Skip to content

List sources: parse real headlines, skip past events - #139

Merged
ralyodio merged 1 commit into
mainfrom
fix/list-sources-stale-events
Oct 7, 2026
Merged

ralyodio merged 1 commit into
mainfrom
fix/list-sources-stale-events

Conversation

@ralyodio

@ralyodio ralyodio commented Oct 7, 2026

Copy link
Copy Markdown
Contributor

Follow-up to #137 after testing against live Google News (3 ValueSERP credits):

  • Funding: 10 live headlines, only ~2 matched the strict pattern. The parser now finds the verb and walks back past prefixes ("Exclusive:"), descriptors ("Egyptian fintech startup", "Stablecoin payments firm") and appositives ("Latitude, founded by Stripe and Uber alums, raises"). All 7 real headlines in the tests now resolve to the right company.
  • Leadership: the OR-heavy query returned 0 results; <market> company appoints new chief returned 10. The parser takes Senior/Executive/Managing titles, , Inc. names, cuts titles at "and", and ignores board seats. Real cases (Solana Foundation, BNB Chain, Usio, Inc.) in tests.
  • Events: pages naming only past years ("PlatformCon 2024 speakers" in 2026) are skipped — one was crawled in prod on the first run.

Tests: list-sources domain 24/24, engine 2/2.

🤖 Generated with Claude Code

Tested against live Google News: most funding headlines do not start with
the company ('Exclusive: Split Pay raises...', 'Egyptian fintech startup
Paymob raises...', 'Latitude, founded by..., raises...'), so the parser now
finds the verb and walks back past descriptors and appositives. The
leadership query with ORs returned nothing; it is now '<market> company
appoints new chief', and the parser takes Senior/Executive titles, ', Inc.'
names, and cuts at 'and'. Board seats are not decision-makers. Event pages
that name only past years (PlatformCon 2024 in 2026) are skipped.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@ralyodio
ralyodio merged commit 886a38e into main Oct 7, 2026
4 checks passed
@ralyodio ralyodio mentioned this pull request Oct 7, 2026
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