normalization-insensitive is identical to case-insensitive, except it is insensitive to Unicode normalization, rather than case. The module Data.Unicode.NormalizationInsensitive provides the NI type constructor which can be parameterized by a string-like type like: String, ByteString, Text, etc.. Comparisons of values of the resulting type will be insensitive to normalization. https://hackage.haskell.org/package/normalization-insensitive https://github.com/ppelleti/normalization-insensitive I created this to solve a problem I was running into comparing filenames on Mac OS X. Because Mac OS X converts all filenames to Normalization Form D, the filenames I read back from disk were different than the filenames I had written to disk. This package solved the problem very neatly. Under the hood, normalization-insensitive uses the unicode-transforms package, which Harendra Kumar announced a few days ago on haskell-cafe. This avoids having to pull in any heavyweight dependencies like ICU. Thanks to Bas van Dijk for writing case-insensitive, upon which normalization-insensitive is based. And thanks for Harendra Kumar for reviewing normalization-insensitive and providing feedback, as well as writing unicode-transforms. Although this is the first release of normalization-insensitive, the version number is 2.0 to avoid any confusion with the version numbers of case-insensitive. --Patrick