Function context_from_tokens 

pub fn context_from_tokens(tokens: &[Token], input_len: usize) -> GrammarContext

Derive grammar context from a token sequence (skipping whitespace/comments). input_len is the total length of the original input, used to detect trailing whitespace.