py3k-finish_time

View source
py3k-finish_time
  • Contributors
    1. Loading...
Author Commit Message Date Builds
250+ commits behind default.
Mark Young
Make all tests pass for the time module. lib-python/test_time.py still fails test_mktime_error when run untranslated, but this test is skipped when run translated anyway.
Mark Young
Fix faulty indentation.
Mark Young
Figured out I was wrongly inverting the meaning of the flag returned by QueryPerformanceCounter.
Mark Young
Get rid of HAVE_CLOCK as it should be present for all supported platforms.
Mark Young
Went through all flows and everything seems to match cpython now. Tested everything on windows and it appears to be okay. Now just to test post-translation.
Mark Young
Figuring out what was going wrong.
Mark Young
Fix a few silly typos.
Mark Young
Hopefully final tweaks to match all fallback scenarios with cpython for time.get_clock_info .
Ronan Lamy
fix translation
Ronan Lamy
hg merge default
Ronan Lamy
Fix missing wrapbytes->newbytes renamings
Ronan Lamy
Reduce diff with default branch
Ronan Lamy
Use space.newbytes() and space.newunicode()
Ronan Lamy
Create space.newunicode()
Ronan Lamy
Rename wrapbytes to newbytes, for consistency with newint, newfloat, etc.
Ronan Lamy
hg merge default
mattip
fix, enable tests
Ronan Lamy
hg merge default
Ronan Lamy
fix bad merge (yet again)
Ronan Lamy
fixes
Ronan Lamy
fix missing imports
Ronan Lamy
manually merge default's stringobject.h changes to bytesobject.h
Ronan Lamy
hg merge default
Ronan Lamy
merge branch testing-cleanup-py3k
Ronan Lamy
Update _sre.c to CPython 3.3.5 (doesn't compile because we're missing most of the PEP393 API)
Ronan Lamy
fix warnings
Philip Jenvey
Merged in marky1991/pypy_new/py3k (pull request #454) Py3k: Update Fallback Code in Time to Match CPython
Mark Young
Fixing monotonic.
Mark Young
Fixing segfault in time.monotonic on windows.
Mark Young
Get rid of wierd space.
Mark Young
Responding to feedback and fixing a couple of bugs.
Mark Young
Wrap up current work. All that's left is to figure out win_perf_counter.
Mark Young
Merge with default
Mark Young
The mentioned revision has already been pulled in.
Mark Young
Hopefully final cleanup.
Mark Young
Mord Reducing unnnecessary diffs with upstream.
Mark Young
Reducing unnnecessary diffs with upstream.
Mark Young
Merging with upstream.
Mark Young
Add fallbacks to gettimeofday.
Mark Young
Commit to ask about the error.
Mark Young
Merging with upstream.
Mark Young
Merge with upstream.
Mark Young
Committing so I can diff between mine and upstream.
Mark Young
Commit what I have for now.
Mark Young
Testing
Mark Young
Implement dynamic lookup properly.
Mark Young
Let's try this.
Mark Young
Let's try this.
Mark Young
Merge with upstream.
Mark Young
Merge with upstream.
Mark Young
Fixes and useless logs. Will remove useles logs later.
Mark Young
Merge heads
Mark Young
Trying something. (Working on my linux box so as to avoid work in cmd.exe)
Mark Young
Merge with upstream I hope.
Mark Young
Merge with upstream.
Mark Young
Commit what I have so far so I can test on windows. Everything added might not work properly. (Tests certainly don't pass)
Ronan Lamy
hg merge py3k
Ronan Lamy
hg merge testing-default
Ronan Lamy
backout a02677ec50cc: does not work, actually
Ronan Lamy
hg merge default
Ronan Lamy
Merge branch 'testing-cleanup-py3k'
Ronan Lamy
Don't attempt to modify sys.filesystemencoding; skip tests instead if it's not good enough
Ronan Lamy
Backed out changeset dc234c6c4b34 (didn't work)
Ronan Lamy
Blindly try to get tests to pass on bencher4
Ronan Lamy
hg merge testing-default
Ronan Lamy
fix after the merge
Ronan Lamy
hg merge default
Philip Jenvey
seems to work now
Ronan Lamy
hg merge default Port cpyext changes to bytearray
Ronan Lamy
hg merge default (f731f9ef48f8)
Ronan Lamy
hg merge default
Philip Jenvey
use wrap_fsdecoded
Philip Jenvey
fix certain bootstrap issues w/ fsdecoding (refs #2300): bootstrap check for delayed filesystemencoding initialization after eb02742ce71d
mattip
update whatsnew for pypy3
Manuel Jacob
2to3
Philip Jenvey
only encode unicode (grafted from 30cf97b5462bf8d9e94cd8946b2871877ca3197d)
Philip Jenvey
more fsdecode on paths
Philip Jenvey
merge py3k-get_clock_info adds time.get_clock_info for posix, windows is still a work in progress
Philip Jenvey
fill_clock_info -> _setinfo
Philip Jenvey
cleanup the in progress windows code
Philip Jenvey
fix platform detection
Philip Jenvey
older glibc needs -lrt for clock_getres and friends
Philip Jenvey
woops
Philip Jenvey
naming
Philip Jenvey
merge the non darwin posix monotonic()s
Philip Jenvey
bring over test_time's test
Philip Jenvey
kill
Philip Jenvey
whitespace
Philip Jenvey
fixes/use _timespec_to_seconds
Philip Jenvey
plug clock_gettime into time()
Philip Jenvey
probably a temporary get_clock_info for posix via host time.time
Philip Jenvey
cleanup
Philip Jenvey
add get_clock_info('process_time')
Philip Jenvey
get tests running
Philip Jenvey
less confusing branch name
Philip Jenvey
merge py3k
Mark Young
Commit what I have for now.
Mark Young
Testing
Mark Young
Implement dynamic lookup properly.
Mark Young
Let's try this.
Mark Young
Let's try this.
Mark Young
Merge with upstream.
Mark Young
Merge with upstream.
Mark Young
Fixes and useless logs. Will remove useles logs later.
Mark Young
Merge heads
Mark Young
Trying something. (Working on my linux box so as to avoid work in cmd.exe)
Mark Young
Merge with upstream I hope.
Mark Young
Merge with upstream.
Mark Young
Commit what I have so far so I can test on windows. Everything added might not work properly. (Tests certainly don't pass)
Philip Jenvey
branch for clock_get_info
Philip Jenvey
add missing msvc includes for lipmpdec
Philip Jenvey
a few more fsdecodes
Philip Jenvey
fix
Philip Jenvey
more fsdecoding fixes
Philip Jenvey
fsdecode/encode in zip_dict
Philip Jenvey
apply da4861ab2c56
Philip Jenvey
fix the error message
Philip Jenvey
fsencode zipimport filenames
Philip Jenvey
fix, we can't use **kwargs in this case
Philip Jenvey
fix
Philip Jenvey
fix
Philip Jenvey
fix
Philip Jenvey
add new funcs for subprocess. there's more needed for multiprocessing
Philip Jenvey
merge default
Philip Jenvey
_subprocess -> _winapi
Philip Jenvey
refactor into do_utimens, parse_utime_args
Philip Jenvey
oups
Philip Jenvey
win32 translation fix
Philip Jenvey
--version still prints to stdout in 3.3 (changed to stderr in 3.4)
Philip Jenvey
run test_ensurepip
Philip Jenvey
ensurepip from the latest cpython w/ the latest pip/setuptools
Philip Jenvey
rearrange
Philip Jenvey
try to cleanup by unfortunately duplicating call_rposix here
Philip Jenvey
add futimes support
Philip Jenvey
refactor into do_utimes
Philip Jenvey
merge default
Philip Jenvey
add 32bit only unsafe setters for test_decimal
Philip Jenvey
need unicode0_w here
Philip Jenvey
fix
Philip Jenvey
kill dead code
Philip Jenvey
sys.version isn't boostrapped yet when freezing app_main (win32)
Philip Jenvey
fix
mattip
fixes for cpyext tests
Philip Jenvey
Backed out changeset 0587b09200f1 reenable faulthandler for now, regrtest depends on it
Philip Jenvey
adjust per py3k and list/tuples now need hardcoding (grafted from bc1be9c89004d3d5a494648b8064a67793c49bcd)
Philip Jenvey
fix l/utimes args
Philip Jenvey
Philip Jenvey
fix on python3 -A
Philip Jenvey
wrong default
Philip Jenvey
fallback to a/mtime_s to basic utime when ns is specified, fixup lutimes
Philip Jenvey
merge default
Philip Jenvey
always ensure _winreg for importlib
Philip Jenvey
o utilize lutimes to fix follow_symlinks=False on osx o fix utime(None) on osx
Philip Jenvey
merge default
Philip Jenvey
oops
Philip Jenvey
handle #ifndef AT_FDCWD
Philip Jenvey
missed a couple more osx (10.9?) translation fixes
Philip Jenvey
more osx (10.9?) translation fixes
Philip Jenvey
osx (10.9?) fixes
Philip Jenvey
fix _DirFD_Unavailable handling of None/not specified
Ronan Lamy
Prevent duplicate winreg module creation on win32
Philip Jenvey
skip without hypothesis installed
Philip Jenvey
fix from probably a bad merge: the default branch removed this check
Philip Jenvey
this workaround is now disabled and seemingly no longer necessary
Philip Jenvey
reapply lost sysconfig changes from old py3k, add a skip
Philip Jenvey
Backed out changeset ba47fac77ffc, this is still needed unfortunately
Philip Jenvey
kill these, importlib handles them now (and we don't pass)
Philip Jenvey
skip for now: it's pretty annoying to fix and not that important
Philip Jenvey
fix test_qualname
Philip Jenvey
merge default
Philip Jenvey
rearrange
Philip Jenvey
o print out PYTHON3 version info in pypy/interpreter/test pytest header o try to more gracefully fail app_main tests when no PYTHON3 is found
Philip Jenvey
cleanup TempoaryDirectorys atexit
Philip Jenvey
kill -R/update usage
Raffael Tfirst
Fix test_gendfa with correct whitespaces
Richard Plangger
missing 3 in states_double3DFA
Raffael Tfirst
Merge
Raffael Tfirst
Fix whitespaces printed in gendfa
Philip Jenvey
rearrange
Philip Jenvey
space.is_true/space.issubtype -> space.issubtype_w
Philip Jenvey
merge default
Raffael Tfirst
Apply gendfa patch from fix-gen-dfa (rev 84575) of Richard Plangger to py3k
Philip Jenvey
utilize enumerate
Philip Jenvey
merge default
Philip Jenvey
kill comment
Philip Jenvey
fix test_context_with_suppressed, always reset the last exception after __exit__, even if it raises
Philip Jenvey
backout 25989b3230e5: this bootstrap hack is still required (CPython also does it) to avoid obscure recursion issues, e.g. test_importhooks.testImpWrapper under linux w/ LANG=C. StdErrPrinter should make the imports safe now under -v
Philip Jenvey
__qualname__ for getset_descriptors
Philip Jenvey
no longer subclassable
mattip
clean up _winreg, winreg problems to get tests running
Philip Jenvey
zipimport namespace pkg support, hopefully
Philip Jenvey
py3 switches to socket.timeout exceptions
Philip Jenvey
followup 1a0b1ff00ed7: fsdecode on the way out
Philip Jenvey
error msg impl details
Philip Jenvey
fix _finalize_ (register it)
Philip Jenvey
fix translation maybe
Manuel Jacob
hg merge
Manuel Jacob
hg merge default (+ fixes) Some tests still fail. E.g. test_verbose_flag_* in pypy/module/imp/test/test_import.py and test_sre in pypy/module/cpyext/test/test_typeobject.py.
Manuel Jacob
Fix test after argument error messages are closer to CPython now.
mattip
add replace() for win32 (grafted from 54617a9d23bb8b70763929b214ab24eba547ccc3)
Philip Jenvey
no longer valid
Philip Jenvey
fix translation
Philip Jenvey
slot shadowing now a ValueError
Philip Jenvey
skip some zipimport tests when testfn_unencodable/special_char aren't available and make special_char available to test_zipimport
Philip Jenvey
cpython only (pypy disallows non str key values)
Mark Young
Review changes.
Mark Young
Merge from upstream.
Mark Young
Respond to feedback. Matches cpython's intent more closely.
Mark Young
Fix test_inspect.
Mark Young
Restore 975377190519 and 177135a9fa92, making the two pickle test suites pass.
Philip Jenvey
cpython issue2382: adjust SyntaxError offset w/ multibyte chars
Philip Jenvey
adjust offsets per pypy impl details
Philip Jenvey
reapply lost changes from default/pypy3.2
Philip Jenvey
cpython issue17413: normalize exceptions for settrace callbacks
Philip Jenvey
cpython TypeError impl detail
Philip Jenvey
pypy3 still adheres to the oldgil interface
Philip Jenvey
apply c4c19a588369 from default
Philip Jenvey
merge default
Ronan Lamy
fix translation
Ronan Lamy
fix another compiler warning
Ronan Lamy
Fix bad merge and compilation warnings in test_typeobject.py
Ronan Lamy
Fix some cpyext compilation warnings
Ronan Lamy
This test shouldn't be cpython_only
Philip Jenvey
merge default
Ronan Lamy
fix translation maybe
Philip Jenvey
missing import
Ronan Lamy
hg merge default
Philip Jenvey
reapply xfails from default
Philip Jenvey
adapt to py3k
Philip Jenvey
kill SSLContext.check_hostname, it's for 3.4 (or modern 2.9) ssl.py. exposing it gives the impression that we provide a 3.4 ssl.py that uses it to do ssl hostname matching in do_handshake, e.g.: https://github.com/python/asyncio/blob/309a218/asyncio/selector_events.py#L828 (without this change, this code never matches hostnames!)
Philip Jenvey
fix after merge
Philip Jenvey
kill test_immutabledoc
Philip Jenvey
o set __main__ loader to SourceFileLoader like cpython o workaround subtle test_cmd_line_script impl details
Philip Jenvey
merge
Mark Young
Match cpython's inconsistent warning about the now-deprecated unicode_internal codec.
Philip Jenvey
Close branch py3k
Philip Jenvey
allow setting class docstrings
Philip Jenvey
fix
Philip Jenvey
__qualname__ must be removed from the type dict so it doesn't propagate down to the instance
Philip Jenvey
now an OSError on 3.3
Mark Young
Merged in marky1991/pypy_new/py3k (pull request #444) Py3k Finish Deque Fix
Mark Young
Fix the message to report the type correctly.
Mark Young
Fix deque as I meant to do before. Added tests specifically testing pickleability of deque iterators.
Philip Jenvey
match cpython's frozen importlib name (it's still exposed in some cases)
Philip Jenvey
a failing test for __qualname__
Philip Jenvey
Merged in marky1991/pypy_new/py3k (pull request #442) py3k Deque Fix
Mark Young
Move deque fixes to py3k so I don't have to merge with upstream again.
Ronan Lamy
Fix test_stressdict
Philip Jenvey
oops
Philip Jenvey
reapply lost 2.7/3.2 workarounds
3925 commits not shown.