Status of vidyut-cheda

[Sai Sreemayee Sistla]

Namaste,

I wanted to know about the status of vidyut-cheda. The present README suggests that it implements an HMM-style splitter. I used the tool a few months ago for sandhi-splitting on a section of the Pañcatantra. I think it does have scope for improvement.

Input. brahmā rudraḥ kumāro harivaruṇayamā vahnir indraḥ kuberaś
candrādityau sarasvatyudadhiyuganagā vāyur urvībhujaṅgāḥ

Output. brahmā rudras kum ās ro harivaruṇayamā vahnir indras candrāt ityau sarasvatyudadhiyuganagā vās i ur urvībhujaṅgāḥ 

Is it under active development? What is the future of vidyut-cheda? Is using a different mechanism than the present one considered? 

[I apologise if I am not up-to-date with any relevant discussion here, as I am new here].

Thank you,

Sreemayee

[Arun Prasad]

vidyut-cheda was started before Dharmamitra was widely available. Since Dharmamitra is much better and also freely available via API, vidyut-cheda is in an unclear position.

I think it still has potential as a very fast segmenter for proofreading or other simple tasks, but it will likely never be state of the art. For that reason, development on it has stalled while we figure out where vidyut-cheda is most useful.

Arun