Unicode::ICU - ICU in Perl
This library is a binding to ICU, a library for internationalization (i18n), localization (l10n), Unicode, and all kinds of related stuff.
We currently only expose a subset of ICU’s (quite vast!) functionality. More can be added as need arises.
Most functionality exists in submodules under this namespace. Their names correspond roughly with modules or classes in ICU’s C and C++ APIs:
-
NB: This exposes a lot of other ICU functionality like formatting of numbers, dates/times, and plurals.
The present namespace exposes limited controls of its own, as described below.
Unicode::ICU’s documentation tries to be helpful while avoiding duplication of ICU’s own documentation. If something is unclear, consult the corresponding parts of ICU’s documentation to see if that helps. If confusion persists, file a documentation bug.
Generally speaking, all strings into and out of this distribution’s interfaces are character strings, not byte strings. If you get a wide-character warning or corrupted output, you may have neglected either a decode prior to calling ICU or an encode afterward. CPAN’s Encode::Simple provides a nice, fail-early-fail-often interface for these operations.
This module is tested with ICU versions as far back as 4.2.1 (the version that ships with CloudLinux 6). Some of this module’s functionality, though, is unavailable in certain ICU versions.
Errors from ICU are Unicode::ICU::X::ICU instances. Most others are plain strings; other errors are thrown as documented.
ICU_VERSION- a text string (e.g.,67.1)ICU_MAJOR_VERSION- an unsigned integer (e.g.,67)ICU_MINOR_VERSION- an unsigned integer (e.g.,1)
A wrapper around ICU’s u_errorName(), which gives a human-readable
name (e.g., U_BUFFER_OVERFLOW_ERROR) for an error code.
Some other ICU bindings exist on CPAN that do different things from this module: