01-09-2019, 01:51 PM
(01-09-2019, 11:11 AM)Torsten Wrote: You are not allowed to view links. Register or Login to view.This is exactly the point. EVA is a stroke-based transcription alphabet and therefore it is necessary to parse the strokes into tokens. Remains the question how to parse EVA into tokens.
But EVA is glyph-based, not stroke-based. Benched glyphs are interpreted as ligatures because composing glyphs often appear separately. A stroke-based transliteration that would make sense for a study such as yours would be:
EVA-i = \ (1 stroke)
EVA-e = C (1 stroke)
EVA-o = C> or C] or C) (2 strokes)
EVA-a = C\ (2 strokes)
EVA-ch = CT (2 strokes)
EVA-Sh = CT^ (3 strokes)
EVA-cKh = CIPT (4 strokes)
EVA-cTh = CIQT (4 strokes)
etc.
Why would there be a need to tokenize?