Standardize granicus.site.Site.scrape

[zstumgoren]

Site.scrape method should be updated to match the features supported in the scrape methods for CivicPlus and Legistar Site classes. Specifically, we should attempt to implement the following features:

• File download should be set to False by default
• date-based filtering, with default set to current day
• file type filtering for downloads, with Agendas and Minutes set to default
• file size filtering for downloads

We should also attempt to implement the cache flag as in CivicPlus, although we didn't do so for Legistar since the underlying python-legistar framework would need to be refactored to support caching of source HTML.