Crate llama_cpp_log_decoder 

Decoder for the llama.cpp / ggml log callback stream.

The C side delivers log lines in fragments: a missing trailing newline signals that more fragments will follow at GGML_LOG_LEVEL_CONT. This crate is a pure &mut self transducer — feed (level, text) pairs, get complete LogLines back when the trailing newline arrives. No globals, no atomics, no FFI, no logger.

Modules

decode_anomaly

decode_output

decode_result

incoming_log_level

log_decoder

log_level

log_line