Commits

Show all
Author Commit Message Labels Comments Date
Philip Jenvey
this eq_w doesn't seem necessary any longer and is broken (doesn't try is_w first)
Philip Jenvey
convert to formal app level tests so they have a chance of running on the py3k branch. they'll run slower but that can be avoided w/ appdirect mode
Philip Jenvey
py3k branch compat.
Philip Jenvey
fill in is_none
Alex Gaynor
added statvfs_result to posix
Armin Rigo
Kill this now-not-useful variable
Armin Rigo
Test for the "except:" path. Tests that the exception class is only instantiated once, and fix.
Armin Rigo
Fix the second test.
Armin Rigo
Fix the first of the two tests
Armin Rigo
More tests, checked to work on CPython's greenlet
Armin Rigo
Import cffi/b3729934bf48
Armin Rigo
int_rshift needs to be there too, because 'n >> m' at app-level checks for more
Carl Friedrich Bolz
kill no longer needed constant
Carl Friedrich Bolz
somewhat annoying hack: if digits are the default base 10 ones, use a specialized recursive function that calls the builtin RPython str method. Helps a lot, though.
Carl Friedrich Bolz
compute mindigits based on the base, instead of taking a worst-case number
Carl Friedrich Bolz
woops, fix test
Carl Friedrich Bolz
use str if fitting into ints
Carl Friedrich Bolz
make small str(small long) much faster by caching the powers of ten needed
Armin Rigo
int_lshift, uint_rshift
Armin Rigo
Use a consistent style and avoid intmask()
Armin Rigo
int_mod
Armin Rigo
Cannot give uint_floordiv a useful sense without an "unsigned" type
Armin Rigo
int_floordiv
Armin Rigo
int_mul
Armin Rigo
int_sub
Armin Rigo
Might go away again, but: attempting to give app-level code direct access to the wrap-around C-ish versions of some arithmetic operations.
David Schneider
import updated _cffi_backend/test/_backend_test_c.py
David Schneider
mention package-tk branch in whatsnew-2.1
Armin Rigo
Fix test_whatsnew
Armin Rigo
Skip this test which was never working, with a comment
  1. Prev
  2. Next