Fix for purge crawl failing if primary crawl comes up empty

[mattnowzari]

Closes #381

When the primary crawl finishes without indexing any documents into a freshly-created index (e.g. every page returns 4xx, robots.txt is forbidden, the seed URL is unreachable, etc.), the index has no field mappings.

The purge phase then issues a search sorted by last_crawled_at, and Elasticsearch responds with 400 query_shard_exception: No mapping found for [last_crawled_at] in order to sort on. The retry logic in ES::Client#execute_with_retry blindly re-issues the same bad query 4 times and then the entire crawl fails.

This PR fixes this by adding unmapped_type: 'date' to the sort clause in fetch_purge_docs.

Per the official Elasticsearch sort docs, this is the canonical idiom for telling ES "treat this field as a date if it has no mapping yet." With the fix, the search succeeds against an empty/unmapped index, returns 0 hits, and run_purge_crawl! in Coordinator.rb takes over and finishes the crawl successfully.

Checklists

Pre-Review Checklist

• This PR does NOT contain credentials of any kind, such as API keys or username/passwords (double check crawler.yml.example and elasticsearch.yml.example)
• This PR has a meaningful title
• This PR links to all relevant GitHub issues that it fixes or partially addresses
  • If there is no GitHub issue, please create it. Each PR should have a link to an issue
• this PR has a thorough description
• Covered the changes with automated tests
• Tested the changes locally
• Added a label for each target release version (example: v0.1.0)
• Considered corresponding documentation changes
• Contributed any configuration settings changes to the configuration reference
• Ran make notice if any dependencies have been added

Release Note

Fixed a bug where a crawl would fail at the purge phase if the primary crawl didn't index any documents into a newly-created index.

[artem-shelkovnikov]

🚀 🌖