probcomp/bayeslite
use flat token sequence, not nested lists of strings, for CREATE GENERATOR schema
Aperta
#395 aperta il 19 apr 2016
help wantedk dayslanguagemetamodel revision
Metriche repository
- Star
- (915 stelle)
- Metriche merge PR
- (Nessuna PR mergiata in 30 g)
Descrizione
The nested lists of strings were convenient for simple-minded hand-written parsers, but we want to expand the language of population schemas substantially. This will make it easier to write standard grammars, e.g. with lemonade, for complex languages. We should then rewrite the Crosscat population schema grammar with lemonade.