Add CLI entrypoints for module and program invocation (#152)

* Add CLI entrypoints for module and program invocation

Supports `nnbench <benchmark> <options>` as well as
`python -m nnbench <benchmark> <options>`.

Like pytest, we give the option to omit the target location, in which
case it becomes a directory named "benchmarks".

Supports a very rudimentary context building from switches method, and
tag filtering via switch.

pyproject.toml

@@ -65,6 +65,9 @@ docs = [
     "neoteroi-mkdocs",
 ]
 
+[project.scripts]
+nnbench = "nnbench.cli:main"
+
 [tool.setuptools]
 package-dir = { "" = "src" }
 

src/nnbench/__init__.py

@@ -11,4 +11,13 @@
 from .core import benchmark, parametrize, product
 from .reporter import BenchmarkReporter
 from .runner import BenchmarkRunner
-from .types import Memo, Parameters
+from .types import Benchmark, BenchmarkRecord, Memo, Parameters
+
+
+# TODO: This isn't great, make it functional instead?
+def default_runner() -> BenchmarkRunner:
+    return BenchmarkRunner()
+
+
+def default_reporter() -> BenchmarkReporter:
+    return BenchmarkReporter()

src/nnbench/__main__.py

@@ -0,0 +1,6 @@
+import sys
+
+from nnbench.cli import main
+
+if __name__ == "__main__":
+    sys.exit(main())

src/nnbench/cli.py

@@ -0,0 +1,49 @@
+import argparse
+from typing import Any
+
+from nnbench import default_reporter, default_runner
+
+
+def main() -> int:
+    parser = argparse.ArgumentParser("nnbench")
+    # can be a directory, single file, or glob
+    parser.add_argument(
+        "benchmarks",
+        nargs="?",
+        metavar="<benchmarks>",
+        help="Python file or directory of files containing benchmarks to run.",
+        default="benchmarks",
+    )
+    parser.add_argument(
+        "--context",
+        action="append",
+        metavar="<key>=<value>",
+        help="Additional context values giving information about the benchmark run.",
+        default=list(),
+    )
+    parser.add_argument(
+        "-t",
+        "--tag",
+        action="append",
+        metavar="<tag>",
+        dest="tags",
+        help="Only run benchmarks marked with one or more given tag(s).",
+        default=tuple(),
+    )
+
+    args = parser.parse_args()
+
+    runner = default_runner()
+    reporter = default_reporter()
+
+    context: dict[str, Any] = {}
+    for val in args.context:
+        try:
+            k, v = val.split("=")
+        except ValueError:
+            raise ValueError("context values need to be of the form <key>=<value>")
+        context[k] = v
+
+    record = runner.run(args.benchmarks, tags=tuple(args.tags))
+    reporter.display(record)
+    return 0