Improve URL boundary regarding commas (Old name: Wrong URL Pattern being Parsed)

[rahulv017]

Input : 'www.safenet.com/abc,False,False'
Output: urls:['www.safenet.com/abc,False,False']
Expected Result : urls: ['www.safenet.com/abc']

[fhightower]

Thanks for reporting! I'll get this fixed. I think this will be resolved as part of #254.

[fhightower]

Thanks again for raising this issue. Just want to post an update on this work:

I'll have to give this more thought because this technically (per the RFC) is not a bug. Commas are allowed to be part of a URL path, so parsing www.safenet.com/abc,False,False as an entire URL is correct.

That being said, this library should absolutely handle this case better b/c commas are not frequently used in URLs and are frequently used as delimiters in plain text. At the moment, I am inclined to do something similar to what I've done for email addresses where we have complete_email_addresses to capture email addresses per the spec and simple_email_addresses to capture email addresses as most people expect. I will likely take the same approach with URLs, but need to give this some thought.

In the mean time, if you have a CSV file with a column containing indicators, it's probably best to only send data from that column into this library (if possible) or to replace commas w/ spaces.

[fhightower]

I've renamed this issue to more generally reflect the root problem we need to solve.

[fhightower]

@rahulv017 - sorry for the delay on this, but I will focus on this ticket over the next week or so. I still need to give some thought to what the best approach is, but I think there is likely something we can do to improve this use-case.