Re:State of the Haskell ecosystem
by PaulAJ on Monday August 17, 2015 @12:41PM
(#10074255)
I haven't used tagsoup, but I believe that is the point; it was designed to let you scrape data from badly formed HTML you got from somewhere else, rather than helpfully pointing out the broken tags.
[ Parent
| on HN ]
The HTML spec defines how to parse any arbitrary stream of characters, and it is what is implemented in browsers and hence is what best supports badly-formed HTML (because it's typically written aimed at browsers!). Therefore, by not following the spec, TagSoup has worse compatibility with badly-formed HTML.