Simplified WARC Reader

[gthole]

Don't use Content-Length header to read content in.

Potential issues:

• It uses the "WARC/1.0" line stamp as a delimiter. If a record has "\nWARC/1.0" in the response body somewhere, this could cause issues. That seems like an outlandish outlier, though, in comparison with the grief we've had with Content-Length reads.
• Interface isn't as pretty. Have to:

from warc.warc import SimpleWARCReader
reader = SimpleWARCReader(open('path/to/file.warc', 'r'))

• Can't read gzipped files (but then the original library can't either, not really.)