- there was a typo and another bug that passed some tests on BeautifulSoup parsing. they have been fixed. todo- migrate tests to public repo
0.6.13
- trying to integrate a "safe read"
0.6.12
- now passing "stream=True" to requests.get. this will fetch the headers first, before looping through the response. we can avoid many issues with this approach
0.6.11
- now correctly validating urls with ports. had to restructure a lot of the url validation
0.6.10
- changed how some nodes are inspected. this should lead to fewer errors
0.6.9
- added a new method `get_metadata_link()`, which applies link transformations to a metadata in an attempt to ensure a valid link