Function generate_markdown_report

benchmarks/python/benchmark_report.py:234–363

(raw, data, sizes, output_dir: Path, plot_prefix: str)

def generate_markdown_report(raw, data, sizes, output_dir: Path, plot_prefix: str):
    context = raw.get("context", {})
    system_info = get_system_info()

    if context.get("python_implementation"):
        system_info["Python Implementation"] = context["python_implementation"]
    if context.get("platform"):
        system_info["Benchmark Platform"] = context["platform"]

    datatypes = [dt for dt in DATATYPE_ORDER if dt in data]
    operations = ["serialize", "deserialize"]

    md = [
        "# Python Benchmark Performance Report\n\n",
        f"_Generated on {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}_\n\n",
        "## How to Generate This Report\n\n",
        "```bash\n",
        "cd benchmarks/python\n",
        "./run.sh\n",
        "```\n\n",
        "## Benchmark Plot\n\n",
        "The plot shows throughput (ops/sec); higher is better.\n\n",
        f"![Throughput]({plot_prefix}throughput.png)\n\n",
        "## Hardware & OS Info\n\n",
        "| Key | Value |\n",
        "|-----|-------|\n",
    ]

    for key, value in system_info.items():
        md.append(f"| {key} | {value} |\n")

    md.append("\n## Benchmark Configuration\n\n")
    md.append("| Key | Value |\n")
    md.append("|-----|-------|\n")
    for key in ["warmup", "iterations", "repeat", "number", "list_size"]:
        if key in context:
            md.append(f"| {key} | {context[key]} |\n")

    md.append("\n## Benchmark Results\n\n")
    md.append("### Timing Results (nanoseconds)\n\n")
    timing_headers = [
        f"{SERIALIZER_LABELS.get(lib, lib)} (ns)" for lib in SERIALIZER_ORDER
    ]
    md.append(
        "| Datatype | Operation | " + " | ".join(timing_headers) + " | Fastest |\n"
    )
    md.append(
        "|----------|-----------|"
        + "|".join("-" * (len(header) + 2) for header in timing_headers)
        + "|---------|\n"
    )

    for datatype in datatypes:
        for operation in operations:
            times = {
                lib: data.get(datatype, {}).get(operation, {}).get(lib, 0)
                for lib in SERIALIZER_ORDER
            }