indexing new features without crawling everything all over again

[ksonda]

In the medium term, we anticipate a scenario where organizations are adding new tranches of PIDs on a somewhat frequent, if irregular basis. (e.g. 1000 PIDS week1, 0 pids week2, 100 PIDS week3, etc.)

We need to come up with a way for gleaner to only need to crawl new/updated sitemap portions. Things to do:

1. Gleaner should be able to target sitemap entries according to lastmod >= yesterday or some such thing that triggers frequent checking. @fils
2. sitemap.xml should keep track of lastmod of the constituent sitemap links in a way that accurately reflects the last time PIDs were submitted to github.com/internetofwater/geoconnex.us @webb-ben