def extract_movement_rep_and_sample()

in scheduler.py [0:0]

• 22 lines of code
• 8 McCabe index (conditional complexity)

def extract_movement_rep_and_sample(log_file):
    """Extracts data movement ingress and egress bytes from log.txt"""
    movement_ingress_bytes = movement_egress_bytes = 0
    replication_size = None
    sample_rate = None

    if os.path.exists(log_file):
        with open(log_file, "r") as f:
            for line in f:
                # Extract data movement bytes
                movement_match = re.search(r"Data movement:\s([\d.]+\s*[A-Z]*B) ingress,\s([\d.]+\s*[A-Z]*B) .* egress",
                                           line)
                if movement_match:
                    movement_ingress_bytes = parse_size(movement_match.group(1))
                    movement_egress_bytes = parse_size(movement_match.group(2))

                # Extract replication size (overlap)
                replication_match = re.search(r"Storage: .*?(\d+\.\d+\s*[A-Z]B)\s+overlap", line)
                if replication_match:
                    replication_size = parse_size(replication_match.group(1))

                # Extract sample rate (k=0.XXX) or k=1.000
                sample_match = re.search(r"k=(1\.000)", line)
                if sample_match:
                    sample_rate = 1.0
                sample_match = re.search(r"k=(0\.\d+)", line)
                if sample_match:
                    sample_rate = float(sample_match.group(1))  # Convert to float


    return movement_ingress_bytes, movement_egress_bytes, replication_size, sample_rate