Use lazy import

Author: vstinner

Lib/test/test_tomllib/test_misc.py

@@ -9,8 +9,11 @@
 from pathlib import Path
 import sys
 import tempfile
+import textwrap
 import unittest
 from test import support
+from test.support import os_helper
+from test.support.script_helper import assert_python_ok
 
 from . import tomllib
 
@@ -139,3 +142,26 @@ def test_parse_simple_number(self):
         self.assertIsNone(parse_simple_number("x123\n", 0))
         self.assertIsNone(parse_simple_number("o123\n", 0))
         self.assertIsNone(parse_simple_number("b100\n", 0))
+
+    def test_lazy_import(self):
+        # Test that _parse_simple_number() can parse the TOML file without
+        # importing regular expressions (tomllib._re)
+        filename = os_helper.TESTFN
+        self.addCleanup(os_helper.unlink, filename)
+        toml = textwrap.dedent("""
+            [metadata]
+            int = 123
+            list = [+1, -2, 3]
+            table = {x=1, y=2}
+        """)
+        with open(filename, "w") as fp:
+            fp.write(toml)
+
+        code = textwrap.dedent(f"""
+            import sys, tomllib
+            with open({filename!a}, "rb") as fp:
+                tomllib.load(fp)
+            print("lazy import?", 'tomllib._re' not in sys.modules)
+        """)
+        proc = assert_python_ok('-c', code)
+        self.assertIn(b'lazy import? True', proc.out)

Lib/tomllib/_parser.py

@@ -6,26 +6,24 @@
 
 from types import MappingProxyType
 
+__lazy_modules__ = ["tomllib._re"]
+
+from ._re import (
+    RE_DATETIME,
+    RE_LOCALTIME,
+    RE_NUMBER,
+    match_to_datetime,
+    match_to_localtime,
+    match_to_number,
+)
+
 TYPE_CHECKING = False
 if TYPE_CHECKING:
     from collections.abc import Iterable
     from typing import IO, Any, Final
 
     from ._types import Key, ParseFloat, Pos
 
-    _REGEX_IMPORTED = True
-    from ._re import (
-        RE_DATETIME,
-        RE_LOCALTIME,
-        RE_NUMBER,
-        match_to_datetime,
-        match_to_localtime,
-        match_to_number,
-    )
-else:
-    # Regular expressions are lazy imported to speed up startup time
-    _REGEX_IMPORTED = False
-
 ASCII_CTRL: Final = frozenset(chr(i) for i in range(32)) | frozenset(chr(127))
 
 # Neither of these sets include quotation mark or backslash. They are
@@ -727,24 +725,11 @@ def parse_value(
     if char == "{":
         return parse_inline_table(src, pos, parse_float)
 
-    global _REGEX_IMPORTED, RE_DATETIME, RE_LOCALTIME, RE_NUMBER
-    global match_to_datetime, match_to_localtime, match_to_number
-    if not _REGEX_IMPORTED:
-        # Simple number parser avoiding regex
-        if char in _DECDIGIT_CHARS:
-            res = _parse_simple_number(src, pos)
-            if res is not None:
-                return res
-
-        from ._re import (
-            RE_DATETIME,
-            RE_LOCALTIME,
-            RE_NUMBER,
-            match_to_datetime,
-            match_to_localtime,
-            match_to_number,
-        )
-        _REGEX_IMPORTED = True
+    # Simple number parser avoiding regex
+    if char in _DECDIGIT_CHARS:
+        res = _parse_simple_number(src, pos)
+        if res is not None:
+            return res
 
     # Dates and times
     datetime_match = RE_DATETIME.match(src, pos)