cfbolz changed the topic of #pypy to: PyPy, the flexible snake (IRC logs: https://quodlibet.duckdns.org/irc/pypy/latest.log.html#irc-end ) | use cffi for calling C | if a pep adds a mere 25-30 [C-API] functions or so, it's a drop in the ocean (cough) - Armin
CrazyPython has joined #pypy
jcea has quit [Quit: jcea]
tsaka_ has joined #pypy
CrazyPython has quit [Read error: Connection reset by peer]
speeder39_ has quit [Quit: Connection closed for inactivity]
tsaka_ has quit [Ping timeout: 258 seconds]
xcm has quit [Remote host closed the connection]
xcm has joined #pypy
andi- has quit [Remote host closed the connection]
andi- has joined #pypy
lritter has quit [Ping timeout: 276 seconds]
lritter_ has joined #pypy
dddddd has quit [Remote host closed the connection]
speeder39_ has joined #pypy
dansan has quit [Read error: Connection reset by peer]
dansan has joined #pypy
glyph has quit [Quit: End of line.]
glyph has joined #pypy
mattip has quit [Ping timeout: 240 seconds]
xcm has quit [Remote host closed the connection]
xcm has joined #pypy
mattip has joined #pypy
mattip has quit [Client Quit]
mattip has joined #pypy
mattip has quit [Quit: ZNC 1.6.6+deb1ubuntu0.2 - http://znc.in]
mattip has joined #pypy
mattip has quit [Client Quit]
mattip has joined #pypy
xcm has quit [Remote host closed the connection]
xcm has joined #pypy
xcm has quit [Remote host closed the connection]
xcm has joined #pypy
speeder39_ has quit [Quit: Connection closed for inactivity]
antocuni has joined #pypy
tsaka_ has joined #pypy
<kenaan> cfbolz default 28102f96dc5c /: for some reason we have two rutf8 functions that do the same thing, codepoints_in_utf8 and get_utf8_length. The fo...
<kenaan> cfbolz py3.6 a94f909131d8 /: merge default
xcm has quit [Remote host closed the connection]
xcm has joined #pypy
<kenaan> cfbolz default 6e05da20e7bd /pypy/objspace/std/unicodeobject.py: call prev_codepoint_pos only once
xcm has quit [Remote host closed the connection]
xcm has joined #pypy
ronan__ has joined #pypy
ronan__ has quit [Quit: Ex-Chat]
ronan has joined #pypy
ronan has quit [Ping timeout: 265 seconds]
xcm has quit [Remote host closed the connection]
antocuni has quit [Ping timeout: 258 seconds]
<kenaan> andrewjlawrence winconsoleio fa6bd070da3d /: Fixed a bunch of translation issues
xcm has joined #pypy
ajlawrence has joined #pypy
dddddd has joined #pypy
ajlawrence has quit [Ping timeout: 260 seconds]
ajlawrence has joined #pypy
<kenaan> cfbolz py3.6 03c11d539a67 /pypy/interpreter/: fix error message of barry_as_FLUFL, important work!
<kenaan> cfbolz py3.6 c4bc6937bce1 /pypy/interpreter/test/test_compiler.py: unskip these two tests
<bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/4819 [Andy Lawrence: force build, winconsoleio]
tsaka_ has quit [Ping timeout: 268 seconds]
ajlawrence has quit [Remote host closed the connection]
ronan__ has joined #pypy
tsaka_ has joined #pypy
antocuni has joined #pypy
ronan__ has quit [Ping timeout: 265 seconds]
<bbot2> Exception: http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/4819 [Andy Lawrence: force build, winconsoleio]
<mattip> ajlawrence: something is not right with stderr/stdout
xcm has quit [Remote host closed the connection]
<kenaan> cfbolz py3.6 cedd17bf511f /pypy/interpreter/astcompiler/: don't fold too huge left shift either
<kenaan> cfbolz py3.6 ce752f8d7d1e /lib-python/3/test/test_readline.py: use getattr here as well (already done for another test)
<kenaan> cfbolz py3.6 a741689e58e9 /lib-python/3/test/test_regrtest.py: this restriction was lifted, and we didn't import the test change correctly or something
<cfbolz> ouch, our poll test is passing completely by accident
xcm has joined #pypy
<mattip> it seems py3.6 own tests have not been run on windows in quite a while. They _io ones import fcntl
<mattip> so of course ajlawrence could not run own tests
xcm has quit [Remote host closed the connection]
xcm has joined #pypy
ronan__ has joined #pypy
marky1991 has joined #pypy
Dejan has joined #pypy
xcm is now known as Guest43262
Guest43262 has quit [Killed (cherryh.freenode.net (Nickname regained by services))]
xcm has joined #pypy
ronan__ has quit [Ping timeout: 265 seconds]
<marmoute> mattip: https://ci.octobus.net/job/MercurialPypy/362/console ← can you make something out of this.
<mattip> marmoute: maybe "Unknown exception encountered with possibly-broken third-party extension drawdag"?
<mattip> are you using pypy HEAD or some older version?
<marmoute> mattip: not sure which version
<marmoute> ** Python 2.7.13 (7.0.0+dfsg-2~bpo9+1, Mar 15 2019, 17:22:43)[PyPy 7.0.0 with GCC 6.3.0 20170516]
<mattip> ahh, I see. PyPy 7.0.0. Please update, there were some issues around some zlib or bzlib support
<marmoute> mattip: which version should we target ?
marky1991 has quit [Remote host closed the connection]
marky1991 has joined #pypy
<marmoute> looks like 7.1.1 is the latest
<mattip> nightly would be better, from http://buildbot.pypy.org/nightly/trunk/
<mattip> we really should release soon, but have that "just a little more" thing going right now
<mattip> yup, look for zlib on the changelog http://doc.pypy.org/en/latest/release-v7.1.0.html#changelog
<mattip> "Add copying to zlib"
<Dejan> And update the website - that new logo is awesome :)
<kenaan> mattip py3.6 b6cec0565e9d /pypy/module/_io/test/: enable _io own tests on windows
<kenaan> mattip winconsoleio 905c730547a7 /: merge py3.6 into branch
<marmoute> mattip: good news, the hg-git crash comes from a corruption specific to my local repo
<cfbolz> antocuni, arigato: hi! Does either of you feel like reviewing my json branch in the next days?
<antocuni> cfbolz: sure. Probably not today though
<antocuni> is it ready or you still need to finish it?
<cfbolz> antocuni: it's ready, I think
<cfbolz> (there are improvements possible)
<antocuni> cool, I'll look at it tomorrow
<cfbolz> antocuni: awesome, thanks
<Dejan> is csv module that is distributed with pypy a copy of the upstream one?
<Dejan> i hoped pypy has own implementation that is faster...
<cfbolz> Dejan: the CPython one is written in c
<cfbolz> We ha e our own, it's not a copy, but I doubt that it's faster
tsaka_ has quit [Ping timeout: 265 seconds]
<Dejan> it is terribly slow
BPL has joined #pypy
BPL has quit [Remote host closed the connection]
<cfbolz> Dejan: as in, slower than CPython?
<Dejan> slower than ... a CSV reader I have implemented in-house in D
<Dejan> by slower I mean good 10-15% slower
marky1991 has quit [Ping timeout: 240 seconds]
ajlawrence has joined #pypy
<ajlawrence> mattip: What was going wrong with stdin/stderr in those tests?
marky1991 has joined #pypy
<ajlawrence> mattip: I am not sure what I am looking for in the logs
<mattip> ajlawrence: I added a commit so the own _io tests run (and fail) like after translation
<mattip> so if you hg up, you can run "python2 pytest.py pypy\module\_io\test\test_bufferedio.py"
<mattip> which will fail with something about expecting an int and getting a str
<mattip> when doing open(fname)
<ajlawrence> OK will do. Is that a bug on windows or something that has not been implemented?
<mattip> it seems to be a bug in windows, but I have not been able to give it my full attention just now
<mattip> in general, we should be running the own tests on py3.6 windows, but as it is the buildbot is barely keeping up
<mattip> ajlawrence: got it. adding --pdb to the test command line is magic
ronan__ has joined #pypy
<mattip> it points to interp_winconsoleio.py line 122, where you are doing space.int_w(w_path_or_fd), but it is a path
jcea has joined #pypy
<ajlawrence> In CPython if it is a path then the result is a negative number.
<mattip> rpython is a bit different, you can try/except or check first
ajlawrence27 has joined #pypy
<ajlawrence27> mattip: What type of exception is being raised?
<mattip> you should do "if space.isinstance_w(w_path_or_fd, space.w_int))" rather than try/except
ajlawrence has quit [Remote host closed the connection]
<mattip> since otherwise you need to do "except OperationError as e" and then check the type of the error, which is uneccssarily expensive
<ajlawrence27> OK understood.
<mattip> thanks for persevering
<mattip> marmoute: anyway you would be better off with latest HEAD
<kenaan> andrewjlawrence winconsoleio d88f85fb9795 /: Implemented write method for the Windows console io class.
<marmoute> mattip: maybe, but installing the laster debian package is simpler for now
<ajlawrence27> mattip: It is a good learning experience.
<mattip> marmoute: wget ... ; tar -xf ...; path/to/pypy/bin/pypy-c -m ensurepip
<mattip> ajlawrence27: instead of lltype.malloc you can use with lltype.scoped_alloc as a context manager, then the "free" is done for you
<ajlawrence27> Can the memory allocated in a scoped_alloc be reallocated? The buffer changes size in the windows console io code.
marky1991 has quit [Ping timeout: 240 seconds]
<mattip> ahh, right. Then the call on line 167 still uses the old length?
<ajlawrence27> That is a mistake
<mattip> :)
antocuni has quit [Ping timeout: 265 seconds]
<cfbolz> Dejan: if it's much slower than CPython it's worth a bug report
BPL has joined #pypy
BPL has quit [Remote host closed the connection]
ajlawrence27 has quit [Remote host closed the connection]
ajlawrence has joined #pypy
<Dejan> I wonder how far are they with it
xcm has quit [Ping timeout: 265 seconds]
xcm has joined #pypy
<kenaan> mattip winconsoleio bf4120ce59b6 /pypy/module/_io/interp_win32consoleio.py: fix copy-paster
xcm has quit [Remote host closed the connection]
xcm has joined #pypy
ajlawrence has quit [Ping timeout: 260 seconds]
BPL has joined #pypy
ronan__ has joined #pypy
<kenaan> mattip winconsoleio 56316fa56a96 /pypy/module/_io/interp_win32consoleio.py: rewrite in pure python. "decoded" is a ascii string AFAICT
<bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/4820 [mattip: force build, winconsoleio]
<kenaan> andrewjlawrence winconsoleio 01ae8b151011 /pypy/module/_io/interp_win32consoleio.py: Fixed issues
<kenaan> andrewjlawrence winconsoleio 0e928fe22992 /pypy/module/_io/interp_win32consoleio.py: Merge
ajlawrence has joined #pypy
jacob22 has quit [Read error: Connection reset by peer]
jacob22 has joined #pypy
<ajlawrence> mattip: opps I seem to have merged out your changes
<ajlawrence> Do we not need the code in the middle of that block where it looks up the full pathname?
<kenaan> andrewjlawrence winconsoleio dfec1e76b2be /pypy/module/_io/interp_win32consoleio.py: Merged discarded changes back in.
xcm has quit [Ping timeout: 276 seconds]
<ajlawrence> mattip: Is it also possible to replace the calls to WideCharToMultiByte with pure Python?
<mattip> ajlawrence: not clear to me. Is there a chance the (potentially not full) path name would not include the magic strings?
<mattip> sorry that was in response to "looks up the full pathname"
xcm has joined #pypy
<ajlawrence> mattip: Is it possible to have a relative path to one of these things?
<mattip> about WideCharToMultiByte, are you sure read_console_w is returning wchar buffers?
<ajlawrence> Yes
<mattip> then I think we need the call
<mattip> Here is the issue that led to the GetFullPath code, I am not sure I understand the nuaces
<ajlawrence> It is possible to have weird console paths like C:/Temp/conout$ therefore they need to be normalized.
<mattip> it seems the behaviour in that issue changes over different versions of windows
<mattip> where would a path like that come from in the first place?
<mattip> if I do " notepad c:\temp\conout$" I get a messgage "Cannot create \\.\conout$.txt file"
<mattip> but who would try to create a file like that?
<ajlawrence> I could imagine my cat walking on the keyboard and doing it but I am not sure it is something that would occur during normal usage.
<mattip> if my fix holds for most of the tests, we can revisit this later
<ajlawrence> There is a translation error now
<ajlawrence> I am not sure what is causing it
<mattip> ?? the buildbot is happy with 56316fa56a96 http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/4820
<mattip> it even uploaded a package
<ajlawrence> There was a return None blocking a path through the code.
<mattip> so that should be returning a void pointer rffi.cast(rffi.VOIDP, 0)
<mattip> (or something like that)
* mattip zzz
tsaka_ has joined #pypy
ajlawrence has quit [Remote host closed the connection]
ajlawrence has joined #pypy
adamholmberg has joined #pypy
<kenaan> cfbolz py3.6 f41451c91702 /pypy/module/select/: un-mess argument handling and test of select.poll().poll(): previously the test was passing completely by accident, ...
<kenaan> cfbolz py3.6 7e303c8f90af /pypy/module/__builtin__/functional.py: update docstrings
ajlawrence has quit [Remote host closed the connection]
ajlawrence has joined #pypy
ajlawrence has quit [Remote host closed the connection]
jcea has quit [Remote host closed the connection]
jcea has joined #pypy
CrazyPython has joined #pypy
CrazyPython has quit []
adamholmberg has quit [Remote host closed the connection]
abrown has quit [Remote host closed the connection]
<marmoute> mattip: using a non-corrupted repository source, the git export worked fine :-D
<marmoute> Now trying what would be heptapod next step during that import (pushing to git, using special ref logic to expose branches).
abrown has joined #pypy
lritter_ has quit [Ping timeout: 268 seconds]
BPL has quit [Quit: Leaving]
marky1991 has joined #pypy
speeder39_ has joined #pypy
_whitelogger has joined #pypy
hexa- has left #pypy ["WeeChat 2.6"]