SPB Git forge
spboucher.ai ↗

spb/rareindex

Public
54commits 1branches 0releases
7.1 MBsize
maindefault branch
10 days agolast push
TypeScript 61.9% HTML 37.2% SQL 0.7%
1.4 KB
View source Raw Blame History

# ComicLink connector (comiclink)

  • Source: https://www.comiclink.com · type: auction_house · country: US · currency: USD
  • Acquisition: Scrapfly without JS rendering (~1 credit/page) — every page is behind a Cloudflare managed challenge for non-browser clients
  • Records: sale (auction results) at MONTH precision

# Behaviour

  • Discovery: /auctions/auctionschedule.asp → past sessions since 2015 (/results/generate_highlights.asp?output=<code>), newest first; art-only sessions skipped unless includeArtSessions.
  • Each results page lists ~300–800 sold books ("TITLE #N / SOLD in CGC 9.4 NM / $1,365" + cover + item link). The closing month comes from the schedule label ("Spring Featured: Comics (5-6/26)") → saleDate = 1st of that month, metadata.sale_date_precision = "month", confidence 0.7. Sessions without a month are skipped (never invented).
  • Item pages hide prices behind login → not fetched. buyerPremiumIncluded = null (not stated on the pages).
  • Incremental: 2 newest sessions per weekly run; backfill: cursor {index} over all sessions (~90 pages).

# Fixtures / tests

data/fixtures/comiclink/results-*.json (live capture, items trimmed to 40, snapshot to 6 blocks) + schedule-trimmed.html (real slice of the schedule page). pnpm vitest run connectors/scrapfly/comiclink. Capture: pnpm tsx connectors/scrapfly/comiclink/_capture.ts [sessionCode].