Skip to content

docs: recommend several example URLs of one page type for prepare_domain() - #68

Draft
zenrows-dev-pool[bot] wants to merge 2 commits into
mainfrom
feat/extract-example-urls-guidance
Draft

zenrows-dev-pool[bot] wants to merge 2 commits into
mainfrom
feat/extract-example-urls-guidance

Conversation

@zenrows-dev-pool

@zenrows-dev-pool zenrows-dev-pool Bot commented Oct 21, 2026 •

Copy link
Copy Markdown
Contributor

What

Docs only: the prepare_domain() docstring and the README's "Preparing a domain" section now steer callers to send several example URLs of the same page type.

  • example_urls takes 1 to 5 example URLs. One works.
  • We recommend 3 or more of the same page type (e.g. 3 product pages). Several page types can share the 5 (e.g. 3 product pages + 2 search-result pages).
  • The README example now passes three product pages from the same domain instead of one listing URL.
  • The old wording "one per kind of page you want to extract" is gone, since it told users to send one URL per type.

Why

Extract compares pages of one type to check the fields it found and fill in fields one page is missing.

Not changed

No behaviour change: example_urls still accepts 1 to 5 URLs, and the client does not reject or warn on 1 or 2 URLs.

Checks

  • make check (ty, ruff check, ruff format --check): pass
  • make test: 304 passed, 1 deselected (e2e)

…ain()

One example URL still works, but Extract compares the pages of one page
type against each other, so several related URLs (at least 3 of a type,
up to 5) give more complete, more reliable results. Replace the old
"one per kind of page" wording, which suggested the opposite, and show
three product pages in the README example. No behaviour change.
…wording

Say "we recommend 3 or more of the same page type" instead of "at least 3
pages of each page type", in the prepare_domain() docstring and the README's
"Preparing a domain" section. One URL still works; no behaviour change.

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.

0 participants