# Changelog of Naukrigulf Scraper — Gulf Jobs, Salaries, Companies & Contacts (`logiover/naukrigulf-scraper`) Actor

- **URL**: https://apify.com/logiover/naukrigulf-scraper/changelog.md
- **Full Actor documentation**: https://apify.com/logiover/naukrigulf-scraper.md

## Changelog

### 2026-10-20

- New columns `contactEmail`, `contactEmails` and `contactPhone`: the hiring contact's e-mail address(es) from Naukrigulf's job data (including addresses the job page only reveals through its apply flow) and any written into the job text — on 26.8% of 600 recent UAE/Saudi/Qatar jobs — plus the contact's phone where given.

### 1.0 (2026-10-18)

- First release.
- Naukrigulf's search interface, 50 jobs per request, every row with the full job description (text and HTML).
- Relevance order for keyword searches (Naukrigulf's search is semantic — closest matches first); searches without a keyword and complete reads are paged newest-first, which is stable on Naukrigulf (relevance paging repeated/skipped 161 of 6,447 jobs in a measured full read; date paging 0).
- Automatic splitting past Naukrigulf's 45,000-results-per-search limit (employer type and gender requirement as exact partitions, experience bands, then country / industry / functional area).
- Job-detail records by default: monthly salary with currency and US-dollar conversion, industry, functional area, employment type, workplace (on-site / remote / hybrid), education, gender and nationality requirement, desired-candidate profile, company profile, public contact block (name, role, address, website), external apply link. Hidden e-mail addresses are not exported.
- Company search by name, Naukrigulf company page or company id; Naukrigulf search-page and job-page URLs.
- Filters: posted within, years of experience, experience bands, countries, industries (69), functional areas (59), salary bands, employer type, Easy Apply.
- Direct connection for small runs; detail-heavy runs read job details through Apify datacenter proxies (a fresh IP per call), with per-call timeouts, route switching and hedged calls for stragglers.
- De-duplication by job id across all searches; clean stop at the run's maximum cost per run (SUMMARY `stoppedBy: MAX_COST`); plain-language outcomes (COMPLETED / PARTIAL / EMPTY_RESULT succeed, BLOCKED / FETCH_FAILED / BAD_INPUT fail) and a SUMMARY record.