CLDR 23 Release Note

No. Date Rel. Note Data Charts Spec Delta SVN Tag DTD Diffs
23 2013-03-15 v23 CLDR23  Charts23 LDML23 Δ23 release-23

Unicode CLDR 23.0 contains data for 215 languages and 227 territories—654 locales in all. This release focused primarily on improvements to the LDML structure and tools, and on consistency of data. It includes substantially improved support for non-Gregorian calendars (such as the Japanese Imperial calendar used extensively in Japan). The data and structure has also been modified to easily permit changing between 12 and 24 hour formats, and between 2 digit and 4 digit years. The new Unicode character is used for the Turkish Lira, and information is provided for currencies that round to 5 cents (or other subunits) in cash transactions. For most languages that use non-Latin scripts, characters in the language’s script now collate before those in other scripts (including A-Z). Language-specific letter-casing changes (Lower, Upper, Title) have been added for Turkish, Azerbaijani, Lithuanian, and Greek. Keyboard data has also been updated for Android.

The release had a short cycle so that we could move to the new regular semi-annual scheduleIt thus only included a limited data submission phase, for 4 languages only: Armenian (hy), Georgian (ka), Mongolian (mn), and Welsh (cy). For those languages, the data increased by over 100%.

Other major features are described below. The table above points to the files for this release. For a description of their purpose and format, see the Key.


The main features include the following (see the Delta link above for a full list of changes, aside from data additions and corrections):

Structural additions and changes

  • The <fields> element has been moved from under each calendar (data only existed for gregorian) to just under the <dates> element [#5512]
  • A new "generic" calendar type is used to provide more consistent support for non-Gregorian calendars. This calendar type inherits from "gregorian", and all other calendar types (except chinese and dangi) inherit their format data from "generic", which typically has date patterns that include era. [#5385]
  • The <orientation> element attributes "characters" and "lines" are deprecated, replaced by new <characterOrder> and <lineOrder> elements under <orientation>. [#1297]
  • The ldml <singleCountries> element under <timeZoneNames> has been deprecated; it is replaced by the supplementalData top-level element <primaryZones>. [#5439]
  • In supplementalData, a new <timeData> element provides information on preferred and acceptable time cycles in various locales. [#5488]
  • In supplementalData, under <currencyData>: [#5548]
    • The <info> element under <fractions> now has an additional optional "cashRounding" attribute used for distinguishing rounding in cash transactions from rounding in other (e.g. electronic) transactions.
    • Individual <currency> elements can now use the digits, rounding, and cashRounding attributes.
  • The supplementalData <coverageAdditions> element and its subelements have been deprecated; the purpose for which this was formerly used is now handled by the <coverageLevels> element. [#5496]

Data changes

  • Update to support tzdata through 2013a. [#5450n, #5761]
  • Significant additions to the locale data for cy (Welsh), hy (Armenian), ka (Georgian), and mn (Mongolian in Cyrillic script).
  • For most languages that use non-Latin scripts, characters in the language’s script now collate before those in other scripts. [#4020, #5577]
  • For Assamese: Updated collation rules to support U+09CE KHANDA TA, added KSSA sequence to main exemplars. [#2998, #5441]
  • The Turkish locale now uses the new U+20BA TURKISH LIRA SIGN. [#4448]
  • Most date formats with years were changed to use single 'y' (standard Gregorian short formats using “yy” were not changed except in a few specific cases such as French). [#5235, #5347]
  • Initial "generic" format data is provided for all locales that had calendar format data. [#5385]
  • Several transforms have been added to support language-specific lettercasing changes (Lower, Upper, Title). [#2645, #4779]
  • Number spellout (RBNF) fixes for Chinese, Croatian, Finnish, Hebrew, Japanese, Polish, and Ethiopic numerals. [#5352, #5363, #5418, #5462, #5663, #5666, #5670, #5685, #5460]
  • Added data for more Android key layouts. [#5420]
  • A number of spelling updates to Hindi names of languages & regions. [#5224]
  • Many other specific fixes and updates, please see the delta link above.

Spec changes

  • The LDML specification has had a major reorganization. It has been split into multiple Parts, with all of the material relevant to a particular topic being gathered in each Part.

Conversion tools (including JSON support)

  • The new Ldml2JSonConverter supports conversion of LDML data to JSON format. [#2733 and many others]
  • The LDML2ICUConverter has been enhanced to convert more types of CLDR data for ICU use, including supplementalData <territoryInfo> and the new <timeData>. [#2404, #5660]

BCP47 enhancements (after CLDR 22.1)

  • New currency ("cu") type: "zmw" [Zambian Kwacha]
  • New time zone ("tz") types: "debsngn" [Europe/Busingen - Büsingen, Germany], "rukhndg" [Asia/Khandyga - Khandyga Tomponsky, Russia] and "ruunera" [Asia/Ust-Nera - Ust-Nera Oymyakonsky, Russia]


  • The Release Note contains a general description of the contents of the release, and any relevant notes about the release.
  • The Data link points to a set of zip files containing the contents of the release (the files are complete in themselves, and do not require files from earlier releases -- for the structure of the zip file, see Repository Organization).
  • The Spec is the version of UTS #35: LDML that corresponds to the release.
  • The Delta document points to a list of all the bug fixes and features in the release, which be used to get the precise corresponding file changes using BugDiffs.
  • The SVN Tag can be used to get the files via Repository Access.
  • For more details see CLDR Releases (Downloads).

The Unicode Terms of Use apply to CLDR data; in particular, see Exhibit 1.

For web pages with different views of CLDR data, see