Implement XML fetcher / PastPerfect mapper

[lthurston]

No description provided.

[lthurston]

@amywieliczka, if you want to take a sneak peek at this XML fetcher, I invite you to do so. It works, and I fetched collection 26935 of the reported 77k records in about 20 seconds locally. It reported there are actually more than 110k records though, so there might be an issue there, or maybe there's actually more records.

I haven't written any mapping code yet, so I consider this to be a little naive, a little optimistic, but nevertheless it does what it's supposed to do. Let me know your thoughts!

[aturner]

@lthurston I think our legacy harvester code has some logic built in to leave out "metadata only" records; the source collection has some items that don't have a digital image -- just metadata records only. That may account for the count difference that you're seeing

[lthurston]

@aturner That makes sense, thanks for the explanation. My instinct is to leave those records in our imported files in order to stay as true to the original source data as possible (despite the fact that we have to rewrite it to paginate), but am only too happy to be overruled.