References : 1158955 1159131 1161007 1162882 1167603 1182252 1182645
This update for icu.691 fixes the following issues:
- Renamed package from icu 69.1 for SUSE:SLE-15-SP3:Update. (jsc#SLE-17893)
- Fix undefined behaviour in 'ComplexUnitsConverter::applyRounder'
- Update to release 69.1
- For Norwegian, 'no' is back to being the canonical code, with
'nb' treated as equivalent. This aligns handling of Norwegian
with other macro language codes.
- Binary prefixes in measurement units (KiB, MiB, etc.)
- Time zone offsets from local time with new APIs.
- Don't disable testsuite under 'qemu-linux-user'
- Fixed an issue when ICU test on 'aarch64 fails. (bsc#1182645)
- Drop 'SUSE_ASNEEDED' as the issue was in binutils. (bsc#1182252)
- Fix 'pthread' dependency issue. (bsc#1182252)
- Update to release 68.2
- Fix memory problem in 'FormattedStringBuilder'
- Fix assertion when 'setKeywordValue w/' long value.
- Fix UBSan breakage on 8bit of rbbi
- fix int32_t overflow in listFormat
- Fix memory handling in MemoryPool::operator=()
- Fix memory leak in AliasReplacer
- Add back icu.keyring.
- Update to release 68.1
- PluralRules selection for ranges of numbers
- Locale ID canonicalization now conforms to the CLDR spec including edge cases
- DateIntervalFormat supports output options such as capitalization
- Measurement units are normalized in skeleton string output
- Time zone data (tzdata) version 2020d
- Add the provides for libicu to Make .Net core can install successfully. (bsc#1167603, bsc#1161007)
- Update to version 67.1
- Unicode 13 (ICU-20893, same as in ICU 66)
- Total of 5930 new characters
- 4 new scripts
- 55 new emoji characters, plus additional new sequences
- New CJK extension, first characters in plane 3: U+30000..U+3134A
- New language at Modern coverage: Nigerian Pidgin
- New languages at Basic coverage: Fulah (Adlam), Maithili,
Manipuri, Santali, Sindhi (Devanagari), Sundanese
- Region containment: EU no longer includes GB
- Unicode 13 root collation data and Chinese data for collation and transliteration
- DateTimePatternGenerator now obeys the 'hc' preference in the locale identifier
- Various other improvements for ECMA-402 conformance
- Number skeletons have a new 'concise' form that can be used in MessageFormat strings
- Currency formatting options for formal and other currency display name variants
- ListFormatter: new public API to select the style & type
- ListFormatter now selects the proper “and”/“or” form for Spanish & Hebrew.
- Locale ID canonicalization upgraded to implement the complete CLDR spec.
- LocaleMatcher: New option to ignore one-way matches
- acceptLanguage() reimplemented via LocaleMatcher
- Data build tool: tzdbNames.res moved from the 'zone_tree' category to the 'zone_supplemental' category
- Fixed uses of u8'literals' broken by the C++20 introduction of the incompatible char8_t type
- and added a few API overloads to reduce the need for reinterpret_cast.
- Support for manipulating CLDR 37 unit identifiers in MeasureUnit.
- Drop icu-versioning. (bsc#1159131)
- Update to version 66.1
- Unicode 13 support
- Fix uses of u8'literals' broken by C++20 introduction of incompatible char8_t type.
- Fixed an issue when Qt apps can't handle non-ASCII filesystem path. ([bsc#1162882)
- Remove '/usr/lib(64)/icu/current'. (bsc#1158955)
The following package changes have been done:
- libicu69-ledata-69.1-7.3.2 added
- libicu69-69.1-7.3.2 added
- libgdbm4-1.12-1.418 removed
- libicu-suse65_1-65.1-4.2.1 removed
- libicu65_1-ledata-65.1-4.2.1 removed
- perl-5.26.1-15.87 removed
- update-alternatives-188.8.131.52-4.3.1 removed
- vim-8.0.1568-5.17.1 removed
- vim-data-common-8.0.1568-5.17.1 removed