KineticCore's built-in Han-to-Mandarin reading table is derived from Unicode 18.0.0 Unihan,
Unihan_Readings.txt, kMandarin. KineticCore's transliteration and search-matching code is
implemented independently and has no runtime or build dependency on a third-party pinyin library.

Source: https://www.unicode.org/Public/18.0.0/ucd/Unihan.zip
Documentation: https://www.unicode.org/reports/tr38/
License: https://www.unicode.org/license.txt

When kMandarin provides multiple values, the first reading is selected for zh-Hans as specified
by UAX #38. Tone marks are omitted for search, and ü is represented as v.
