Use translate to improve performance of normalize

Author: hugovk

importlib_metadata/__init__.py

@@ -910,6 +910,14 @@ def search(self, prepared: Prepared):
         return itertools.chain(infos, eggs)
 
 
+# Translation table for Prepared.normalize: lowercase and
+# replace "-" (hyphen) and "." (dot) with "_" (underscore).
+_normalize_table = str.maketrans(
+    "ABCDEFGHIJKLMNOPQRSTUVWXYZ-.",
+    "abcdefghijklmnopqrstuvwxyz__",
+)
+
+
 class Prepared:
     """
     A prepared search query for metadata on a possibly-named package.
@@ -945,7 +953,13 @@ def normalize(name):
         """
         PEP 503 normalization plus dashes as underscores.
         """
-        return re.sub(r"[-_.]+", "-", name).lower().replace('-', '_')
+        # Emulates ``re.sub(r"[-_.]+", "-", name).lower()`` from PEP 503
+        # About 3x faster, safe since packages only support alphanumeric characters
+        value = name.translate(_normalize_table)
+        # Condense repeats (faster than regex)
+        while "__" in value:
+            value = value.replace("__", "_")
+        return value
 
     @staticmethod
     def legacy_normalize(name):

tests/test_api.py

@@ -6,6 +6,7 @@
 from importlib_metadata import (
     Distribution,
     PackageNotFoundError,
+    Prepared,
     distribution,
     entry_points,
     files,
@@ -317,3 +318,36 @@ class InvalidateCache(unittest.TestCase):
     def test_invalidate_cache(self):
         # No externally observable behavior, but ensures test coverage...
         importlib.invalidate_caches()
+
+
+class PreparedTests(unittest.TestCase):
+    def test_normalize(self):
+        tests = [
+            # Simple
+            ("sample", "sample"),
+            # Mixed case
+            ("Sample", "sample"),
+            ("SAMPLE", "sample"),
+            ("SaMpLe", "sample"),
+            # Separator conversions
+            ("sample-pkg", "sample_pkg"),
+            ("sample.pkg", "sample_pkg"),
+            ("sample_pkg", "sample_pkg"),
+            # Multiple separators
+            ("sample---pkg", "sample_pkg"),
+            ("sample___pkg", "sample_pkg"),
+            ("sample...pkg", "sample_pkg"),
+            # Mixed separators
+            ("sample-._pkg", "sample_pkg"),
+            ("sample_.-pkg", "sample_pkg"),
+            # Complex
+            ("Sample__Pkg-name.foo", "sample_pkg_name_foo"),
+            ("Sample__Pkg.name__foo", "sample_pkg_name_foo"),
+            # Uppercase with separators
+            ("SAMPLE-PKG", "sample_pkg"),
+            ("Sample.Pkg", "sample_pkg"),
+            ("SAMPLE_PKG", "sample_pkg"),
+        ]
+        for name, expected in tests:
+            with self.subTest(name=name):
+                self.assertEqual(Prepared.normalize(name), expected)