* much more tests
* better support for unicode characters and template classes that are instantinated with unicode character types
* meta code refactored (it is now shorter and clearer)
[SVN r77824]
* reduces templates count
* leaves only one lexical converting class (at least for modern compilers)
* fixes small TODOs and XXXs from source code
* makes lexical_stream_limited_src more readable
* updates status/explicit-failures-markup.xml
* makes lexical_cast_inf_nan_test.cpp pass on Itanium pltform
* makes lexical_cast able to convert signed and unsigned chars to wchar_t
and updates lexical_cast_wchars_test
[SVN r73313]