lib/ai: remove randomness in puzzle so benches can compare sanely
[nit.git] / tests / error_parser_oneline_string.nit
1 "hello"