This would involve a backwards-compatible addition to the current API.
Basically, we should have a way of sending not only the text to be glossed, but also a large chunk thereof that could be used for part-of-speech / word sense disambiguation. (Right now, that would only be useful for Madamira and possibly WWWJDIC, but once the API is in place, one could imagine other services being upgraded to take advantage of it later.)
I think the simplest way to do this would be to just add "start" and "end" indices to the request API, which default to "0" and "length of the text", respectively, which would indicate the segment of the text that the client wants glossed, with the rest being treated as context.
This would involve a backwards-compatible addition to the current API.
Basically, we should have a way of sending not only the text to be glossed, but also a large chunk thereof that could be used for part-of-speech / word sense disambiguation. (Right now, that would only be useful for Madamira and possibly WWWJDIC, but once the API is in place, one could imagine other services being upgraded to take advantage of it later.)
I think the simplest way to do this would be to just add "start" and "end" indices to the request API, which default to "0" and "length of the text", respectively, which would indicate the segment of the text that the client wants glossed, with the rest being treated as context.