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.Utf8Only
Besides speed fast-tagsoup correctly handles HTML <script> and <style> tags and converts tags to lower case. This fork purposefully removes support for parsing non-utf8 documents, to avoid dependency on text-icu. If you need to handle other encodings, refer to the original http://hackage.haskell.org/package/fast-tagsoup
This parser is used in production in BazQux Reader feeds and comments crawler.