Author: afa
Revision Date: 2010-11-28 22:35:48 UTC

Implement the sq_contains slot,
and PyStructSequence objects

Author: magcius
Revision Date: 2010-06-24 05:27:30 UTC

Merged from trunk

Author: afa
Revision Date: 2010-04-20 23:46:01 UTC

PyString_Format

Author: arigo
Revision Date: 2010-03-20 14:40:22 UTC

Fix.

Author: fijal
Revision Date: 2009-03-20 07:49:34 UTC

an in-progress checkin of changes so-far

Author: cfbolz
Revision Date: 2010-01-26 14:36:20 UTC

mark character finding as pure. at some point we should do that systematically
about string functions.

Author: afa
Revision Date: 2009-04-24 16:17:08 UTC

Fix os.getcwdu() to return a unicode as advertised.
This fixes the last failure on Windows!

Author: cfbolz
Revision Date: 2009-06-27 17:26:40 UTC

refactor way the dict_w attribute of types is accessed

Author: arigo
Revision Date: 2005-06-09 20:23:30 UTC

Copied from the head.

Author: arigo
Revision Date: 2009-11-20 19:37:44 UTC

Don't crash when printing a FailedToImplement error.

Author: xoraxax
Revision Date: 2009-12-09 11:15:29 UTC

Add example md5demo module.

Author: afa
Revision Date: 2010-01-22 16:32:50 UTC

A test that I would like to pass:
a Controller should be able to support method calls on the Controlled object.

Help needed!

Author: arigo
Revision Date: 2009-12-20 12:19:50 UTC

Progress.

Author: arigo
Revision Date: 2006-03-23 10:30:47 UTC

Typo (thanks Ben Young). Quite hard to test because CPython
eats all exceptions in __eq__ called by the dict lookup
algorithm! Bad CPython!

Author: arigo
Revision Date: 2007-09-04 11:39:50 UTC

If a CConfig has only 'rffi_platform.Has' entries, and if the
CConfig tries to #include non-existing headers, then it is not
a fatal error but simply an indication that the platform does
not have the functions or names in question.

Author: arigo
Revision Date: 2007-12-10 11:17:45 UTC

The ReadBZ2Filter acts like a read buffer itself, so we don't need
to layer another one on top of it - at least if the peek() method
is present.

Author: arigo
Revision Date: 2004-09-06 13:53:14 UTC

Some more tweaking reference counters. The resulting C code looks reasonable
now -- it tries hard to minimize the Py_{INC,DEC}REF().

Moved a couple of private tests to test_ctrans.py. I'll try to figure out
later why you cannot call two test functions 'fn' without the two Translator
instances getting confused.

 --This line, and those below, will be ignored--

M translator/test/test_ctrans.py
M translator/typer.py

Author: igorto
Revision Date: 2009-04-26 13:17:13 UTC

a simple example using reflex and pypy(i still need to create a Makefile)

Author: arigo
Revision Date: 2003-07-07 11:40:22 UTC

reorganizing branch, part 1

Author: cfbolz
Revision Date: 2009-04-20 13:27:19 UTC

replace it with the version of the old branch

Author: pedronis
Revision Date: 2009-09-04 15:12:22 UTC

(micke, pedronis) disambiguate the usage of _x86_stack_depth minimizing confusion, one overlapping and potentially
dangerous usage was test only

Author: misto
Revision Date: 2007-12-12 21:07:04 UTC

added test files and temp refactoring modules

Author: arigo
Revision Date: 2010-03-11 12:41:07 UTC

Merge branch/multijit. Start to fix a few test failures. There are more.

Author: dan
Revision Date: 2009-12-30 08:21:02 UTC

This broke all tests, keeping the code alive in this branch though.

Author: electronicru
Revision Date: 2010-03-11 09:13:27 UTC

Added many tests and fixed many old errors for sd/mdarray;
fixed Ellipsis behavior, None behavior in indices; newaxis added and tested.
TODO: multiindexing sdarrays.

Author: arigo
Revision Date: 2007-09-05 11:26:41 UTC

Repeat the deletion of these two files that occurred on the
pypy-more-rtti-inprogress branch.

Author: pedronis
Revision Date: 2005-06-07 14:40:18 UTC

oops, debug print

Author: arigo
Revision Date: 2007-11-20 19:15:15 UTC

Revert unintended change.

Author: cfbolz
Revision Date: 2010-02-17 14:23:04 UTC

I would like to pass this function

Author: fijal
Revision Date: 2008-06-01 15:35:31 UTC

* Add an explicit free and pass a test
* Add exception_is_here/cannot_occur

Author: arigo
Revision Date: 2006-04-04 14:11:13 UTC

Fixed imports.

Author: fijal
Revision Date: 2010-01-18 14:31:48 UTC

(pedronis, fijal) A tentative checkin, try to kill frame.is_being_profiled.

Author: arigo
Revision Date: 2010-03-14 16:40:05 UTC

Slowly learning the proper way of unit tests: make codebuf
independent of rx86, including tests.

Author: arigo
Revision Date: 2010-03-08 16:27:52 UTC

Intermediate stage. Random progress, but I still need to write tests
and implement code that notes where in the assembler the
convert_to_imm(ConstPtr(..)) results get written to.

Author: cfbolz
Revision Date: 2007-12-14 10:36:11 UTC

allow weakrefability

Author: arigo
Revision Date: 2007-01-14 14:54:51 UTC

Remove the odd-numbers hack for the llinterp in codebuf.py; seems to work without.
Add an --interp flag for py.test to run test_interp_ts.

Author: arigo
Revision Date: 2010-03-05 11:43:12 UTC

Check in my working copy. That branch will probably be abandonned
and maybe restarted some day.

Author: arigo
Revision Date: 2007-06-15 14:01:15 UTC

Fix imports.

Author: arigo
Revision Date: 2010-02-23 16:50:19 UTC

Intermediate check-in. Mess. It's not going to work unless I find
a new idea.

Author: afa
Revision Date: 2010-02-20 16:29:35 UTC

a branch for trackgcroot.py to write the map table in a C array
(the MASM assembler on windows seems limited to 64K external symbols)

Author: arigo
Revision Date: 2009-11-05 10:46:09 UTC

This is the best result so far. At least in this case I understand the
results: they seem to clearly show that a lot of the objects allocated
during jit-tracing actually survive. The results are not good enough --
they say that jit-tracing takes only half the time, but the truth is
that it's always soon followed by a lengthy minor collection, so that
the total time (as well as the total GC time) end up not really
improved. Closing this branch for now.

Author: fijal
Revision Date: 2010-01-29 13:07:31 UTC

Merge changes from trunk

Author: arigo
Revision Date: 2009-11-08 16:34:33 UTC

svn merge -r68999:69075 svn+ssh://codespeak.net/svn/pypy/trunk

Author: arigo
Revision Date: 2009-10-21 16:33:11 UTC

Merge r68702 from trunk.

Author: hpk
Revision Date: 2003-10-15 13:44:09 UTC

made snapshot of funding 16_44

Author: santagada
Revision Date: 2009-11-25 22:57:18 UTC

reverted 69623

Author: niko
Revision Date: 2008-03-28 11:42:08 UTC

fix bug related to boolean arrays: we used to map them
to byte arrays, which caused us to incorrectly try
to box their arguments (since byte != boolean), etc

Author: arigo
Revision Date: 2009-11-21 17:13:47 UTC

Two tests about raising and catching exceptions in a loop.

Author: fijal
Revision Date: 2010-03-10 01:37:16 UTC

Make this newstyle classes as well (avoids bad mro)

Author: antocuni
Revision Date: 2008-06-11 12:39:28 UTC

try yet another way to encode bytecodes; this time, opargs are decoded
in advance, but they are stored togheter with the opcodes in a list of
ints. This code is not very well tested nor well written, I commit it
just to do benchmarks on different machines

Author: verte
Revision Date: 2009-10-24 02:18:03 UTC

Initial commit, mostly just failing tests.

Author: ac
Revision Date: 2007-11-21 15:05:03 UTC

(jacob, arre)
__future__ imports and 'with'-statements work again. Yay!

Author: antocuni
Revision Date: 2010-01-20 08:09:23 UTC

tentative checkin: implement WeakValueDictionary in rpython, without using the
implementation in rlib.rweakref. This way it works also with ootype

Author: arigo
Revision Date: 2007-09-11 19:56:11 UTC

Fix the logic used to skip the _weakref app-level module
from the list of enabled modules.

Author: benjamin
Revision Date: 2009-04-06 22:47:42 UTC

make the 2.5 grammar like 2.6, so test_compiler works

Author: hpk
Revision Date: 2003-09-17 20:58:52 UTC

the final merge of the builtinrefactor branch into the trunk. See

http://codespeak.net/pipermail/pypy-dev/2003q3/001012.html

for in-depth discussion and description of what is new.

Author: mwh
Revision Date: 2006-12-06 12:02:34 UTC

change some "if isinstances" to "assert isinstances" and delete some code.
now a pypy-c-fast is 2-4% faster than pypy-c-multi

Author: afa
Revision Date: 2008-06-10 10:48:46 UTC

merge revisions 55322:55728

Author: arigo
Revision Date: 2008-03-09 16:04:40 UTC

Aaah. socket.send() with a timeout would wait until
the socket was ready to be *read* from before sending
its data.....

Author: cfbolz
Revision Date: 2010-01-21 14:33:18 UTC

Experimental: after we made a new loop, immediately attach a bridge to it using
the same trace. In theory it should cut down on compilation time, in practise it
doesn't seem to matter.

Author: fijal
Revision Date: 2010-03-15 01:33:38 UTC

Try again branching to speedup blackhole

Author: arigo
Revision Date: 2008-01-10 20:43:27 UTC

Run in a predictible (alphabetical) order.

Author: arigo
Revision Date: 2009-09-30 12:57:14 UTC

Intermediate check-in.

Author: fijal
Revision Date: 2008-01-05 09:51:44 UTC

Merge translator/c from trunk as well.

Author: arigo
Revision Date: 2005-03-04 12:48:18 UTC

Mixins bug fix.

Author: cfbolz
Revision Date: 2010-02-08 10:24:11 UTC

a test for r71114

Author: electronicru
Revision Date: 2009-12-25 18:17:30 UTC

Adding branch for Python 2.6 support.

Author: arigo
Revision Date: 2005-05-21 11:04:06 UTC

issue75 testing

Moved a hack elsewhere, basically. Importing stringobject's apphelpers from
the cache didn't find _formatting at interp-level. This is all very broken,
but out of reach of a simple clean-up.

Author: antocuni
Revision Date: 2009-04-09 13:55:49 UTC

debug aid that have been living in my working copy for ages