emoji: repair one-shot candidate selection

tests/mkemoji_test.py

@@ -0,0 +1,85 @@
+#!/usr/bin/env python3
+
+import subprocess
+import sys
+import tempfile
+import unittest
+from pathlib import Path
+
+
+ROOT = Path(__file__).resolve().parents[1]
+MKEMOJI = ROOT / "map" / "mkemoji"
+
+
+def generate(source=None):
+    args = [sys.executable, "-B", str(MKEMOJI)]
+    if source is not None:
+        args.append(str(source))
+    return subprocess.run(args, capture_output=True, text=True, check=False)
+
+
+def table(output):
+    return dict(line.split("\t", 1) for line in output.splitlines())
+
+
+class MkemojiTest(unittest.TestCase):
+    def source(self, text):
+        tmp = tempfile.TemporaryDirectory()
+        path = Path(tmp.name) / "emoji.src"
+        path.write_text(text, encoding="utf-8")
+        self.addCleanup(tmp.cleanup)
+        return path
+
+    def test_production_digit_slots(self):
+        result = generate()
+        self.assertEqual(result.returncode, 0, result.stderr)
+        data = table(result.stdout)
+        self.assertEqual(data["^"].split()[:9], list("¹²³⁴⁵⁶⁷⁸⁹"))
+        self.assertEqual(data["_"].split()[:9], list("₁₂₃₄₅₆₇₈₉"))
+        self.assertEqual(data["<"].split()[2], "♥")
+        self.assertEqual(data["^1"], "¹")
+        self.assertEqual(data["_2"], "₂")
+        self.assertEqual(data["<3"], "♥")
+
+    def test_fold_normalize_and_exact_first(self):
+        source = self.source(
+            "β\tALPHABET\n"
+            "α\talpha\n"
+            "e\u0301\tE\u0301\n"
+            "#\thash\n"
+        )
+        first = generate(source)
+        second = generate(source)
+        self.assertEqual(first.returncode, 0, first.stderr)
+        self.assertEqual(first.stdout, second.stdout)
+        data = table(first.stdout)
+        self.assertEqual(data["alpha"].split(), ["α", "β"])
+        self.assertEqual(data["al"].split(), ["β", "α"])
+        self.assertEqual(data["é"], "é")
+        self.assertEqual(data["hash"], "#")
+
+    def test_rejects_malformed_and_oversized_rows(self):
+        bad = [
+            "x\n",
+            "bad result\ta\n",
+            "x\t\n",
+            "x\t a\n",
+            "\0\ta\n",
+            "x\ta\0b\n",
+            "x" * 65 + "\ta\n",
+            "x\t" + "a" * 65 + "\n",
+        ]
+        for text in bad:
+            with self.subTest(text=repr(text)):
+                result = generate(self.source(text))
+                self.assertEqual(result.returncode, 1)
+
+    def test_accepts_runtime_boundary(self):
+        source = self.source("x" * 64 + "\t" + "a" * 64 + "\n")
+        result = generate(source)
+        self.assertEqual(result.returncode, 0, result.stderr)
+        self.assertIn("a" * 64 + "\t" + "x" * 64, result.stdout)
+
+
+if __name__ == "__main__":
+    unittest.main()