You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Take the content in the ALTLab version of AECD and convert that into the general JSON format, for eventual inclusion into the aggregated importjson format, after comparison with CW and MD.
The text was updated successfully, but these errors were encountered:
We should separate into distinct senses AECD definitions that contain the semi-colon as a delimiter.
There appear to be 1611 such cases on the crk2eng version of AECD, and an examination of random entries does suggest that the semicolon is used systematically to distinguish senses.
In addition, there's some AECD-specific encoding, such as Alt. or Var. for alternative or variant Cree word forms, (Plains) and (Northern) indicating dialect, that we wouldn't want to include in the search. For the 731 Alt. cases and 180 Var. cases, we'd probably want to encode the Cree word-forms appropriately (something for Daniel), perhaps marking them with <crk>...</crk>.
Take the content in the ALTLab version of AECD and convert that into the general JSON format, for eventual inclusion into the aggregated importjson format, after comparison with CW and MD.
The text was updated successfully, but these errors were encountered: