Skip to content

docs: explain source verification in schema extraction tutorial - #1667

Open
roli-lpci wants to merge 1 commit into
simonw:mainfrom
roli-lpci:docs/schema-extraction-evidence
Open

roli-lpci wants to merge 1 commit into
simonw:mainfrom
roli-lpci:docs/schema-extraction-evidence

Conversation

@roli-lpci

Copy link
Copy Markdown
Contributor

Closes #808.

The schema tutorial still shows an incorrect article date without explaining the error. This adds the warning requested in the issue, links the original investigation, and follows it with the strip-tags -t meta command that preserves publication-date metadata.

The explanation distinguishes a schema-conforming value from a source-supported fact and publication time from modification time. It explicitly asks readers to check the extraction after preserving metadata.

Validation:

  • Ran default strip-tags and strip-tags -t meta against a deterministic HTML fixture with publication and modification meta tags plus a placeholder visible date. Default output drops both date values; -t meta retains both.
  • uv run --python 3.11 --with-requirements docs/requirements.txt make -C docs html: build succeeded.
  • git diff --check: passed.
  • No new LLM extraction was run or presented as verified. The incorrect-date example and diagnosis come from the existing tutorial and its issue investigation.

Prepared with AI assistance.

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.

Add note to schemas tutorial about the date failure (and general unreliability of LLMs)

1 participant