelog: v26.05.02...v26.6.0

v26.05.02

Python SDK

Install the wheel for this release with uv:

uv add https://github.com/firefly-operationOS/flydocs/releases/download/v26.05.02/flydocs_sdk-26.5.2-py3-none-any.whl

Or pin it in your pyproject.toml:

[project]
dependencies = ["flydocs-sdk"]

[tool.uv.sources]
flydocs-sdk = { url = "https://github.com/firefly-operationOS/flydocs/releases/download/v26.05.02/flydocs_sdk-26.5.2-py3-none-any.whl" }

What's Changed

• docs: top-level QUICKSTART + payload-reference, mirrored across SDKs by @ancongui in #22
• fix: atomic state transitions + race-loser handling for concurrent workers by @ancongui in #23
• fix: periodic reaper revives orphan jobs left stuck by lost events by @ancongui in #24
• feat(sdk/java): Spring Boot starter, reactive hardening, examples, live integration by @ancongui in #25
• chore: clean up ruff findings from #23 / #24 / #25 by @ancongui in #26
• docs: concurrency model + SDK 26.05.02 + new env settings by @ancongui in #27

Full Changelog: v26.05.01...v26.05.02

v26.05.01

Python SDK

Install the wheel for this release with uv:

uv add https://github.com/firefly-operationOS/flydocs/releases/download/v26.05.01/flydocs_sdk-26.5.1-py3-none-any.whl

Or pin it in your pyproject.toml:

[project]
dependencies = ["flydocs-sdk"]

[tool.uv.sources]
flydocs-sdk = { url = "https://github.com/firefly-operationOS/flydocs/releases/download/v26.05.01/flydocs_sdk-26.5.1-py3-none-any.whl" }

What's Changed

• feat(eda): migrate async jobs to pyfly EDA (Postgres outbox by default) by @ancongui in #1
• feat(ci): GitHub Actions PR gate + multi-arch GHCR publish by @ancongui in #2
• fix(tests): skip pricing test when agentic cost module is absent by @ancongui in #3
• chore(typing): silence the pyright advisory output (31 → 0 errors) by @ancongui in #4
• fix(observability): use genai_prices_cost API for outbound cost logging by @ancongui in #5
• feat(binary): any-binary normalization stage with distroless-friendly Office conversion by @ancongui in #6
• feat(bbox): grounded bbox refinement (sync, multilingual, PyMuPDF + pluggable OCR) by @ancongui in #7
• feat(bbox): async out-of-band refinement via second-stage EDA worker by @ancongui in #8
• feat(bbox): TesseractOcrEngine + PageWords.source so OCR-grounded boxes surface as source=ocr by @ancongui in #9
• feat: bastanteo-driven improvements (multi-file async, hybrid matcher, job filtering) by @ancongui in #10
• feat(transform): post-extraction transformation stage + docs refresh by @ancongui in #11
• refactor: drop legacy single-file shape + provider-agnostic / fireflyframework-* docs sweep by @ancongui in #12
• docs: complete the public-API reference + add docs/README.md index by @ancongui in #13
• feat(bbox): Docling-backed OCR engine + Markdown text-anchor by @ancongui in #14
• docs(prompt-cache): drop "Anthropic-only" framing; cover OpenAI + Google by @ancongui in #15
• fix(ci): docker-publish docling variant -- free disk + drop arm64 by @ancongui in #16
• docs(readme): complete the public API surface table by @ancongui in #17
• chore: rename flydesk-idp to flydocs by @ancongui in #18
• docs: platform-agnostic framing + skip CI on docs-only changes by @ancongui in #19
• feat(sdks): official Python + Java/Spring Boot SDKs for flydocs by @ancongui in #20
• chore: switch to CalVer YY.MM.PP (26.05.01) + async-first SDK docs/examples by @ancongui in #21

New Contributors

• @ancongui made their first contribution in #1

Full Changelog: https://github.com/firefly-operationOS/flydocs/commits/v26.05.01