Closed putstrln closed 9 years ago
Can I know how speakingurl handles cases where the transliteration changes as more diatrics are added? I am assuming I don't have to write every possible combos for each alphabet. The dotted circle represents the space for an alphabet. Given:
'က': 'k', // alphabet
'ေ': 'e', // diatric
'ာ': 'aa', // diatric
'ော': 'aw' // two diatric
Correct Results:
'က' + 'ေ' = 'ကေ' => 'ke',
'က' + 'ာ' = 'ကာ' => 'kaa'
Wrong results
'က' + 'ေ' + 'ာ' = 'ကော' => 'keaa'
'ကေ' + 'ာ' = 'ကော' => 'keaa'
These need to match the last case 'ော': 'aw' and become 'kaw'.
'က' + 'ေ' + 'ာ' = 'ကော' => 'kaw''
I think, speakingurl isn't really build with this specific language in mind. You already insert a translation-table to your pull request, that's the way it will be handled. Furthermore I couldn't verify the specific cases – I'm not experienced. If you can add this correct behavour for the burmese language, pull requests are welcome :-) But please add tests to your pull request, so that future development will not destroy the burmese transliteration support.
@pid, I tried to add some logic for lookahead characters that doesn't impact linear languages. It's now working with examples and unit tests.
awesome :+1: released v0.19.0
thanks for contribution :-)
Thanks for the project too :)
Let me know if something's missing. Thanks!