Carl Friedrich Bolz-Tereick pushed to branch branch/default at PyPy / pypy
Commits: 5ae1bcb2 by Carl Friedrich Bolz-Tereick at 2020-04-04T15:28:44+02:00 use more "official" methods on the W_UnicodeObject to do conversion --HG-- branch : optimize-sre-unicode - - - - - e78e6b4e by Carl Friedrich Bolz-Tereick at 2020-04-04T15:38:35+02:00 reuse the string we already have, no need to re-wrap it --HG-- branch : optimize-sre-unicode - - - - - 77b7bc5f by Carl Friedrich Bolz-Tereick at 2020-04-04T15:59:01+02:00 add an ascii optimization: for ascii unicode strings, use the much faster regular byte matching engine --HG-- branch : optimize-sre-unicode - - - - - 009bd570 by Carl Friedrich Bolz-Tereick at 2020-04-04T16:36:04+02:00 use an ascii-fastpath for getlower and getupper --HG-- branch : optimize-sre-unicode - - - - - 3743aefe by Carl Friedrich Bolz-Tereick at 2020-04-04T21:23:43+02:00 write getlower/upper_ascii in a branch-free way to get less bridges (note that the multiplication * (ord('a') - ord('A')) actually turns into a shift by 5, since that is 32) --HG-- branch : optimize-sre-unicode - - - - - 7d8b39af by Carl Friedrich Bolz-Tereick at 2020-04-04T21:29:02+02:00 test for f0dd6cd6eec2 --HG-- branch : optimize-sre-unicode - - - - - f07956ea by Carl Friedrich Bolz-Tereick at 2020-04-05T13:19:44+02:00 merge default --HG-- branch : optimize-sre-unicode - - - - - 4e2d67a5 by Carl Friedrich Bolz-Tereick at 2020-04-05T16:38:01+02:00 move the flags from the contexts to the CompiledPatterns --HG-- branch : optimize-sre-unicode - - - - - c209237d by Carl Friedrich Bolz-Tereick at 2020-04-05T16:38:58+02:00 oops, this is no longer needed --HG-- branch : optimize-sre-unicode - - - - - 2e80750f by Carl Friedrich Bolz-Tereick at 2020-04-10T15:41:44+02:00 merge default --HG-- branch : optimize-sre-unicode - - - - - d21a77f6 by Carl Friedrich Bolz-Tereick at 2020-04-10T15:43:16+02:00 document branch --HG-- branch : optimize-sre-unicode - - - - - c5bf86cd by Carl Friedrich Bolz-Tereick at 2020-04-10T15:44:58+02:00 merge optimize-sre-unicode: speed up the performance of matching unicode strings in the sre engine: - there is a fast path for the case where the whole string is ASCII - case-insensitive matching of ASCII characters is much faster both for unicode and bytes matching - - - - - 13 changed files: - pypy/doc/whatsnew-head.rst - pypy/module/_sre/interp_sre.py - pypy/module/_sre/test/test_app_sre.py - pypy/module/_sre/test/test_sre_buf.py - rpython/rlib/rsre/rpy/_sre.py - rpython/rlib/rsre/rsre_char.py - rpython/rlib/rsre/rsre_core.py - rpython/rlib/rsre/rsre_re.py - rpython/rlib/rsre/rsre_utf8.py - rpython/rlib/rsre/test/support.py - rpython/rlib/rsre/test/test_char.py - rpython/rlib/rsre/test/test_ext_opcode.py - rpython/rlib/rsre/test/test_zjit.py View it on GitLab: https://foss.heptapod.net/pypy/pypy/compare/024cbd9a1f2835405095133eb6fba7b4edd86851...c5bf86cd92329ac5f5cee2bf6459a585aa7f35b3 --- View it on GitLab: https://foss.heptapod.net/pypy/pypy/compare/024cbd9a1f2835405095133eb6fba7b4edd86851...c5bf86cd92329ac5f5cee2bf6459a585aa7f35b3 You're receiving this email because of your account on foss.heptapod.net.
_______________________________________________ pypy-commit mailing list [email protected] https://mail.python.org/mailman/listinfo/pypy-commit
