Simple script to scrap Glassdoor job listings.
First the html_scraper.py scraps the first 30 pages of every search term defined
in the config.yml file for every country. In this step we look for the job id
of each position and append them to a jobs_ids.txt file.
Having the jobs IDs of our interest we start scrapping the actual information of
each listing. The Glassdoor API returns a json file for each listing. We then collect
the information in blocks of 400 listings and save the json files to the results folder.
I uploaded the data manually to a S3 bucket, where I crawled the data using a Glue Crawler.
It was latter transformed to csv using a Glue Job with the script provided in
glue-job-script.py. No automation on this step though, because I ran it only once.
At the time I scrapped it, there were about 160k listings for the terms I've searched for. The data will be uploaded to a Kaggle Dataset.
I don't have any connection with Glassdoor and this project is neither approved or endorsed by them. The data collected with this script was publicly accessible at the moment it was collected. This script was created for educational purposes.