The parser is currently in development, and 60-80% of MLIR files in the test suite of MLIR, within the Arith, Math, SCF, Tensor, and Linalg dialects parse successfully.