# Changelog of Wongnai Scraper - Thailand Restaurants & Businesses (`haketa/wongnai-scraper`) Actor

- **URL**: https://apify.com/haketa/wongnai-scraper/changelog.md
- **Full Actor documentation**: https://apify.com/haketa/wongnai-scraper.md

## Changelog

### 0.1.8 — 2026-11-01

- **Fixed broken `photo` URLs.** The `photo` field previously emitted a scheme-less relative path (`photos/<hash>`) for every record, which did not load as an image. The actor now builds the full CDN URL from the embedded `contentUrl` (`https://img.wongnai.com/p/1920x0/<date>/<hash>.jpg`), so `photo` is a working, directly loadable image link for 100% of listings.
- **Fixed scheme-less `website` values.** Homepages stored without a protocol (e.g. `www.example.com`) are now normalized to absolute `https://` URLs so they are clickable.
- **Verified a fresh run of 400 unique Wongnai businesses** across 10 broad cuisine keywords in ~220s with **0 duplicates** (deduped by business `id`). Core-field coverage: `name`, `url`, `id`, `rating`, `category`, `lat`/`lng`, `province`, `district`, `address` all **100%**; `phone` **98.8%**, `priceRange` **97.2%**, `openingHours` **97.8%**. Lead fields are data-dependent: `email` 59%, `facebook` 57.5%, `instagram` 43.2%, `line` 36%, `website` 8.8%.
- **Refreshed the README** to the current field set: corrected the `photo` output example to the real `img.wongnai.com` CDN format, updated the field-coverage notes to match the verified run, and tightened the pagination/scale section.