Align DataFrame boilerplate in Python and Scala

[lintool]

Currently, the Scala DF boilerplate is something like:

RecordLoader.loadArchives("src/test/resources/warc/example.warc.gz", sc).extractValidPagesDF()
  .select($"Url")
  .show(20, false)

And Python:

WebArchive(sc, sqlContext, "src/test/resources/warc/example.warc.gz").pages() \
    .select("url") \
    .show(20, False)

It would make sense to align?

So Python would look like:

RecordLoader.loadArchives("src/test/resources/warc/example.warc.gz", sc, sqlContext).pages() \
    .select("url") \
    .show(20, False)

And from the Scala end, let's just change extractValidPagesDF() to pages() to match the Python end? This runs the slight risk of confusion with RDD operations, but I think the risk is minimal.

@ianmilligan1 @ruebot thoughts?

[ruebot]

That was the general consensus in #231

[lintool]

For references, where the current DF extractors are defined:
https://github.com/archivesunleashed/aut/blob/master/src/main/scala/io/archivesunleashed/package.scala#L102

And the Python connectors:
https://github.com/archivesunleashed/aut/blob/master/src/main/python/aut/common.py