[erlang-questions] The state of Unicode libraries

[Thomas Allen]

Hello,

I am working on a project right now for which Unicode string and binary
functions are critical. I find myself adrift in my search for a suitable
library to handle this.

These are the projects I have reviewed this far:

i18n <https://github.com/erlang-unicode/i18n>:

     Uses NIFs, linking with libicu. Promising, but this project has a
     major issue: It depends (in its rebar.config) on some testing
     library that uses pmods, so I cannot provide an easy install for R16
     users if I depend on this library. Author states in the issue queue
     that he has not used it in production himself, so I'm interested in
     how this has worked for other users in production.

ux <https://github.com/erlang-unicode/ux>:

     Pure Erlang solution. This is what I am using right now. It does the
     job, but takes about two seconds to start up (an annoyance) and has
     some strange history, like when `ux_html' was removed but not
     mentioned in the commit message:

 
<https://github.com/erlang-unicode/ux/commit/73e5e7ace9b1c745b29e2a500a9e3d937854bfde>

     A year later now, and `ux_html' functions are still in the README.

starling <https://code.google.com/p/starling/>:

     Something of a toy, and untouched since early 2008. No
     to_upper/to_lower or other important functions.

I hope that I am missing some viable libraries, because none of these
seem satisfactory right now, though `ux' is functional.

Thanks,
Thomas Allen