Imr-u/jobs.et

A collection data for job posts in Ethiopia. # Afriworket Job Scraper Automated scraper for afriworket.com that runs daily via GitHub Actions and saves results as a Parquet file. --- ## What it scrapes Each run collects up to **100 unique job listings**, visiting every detail page for the full data set: | Field | Source | |---|---| | `title` | Detail page h1 | | `company` | Listing card | | `location` | Listing card | | `industry` | Detail page (e.g. Media & Entertainment) | | `posted_relative` | Listing card (e.g. "3 hours ago") | | `posted_date` | Detail page (e.g. "March 16, 2026") | | `deadline` | Both | | `job_type` | Listing card (e.g. Onsite - Full Time) | | `experience_level` | Detail page (Expert / Senior / Intermediate / Junior) | | `vacancies` | Detail page | | `education_qualification` | Detail page | | `applicants_needed` | Detail page (Male / Female / Both) | | `salary` | Detail page (numeric ETB) | | `salary_period` | Detail page (Monthly / etc.) | | `skills` | Detail page (comma-separated) | | `work_address` | Detail page | | `full_description` | Detail page | | `company_profile_url` | Detail page | | `company_jobs_posted` | Detail page | | `detail_url` | Detail page URL | | `scraped_at` | UTC timestamp of the scrape | **Deduplication** is performed on `(title, company, deadline)` — both during scraping and as a final safety pass before saving. --- ## Setup ### 1. Fork / clone this repo ```bash git clone github.com cd afriworket-scraper ``` ### 2. Enable GitHub Actions Go to **Actions** tab in your repo → click **"I understand my workflows, go ahead and enable them"**. ### 3. Give Actions write permission Go to **Settings → Actions → General → Workflow permissions** → select **"Read and write permissions"** → Save. ### 4. Run manually (optional) Go to **Actions → Scrape Afriworket Jobs → Run workflow**. The parquet file will be committed to `data/jobs.parquet` and also available as a downloadable artifact. --- ## Run locally …