CLDR Ticket #10199(new unknown)
Armenian BGN transliterator fails with LHS context on transliterated text, missing digraph
|Reported by:||robert@…||Owned by:||sascha|
The same problem as was fixed some time ago with the Russian-Latin/BGN... The LHS of a context rule was written assuming that the text is still the original, not the transliterated one.
For example ԱԵ converts to AE instead of AYE because the vowel Ա has already been converted to A before the context LHS can test for the presence of the original Ա. (It then fails to convert Ե into YE.)
Either the rules should be re-ordered and re-run (with a ::null;) or written to match transliterated characters on the LHS.
Also PCGN note 3 is not implemented (digraph Եի is equivalent to ԵՎ ref. https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/499634/ROMANIZATION_SYSTEM_FOR_ARMENIAN.PDF).