mirrors/cpython
1
0
Fork 0
mirror of https://github.com/python/cpython.git synced 2025-08-04 17:08:35 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 13

gh-118761: Improve the import time of `gettext` (#128898)

Browse source 
``gettext`` is often imported in programs that may not end up translating
anything. In fact, the ``struct`` module already has a delayed import when
parsing ``GNUTranslations`` to speed up the no ``.mo`` files case. The re module
is also used in the same situation, but behind a function chain only
called by ``GNUTranslations``.

Cache the compiled regex globally the first time it is used. The
finditer function is converted to a method call on the compiled
object which is slightly more efficient, and necessary for the
delayed re import.
This commit is contained in:
Eli Schwartz 2025-01-19 19:01:20 -05:00 • committed by GitHub
parent bbeb219354
commit c9c9fcb8fc
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
2 changed files with 21 additions and 15 deletions
Download patch file Download diff file Expand all files Collapse all files

33
Lib/gettext.py
View file

@ -48,7 +48,6 @@ internationalized, to the local language and cultural habits.
import operator
import os
import re
import sys
@ -70,22 +69,26 @@ _default_localedir = os.path.join(sys.base_prefix, 'share', 'locale')
# https://www.gnu.org/software/gettext/manual/gettext.html#Plural-forms
# http://git.savannah.gnu.org/cgit/gettext.git/tree/gettext-runtime/intl/plural.y
_token_pattern = re.compile(r"""
(?P<WHITESPACES>[ \t]+) | # spaces and horizontal tabs
(?P<NUMBER>[0-9]+\b) | # decimal integer
(?P<NAME>n\b) | # only n is allowed
(?P<PARENTHESIS>[()]) |
(?P<OPERATOR>[-*/%+?:]|[><!]=?|==|&&|\|\|) | # !, *, /, %, +, -, <, >,
# <=, >=, ==, !=, &&, ||,
# ? :
# unary and bitwise ops
# not allowed
(?P<INVALID>\w+|.) # invalid token
""", re.VERBOSE|re.DOTALL)
_token_pattern = None
def _tokenize(plural):
for mo in re.finditer(_token_pattern, plural):
global _token_pattern
if _token_pattern is None:
import re
_token_pattern = re.compile(r"""
(?P<WHITESPACES>[ \t]+) | # spaces and horizontal tabs
(?P<NUMBER>[0-9]+\b) | # decimal integer
(?P<NAME>n\b) | # only n is allowed
(?P<PARENTHESIS>[()]) |
(?P<OPERATOR>[-*/%+?:]|[><!]=?|==|&&|\|\|) | # !, *, /, %, +, -, <, >,
# <=, >=, ==, !=, &&, ||,
# ? :
# unary and bitwise ops
# not allowed
(?P<INVALID>\w+|.) # invalid token
""", re.VERBOSE|re.DOTALL)
for mo in _token_pattern.finditer(plural):
kind = mo.lastgroup
if kind == 'WHITESPACES':
continue

3
Misc/NEWS.d/next/Library/2025-01-15-19-16-50.gh-issue-118761.cbW2ZL.rst Normal file
View file

@ -0,0 +1,3 @@
Reduce import time of :mod:`gettext` by up to ten times, by importing
:mod:`re` on demand. In particular, ``re`` is no longer implicitly
exposed as ``gettext.re``. Patch by Eli Schwartz.