A customized url normalizer for our article #97
Description
As discussed in issue #44 , many websites have url with inconsequential parameters, making them look like "different article" on our db while they are actually the same ones. This flooded our db with unnecessarily duplicated snapshots and article entries. Right now we have a StandardizePipeline that remove parameters in url after ? for appledaily, udn, chinatimes. However, for a better solution that works on other cases discussed in issue #44, we are looking to have a custom-made url normalizer to use in our pipeline.