I've issued new releases of the text and text-icu packages, the fast, comprehensive Unicode text manipulation libraries.

Features of text:

• Compact array-based data representation.
• Library code based on stream fusion automatically reduces copying in your application, compared to a more naive library.
• The lazy text type supports streaming of data much larger than RAM.

Features of text-icu:

• Locale sensitive case mapping.
• [NEW] Language sensitive text boundary analysis.
• Text normalization. (What is normalization? http://unicode.org/faq/normalization.html)
• [NEW] Access to the Unicode Character Database.
• Locale sensitive string collation.
• Conversion to and from a huge number of native encodings.
• Efficient comparison of ByteString and Text.