Fast Haskell tagsoup parser.
Speeds of 20-200MB/sec were observed.
Works only with strict bytestrings.
This library is intended to be used in conjunction with the original tagsoup package:
import Text.HTML.TagSoup hiding (parseTags, renderTags) import Text.HTML.TagSoup.Fast
Besides speed fast-tagsoup correctly handles HTML <script> and <style> tags, converts tags to lower case and can dett non UTF-8 XML for you.
This parser is used in production in BazQux Reader feeds and comments crawler.
Use cabal install fast-tagsoup to get it.