KingfisherFilesStore: Write scrapyd-job.txt

[jpmckinney]

Presently, it's a little complicated to reconcile data directories with log files, especially if multiple crawls are started for the same source in rapid succession. This is relevant to Kingfisher Archive.

If we write the spider's job ID to a scrapyd-job.txt in the crawl directory, this can be much easier. A _job kwarg is passed to the spider's initializer by Scrapyd.