refactor: drop legacy single-file shape + provider-agnostic / fireflyframework-* docs sweep#12
Merged
Merged
Conversation
…ramework-* docs sweep
The HTTP surface now exposes a single shape -- `documents` is always a
non-empty list (one entry = single file), removing the `document`
single-file branch from DTOs, the orchestrator, the submit/job worker,
the bbox-refine worker, and the controller's size-limit check. The
`schema_json` written by `SubmitJobHandler` is now a single shape too:
always `{ documents: [...] }`, no more `document_content_base64`
fallback. Every test was updated to construct requests with
`documents=[DocumentInput(...)]`.
Docs sweep alongside the refactor:
- Rename bare `pyfly` -> `fireflyframework-pyfly` and bare `agentic`
-> `fireflyframework-agentic` in prose across README, overview,
architecture, pipeline, deployment, cicd, prompts, troubleshooting,
and api-reference. Python module paths (`pyfly.eda.*`) are kept
since they are the actual import names.
- Make provider mentions provider-agnostic. `genai-prices` is now
described as the multi-provider pricing source (Anthropic / OpenAI
/ Google / Mistral), not "Anthropic tariffs". Prompt caching is
marked explicitly Anthropic-specific (no-op on other providers).
`outbound_call` log-line target enumeration now lists all
providers `fireflyframework-genai` can resolve.
- Rewrite the README quickstart end-to-end: prerequisites, install,
provider-agnostic env setup, db + migrations, API + worker boot,
first sync extraction, first async multi-file job with a
transformation + webhook, docker stack alternative, and the test
surface. Updates the "what you get back" section to drop the
"single-file only on async" caveat.
ancongui
added a commit
that referenced
this pull request
May 31, 2026
…ramework-* docs sweep (#12) The HTTP surface now exposes a single shape -- `documents` is always a non-empty list (one entry = single file), removing the `document` single-file branch from DTOs, the orchestrator, the submit/job worker, the bbox-refine worker, and the controller's size-limit check. The `schema_json` written by `SubmitJobHandler` is now a single shape too: always `{ documents: [...] }`, no more `document_content_base64` fallback. Every test was updated to construct requests with `documents=[DocumentInput(...)]`. Docs sweep alongside the refactor: - Rename bare `pyfly` -> `fireflyframework-pyfly` and bare `agentic` -> `fireflyframework-agentic` in prose across README, overview, architecture, pipeline, deployment, cicd, prompts, troubleshooting, and api-reference. Python module paths (`pyfly.eda.*`) are kept since they are the actual import names. - Make provider mentions provider-agnostic. `genai-prices` is now described as the multi-provider pricing source (Anthropic / OpenAI / Google / Mistral), not "Anthropic tariffs". Prompt caching is marked explicitly Anthropic-specific (no-op on other providers). `outbound_call` log-line target enumeration now lists all providers `fireflyframework-genai` can resolve. - Rewrite the README quickstart end-to-end: prerequisites, install, provider-agnostic env setup, db + migrations, API + worker boot, first sync extraction, first async multi-file job with a transformation + webhook, docker stack alternative, and the test surface. Updates the "what you get back" section to drop the "single-file only on async" caveat. Co-authored-by: ancongui <andres.contreras@soon.es>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
documentsingle-file branch. The HTTP surface is now one shape only:documentsis a non-empty list (single file = one-element list). Cleaned up across DTOs (ExtractionRequest,SubmitJobRequest,ExtractionResult,DocumentInfo,ExtractedDocument), the orchestrator'sis_multi_fileheuristic, the submit-job handler'sschema_jsonshape (no moredocument_content_base64keys), the async job worker, the bbox-refine worker, and the controller's size-limit check. Every unit + LLM test was updated to construct requests withdocuments=[DocumentInput(...)].pyfly→fireflyframework-pyflyand bareagentic→fireflyframework-agenticin prose across the docs (Python module paths likepyfly.eda.*stay as-is — that is the import name).genai-pricesis now described as the multi-provider pricing source (Anthropic / OpenAI / Google / Mistral) and not "Anthropic tariffs". Anthropic prompt caching is explicitly marked Anthropic-specific (no-op on other providers).Test plan
task lint:check— ruff + ruff-format + pyright all clean (0 errors).task test— 237 passed, 1 skipped, < 6 s. Every previously-touched suite (request validator, submit-job handler, bbox-refine worker, judge escalator, real-LLM extraction) was rewritten for the new shape.