test(tests): added test suite

The test suite is extensive. Note that LLMs were used to aid in the development of the test suite. All LLM generated code was manually reviewed.

tests/serialize/python_reader.py

@@ -0,0 +1,53 @@
+"""Verify the public format through an independent Python implementation."""
+
+import importlib.util
+import math
+from pathlib import Path
+import struct
+import subprocess
+import sys
+import tempfile
+
+
+def main():
+    """Check native/text equality and CSV output using files produced by C++."""
+    sys.dont_write_bytecode = True
+    writer, reader_path = sys.argv[1:]
+    spec = importlib.util.spec_from_file_location("serif_archive", reader_path)
+    module = importlib.util.module_from_spec(spec)
+    spec.loader.exec_module(module)
+    with tempfile.TemporaryDirectory(prefix="serif-python-reader-") as directory:
+        subprocess.run([writer, directory], check=True)
+        representations = []
+        for encoding in ("binary", "text"):
+            path = Path(directory) / f"{encoding}.serif"
+            with module.Archive(path) as archive:
+                values = archive.read("values")
+                assert len(values) == 5
+                assert math.copysign(1.0, values[0]) < 0
+                assert values[1] == 1.2345678901234567
+                assert values[2] == float.fromhex("0x0.0000000000001p-1022")
+                assert values[3] == math.inf
+                assert struct.pack("<d", values[4]).hex() == "bc9a78563412f87f"
+                assert archive.read("integers") == [-(1 << 63), (1 << 63) - 1]
+                assert archive.read("label") == 'ξ\n"quote"\\end'
+                assert archive.read("empty") == []
+                assert archive.read("provenance/author") == "Cross-language regression"
+                representations.append(struct.pack("<5d", *values))
+            csv = subprocess.run([sys.executable, reader_path, str(path), "--read", "values", "--csv"], check=True, capture_output=True, text=True).stdout
+            assert csv.startswith("index,value\n0,-0.0\n")
+            damaged = bytearray(path.read_bytes())
+            damaged[550] ^= 1
+            path.write_bytes(damaged)
+            try:
+                module.Archive(path)
+            except ValueError as error:
+                assert "ledger" in str(error)
+            else:
+                raise AssertionError("Corrupted ledger was accepted")
+        assert representations[0] == representations[1]
+    print("Python reader: binary/text exact values, provenance, CSV and corruption checks passed")
+
+
+if __name__ == "__main__":
+    main()