feat: Measure ad-block rate via control domain init signal

[mattbodle]

Summary

• Fire two hidden iframes from the mParticle kit during launcher initialization to measure ad-block rates by comparing signal counts server-side
• Signal 1: existing domain (e.g. apps.rokt.com) — may be blocked by ad blockers
• Signal 2: apps.roktecommerce.com (control domain) — not on any block lists
• Signals fire from initRoktLauncher (after launcher is attached), ensuring window.__rokt_li_guid__ is reliably available
• 10% sampling rate to limit traffic volume
• pageUrl is stripped of both query params and hash fragments to avoid leaking sensitive state
• Gated to a single allowed hostname via djb2 hash check (no plaintext domain in source)

Changes

• src/Rokt-Kit.js: Add createAutoRemovedIframe helper, sendAdBlockMeasurementSignals with host hash gating, djb2 hash function, call from initRoktLauncher, expose via testHelpers
• test/src/tests.js: 10 new tests covering iframe creation, auto-removal, sampling, GUID gating, hostname gating, fragment stripping, and integration-level init behavior

Companion PR

ecommerce-api changes (new wsdkInitControlStart metric, isControl query param): https://github.com/ROKT/ecommerce-api/pull/2377

Test plan

• All 120 tests pass (npm test with ChromeHeadless)
• npm run lint passes clean
• After deployment, query sdk.TimingEvents Kafka topic for wsdkInitStart vs wsdkInitControlStart event counts
• Verify formula: ad_block_rate = 1 - (wsdkInitStart_count / wsdkInitControlStart_count)