libunistring: upgrade 1.1 -> 1.2

libunistring: upgrade 1.1 -> 1.2

License-Update:
===============
doc/libunistring.texi (Conventions):
 Explain the (resultbuf, lengthp) return value convention from the user's point of view.
doc/libunistring.texi (Include files):
 Fix documentation of _LIBUNISTRING_VERSION and _libunistring_version (incorrect since 2010-05-14).
 Add documentation of _libunistring_unicode_version.

Changelog:
=============
* The data tables and algorithms have been updated to Unicode version 15.1.0.
* New functions u8_pcpy, u16_pcpy, u32_pcpy, similar to mempcpy.
* New functions uc_indic_conjunct_break_name, uc_indic_conjunct_break_byname,
  uc_indic_conjunct_break.
* New functions
    uc_is_property_prepended_concatenation_mark,
    uc_is_property_id_compat_math_start, uc_is_property_id_compat_math_continue,
    uc_is_property_ids_unary_operator
  and new constants
    UC_PROPERTY_PREPENDED_CONCATENATION_MARK,
    UC_PROPERTY_ID_COMPAT_MATH_START, UC_PROPERTY_ID_COMPAT_MATH_CONTINUE,
    UC_PROPERTY_IDS_UNARY_OPERATOR.
* New constant _libunistring_unicode_version.
* The UTF-8 decoder functions, especially u8_mbtouc, are now more Unicode
  Standard compliant.
* The *printf functions no longer support the %n directive, for security
  reasons.
* Fixed a bug in the *printf functions: In the %U, %lU, %llU directives, a
  negative width given as an argument did not trigger left-justification.
* The functions u16_strstr and u32_strstr now operate in worst-case linear
  time.

Signed-off-by: Wang Mingyu <wangmy@fujitsu.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>