Log message:
icu: Revert -fno-addrsig for Clang

This should be handled in a different place.
The addrsig property breaks GNU strip(1) in a cryptic way, creating
misparsable ELF file.

Log message:
icu: Append -fno-addrsig in clang build

The -faddrsig option (enabled by default) generates executables that are
not compatible with GNU strip(1).

Log message:
Apply ICU-20208 uspoof.cpp function checkImpl should be static from
https://github.com/unicode-org/icu/commit/8baff8f03e07d8e02304d0c888d0bb21ad2eeb01

This should fix the firefox build with this version of icu

Bump PKGREVISION

Log message:
icu: updated to 63.1

ICU 63.1:

Common Changes
- CLDR 34
- Segmentation rules and emoji sort order adjusted for Unicode 11
- Somali and Javanese data now up to moderate level (document content)
- Tongan, Konkani, Maori, Dzongkha, Tatar, Kurdish (ku), and Xhosa data now up \ 
to basic level
- Many data additions and corrections in many other languages
- The following languages have been added to ICU: Sindhi, Maori, Turkmen, \ 
Javanese, Interlingua, Kurdish (ku), Xhosa
- New currency: Venezuela's Bolívar Soberano (VES)
- New Japanese calendar era from 2019: CLDR and ICU include data for testing \ 
that can be enabled.
To enable CLDR new Japanese era placeholder name, set environment variable (and \ 
Java system property for ICU4J) ICU_ENABLE_TENTATIVE_ERA=true.
- New API for number and currency range formatting
Support for additional Unicode properties: Indic_Positional_Category & \ 
Indic_Syllabic_Category and Vertical_Orientation
- New API for code point maps and tries, mapping Unicode code points \ 
(U+0000..U+10FFFF) to integer values.
- Java classes CodePointMap, CodePointTrie, MutableCodePointTrie
- C types UCPMap, UCPTrie, UMutableCPTrie
- New API for getting a UnicodeSet per binary property and a code point map per \ 
enumerated/int-value property.
- Full conformance with UAX 14 Line Breaking (required BreakIterator feature work).

ICU4C Specific Changes
- C++ Locale class
- Additional functions forLanguageTag()/toLanguageTag(), and functions that are \ 
easier and safer to use by using StringPiece and ByteSink rather than raw \ 
buffers.
- Move semantics.
- ICU4C: Various Out-Of-Memory (OOM) issues have been fixed. (ticket query)
- The icu-config tool has been deprecated.  You can use the --disable-icu-config \ 
option to disable icu-config from being installed.  Alternately, you can use \ 
--enable-icu-config to enable icu-config. In the future, icu-config will be \ 
disabled by default

Log message:
Rather than playing "reserved define whack-a-mole" on NetBSD by trying to
define _ISOC99_SOURCE to mitigate the damage of defining _XOPEN_SOURCE,
just don't define _XOPEN_SOURCE - nothing in the source defines it and
it breaks at least gcc 6.4 on NetBSD-8.0

No PKGREVISION bump as should not affect any building systems

Log message:
Update to 62.1

Changelog:
Common Changes
    Unicode 11: 684 new characters, including 7 new scripts, Mtavruli Georgian \ 
capital letters, 5 new Han characters, and 66 new emoji characters.
    CLDR 33.1:
        Unicode 11 script metadata, collation, Chinese transliteration.
        Chinese collation stroke order updated from Unicode 7 to Unicode 11 \ 
after tooling bug fixes.
    NumberFormatter
        A NumberFormatter can now be constructed from a locale-neutral skeleton \ 
string (like a DateFormat) (#8610). This is particularly useful in translated \ 
messages where placeholder details should not be translated. MessageFormat \ 
recognizes the style field as a number skeleton if it is prefixed with \ 
"::", as in "Number of files: {num, number, :: round-integer \ 
group-min2}." (#13742)
        New "conversion" functions for getting a NumberFormatter from \ 
a DecimalFormat, and a Format from a NumberFormatter.
        New C API (unicode/unumberformatter.h [permanent API docs link TBD]). \ 
(#13597) Currently it supports formatting settings only via a skeleton string.
        Several still-draft NumberFormatter methods and helper classes have been \ 
modified or renamed; the previous versions remain temporarily (as deprecated) \ 
for one release, to help with the transition.
    Break Iterator Rules: "Safe" rules are no longer required for \ 
correct break iterator operation. For back compatibility, existing rule sets \ 
containing safe rules will continue to work, with the safe rules they contain \ 
being ignored. The Break Iterator binary data format has been updated to reflect \ 
this change.
    Line Break: The boundary rules have been updated to reflect the Unicode 11 \ 
version of UAX #14. Specifically, the handling of Emoji ZWJ sequences has been \ 
improved.

ICU4C Specific Changes
    Under-the-hood overhaul of number parsing. See the design doc for a summary \ 
of changes; behavior is mostly compatible with previous versions, but there are \ 
some known differences.
    DecimalFormat now wraps the new NumberFormatter code.

Log message:
icu: fix build on armeb by adding it to the long list of archs that exist
and don't have 80bit float types.

PR pkg/53408

Log message:
textproc/icu: updated to 61.1

61.1:
Common Changes
* CLDR 33:
  - Two additional locales (Odia, Assamese) were brought up to Modern coverage level.
  - 4 new transforms: fa-fa_FONIPA, ha-ha_NE, nv-nv_FONIPA, vec-vec_FONIPA.
  - New currency code MRU for Mauritania.
  - Arabic native vs. ASCII digits.
  - Data additions & bug fixes.
* Many small API additions, improvements, and bug fixes.

ICU4C Specific Changes
* Added Google double-conversion library for formatting doubles. This is the \ 
library used in V8 and a number of other projects for converting doubles to \ 
decimals. To avoid name collisions, the library is linked internally under the \ 
ICU namespace as icu::double_conversion. Our copy of double_conversion is not \ 
intended for public usage.
* Re-wrote U8_NEXT macros to eliminate all library function calls.

Log message:
icu: Revert previous clang patch, clang was changed instead.

Log message:
icu: Don't perform SunOS _STDC_C99 workaround with clang.