optimize Version.__init__

Version objects are created thousands of times by the check_version method of
decorators.  Creation is inefficient, resulting in five calls to re.match()
that do not even use precompiled regex.  The use of re.match() however is
fully redundant, as finditer() can provide the same information with a better
use of groupings.  Do that and precompile the tokenization regex.

This saves about 3% in the "meson setup" run of QEMU.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

1 files changed, 6 insertions, 9 deletions

diff --git a/mesonbuild/utils/universal.py b/mesonbuild/utils/universal.py
index eede9fd2d..e2634b925 100644
--- a/mesonbuild/utils/universal.py
+++ b/mesonbuild/utils/universal.py
@@ -828,21 +828,18 @@ def current_vs_supports_modules() -> bool:
         return True
     return vsver.startswith('16.9.0') and '-pre.' in vsver
 
+_VERSION_TOK_RE = re.compile(r'(\d+)|([a-zA-Z]+)')
+
 # a helper class which implements the same version ordering as RPM
 class Version:
     def __init__(self, s: str) -> None:
         self._s = s
 
-        # split into numeric, alphabetic and non-alphanumeric sequences
-        sequences1 = re.finditer(r'(\d+|[a-zA-Z]+|[^a-zA-Z\d]+)', s)
-
-        # non-alphanumeric separators are discarded
-        sequences2 = [m for m in sequences1 if not re.match(r'[^a-zA-Z\d]+', m.group(1))]
-
+        # extract numeric and alphabetic sequences
         # numeric sequences are converted from strings to ints
-        sequences3 = [int(m.group(1)) if m.group(1).isdigit() else m.group(1) for m in sequences2]
-
-        self._v = sequences3
+        self._v = [
+                int(m.group(1)) if m.group(1) else m.group(2)
+                for m in _VERSION_TOK_RE.finditer(s)]
 
     def __str__(self) -> str:
         return '{} (V={})'.format(self._s, str(self._v))