Skip to content

fix: correct and complete the langchain-decodo README - #22

Open
pauliusminialga-dc wants to merge 1 commit into
masterfrom
docs/readme-accuracy
Open

pauliusminialga-dc wants to merge 1 commit into
masterfrom
docs/readme-accuracy

Conversation

@pauliusminialga-dc

Copy link
Copy Markdown

What

Documentation-only corrections to libs/partners/decodo/README.md. No source changes, so nothing about the package's behaviour moves.

This is the text PyPI renders on the langchain-decodo project page, so every inaccuracy in it is user-facing on every install.

Changes

Reddit removed from the DecodoSearchTool section — the prose, the example call and the engines table row. Reddit is not documented on any public Decodo surface, and the equivalent cleanup has already landed on the Mintlify pages and the TypeScript package.

Note this is a README change only. _ENGINE_TARGET_MAP in tools.py is untouched, so engine="reddit" still works exactly as before for anyone already passing it. Whether the source should change too is a separate decision.

DecodoLoader's constructor arguments documented. The section listed none of them. Added urls, api_token, auth_mode, timeout and continue_on_error, plus the task_id, created_at and updated_at metadata the loader carries through, and what continue_on_error actually does to a failed URL (empty page_content, error in metadata["error"], which you want to filter out before embedding).

.run() replaced with .invoke() in the two scrape-tool examples. The four newer tools already used .invoke(), so the README was demonstrating two different APIs, one of them deprecated.

Two DecodoSearchTool behaviours documented that currently surprise people, along with num_results itself:

  • engine= on the constructor is a no-op. _run takes engine from the args schema, which defaults to google, so DecodoSearchTool(engine="amazon") still searches Google. Pass engine per call.
  • num_results does not cap the results returned. It is sent as limit, and a Google search with num_results=3 still comes back with the full first page of organic results.

Both were already documented on help.decodo.com; this brings the README in line.

Why fix: and not docs:

Deliberate. python-semantic-release is configured in pyproject.toml with patch_tags = ["fix", "perf", "hotfix"], so a docs: title cuts no release. PyPI serves the README from the last published version, which means the 0.2.0 text — Reddit included — would stay on the project page indefinitely.

fix: cuts 0.2.1 and republishes the corrected README. If you would rather not cut a release for this, retitle to docs: before merging and the text will land in the repo but not on PyPI.

Verification

  • Every documented argument, default and limit checked against langchain_decodo at master: loader defaults timeout=180 and continue_on_error=True, num_results 1–100 default 10, the engine map, and the _run signature that makes the constructor engine a no-op.
  • grep -ic reddit on the pushed file returns 0; grep -c '\.run(' returns 0.

The README is what PyPI serves on the project page, so these inaccuracies are
user-facing on every install.

- Removed Reddit from the DecodoSearchTool section: the prose, the example and
  the engines table. Reddit is not documented on any public Decodo surface.
- Documented DecodoLoader's constructor arguments, which were missing
  entirely: api_token, auth_mode, timeout and continue_on_error. Added the
  task_id/created_at/updated_at metadata and what continue_on_error does to a
  failed URL.
- Replaced the deprecated .run() with .invoke() in the two scrape-tool
  examples, so the whole README uses one API.
- Documented num_results, and the two DecodoSearchTool behaviours that
  surprise people: engine= on the constructor is a no-op, and num_results does
  not cap the results the API returns.

Documentation only; no source changes, so DecodoSearchTool still accepts every
engine value it accepted before.

Typed `fix:` rather than `docs:` deliberately. python-semantic-release lists
only fix, perf and hotfix in patch_tags, so a `docs:` title cuts no release and
the README on PyPI would stay at the 0.2.0 text indefinitely.

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant