url cleanup function

[schochastics]

wonky things in webtracking data

• protocolls see missing protocolls #93
• /#/ (hashtag in path)

[BernhardClemm]

Collecting some (potential) examples for "hashtag in path" to review:

• https://29nunder.com/21-summer-dresses#/page-4
• https://4tire.com/Catalog/tires/Manufacturer:6#u=/Manufacturer:6/Category:8/DisplayAll:0 |  
• https://6pm.com/dockers-clothing-on-sale~5#!/dockers-men-pants/CKvXARDK1wE6ArQaUgLsAcABAuABAeICBhwBChgCBw.zso
• https://arciform.com/#/image-galleries/residential-gallery/kitchens/1900-1919 |  

... looking at these cases (and there are a lot of them), it seems like this may be an undefinable problem, i.e., what looks to us like a "hashtag in the path" is not easy to delineate. Potentially may be better to just add a caveat to the adaR::ada_get_hash() description and leave the issue be...

There are also cases where the hash acts more like the query separator...

https://511ga.org/#zoom=4&lat=33.85303&lon=-84.42153

Browsing data are messy!