Consolidate spending items scraping into this repo

Currently all scraping underlying the search functionality is not running on a regular basis and requires a lot of manual work to publish a new release.

- Create database & tables for scraped data
- Likely a core table with JSONB for source-specific data.
- Need to support:
- Global Affairs Canada

Can be split into two parts:
- Scraping infrastructure (database/cron jobs)
- Search Infrastructure (meillisearch, currently it is backed by typesense, but typesense is more expensive to host because it loads everything into memory)