test Log


Commit Date  
[d83ea5] (Release_2_5beta2) by zyasoft@79b11f02-f210-0410-99aa-fb189b3d1ac4

Use a PyList instead to store the block stack for the PBC-VM.

2009-03-07 16:00:32 Tree
[ecd023] by fwierzbicki@79b11f02-f210-0410-99aa-fb189b3d1ac4

Changing version numbers, README, plus some bugfix reports in NEWS (from some
archeology on http://bugs.jython.org)

2009-03-07 15:57:43 Tree
[b0a0bc] by amak@79b11f02-f210-0410-99aa-fb189b3d1ac4

Now that modjy is fully integrated, we no longer need the .zip version in extlibs.

2009-03-07 15:35:39 Tree
[ee2a3c] by fwierzbicki@79b11f02-f210-0410-99aa-fb189b3d1ac4

increment APIVersion, take --verify out of build.xml

2009-03-07 04:16:00 Tree
[6f316e] by zyasoft@79b11f02-f210-0410-99aa-fb189b3d1ac4

Now uses CompilerFlags instead of the co_flags bit vector uniformly in
the code. This change enables more programmatic direction of the
compiler, including at the very least the desired compiler (once we
include more than one!).

Removed unnecessary fields in PyFrame that supported the former
newcompiler. Made public some additional PyUnicode constructors to
support the Java version of unicodedata, whenever that happens. Fixed
a memory leak in PyBytecode.

2009-03-07 04:05:04 Tree
[d793c6] by otmarhumbel@79b11f02-f210-0410-99aa-fb189b3d1ac4

also unset _BOOT_CP (for symmetry)

2009-03-06 22:18:24 Tree
[8960d5] by otmarhumbel@79b11f02-f210-0410-99aa-fb189b3d1ac4

make byte code verification the default again:
- drop option --verify
- new option --boot

2009-03-06 21:56:19 Tree
[d06cae] by fwierzbicki@79b11f02-f210-0410-99aa-fb189b3d1ac4

Merged revisions 5751,6066-6070 via svnmerge from
https://jython.svn.sourceforge.net/svnroot/jython/branches/modjy

Merging Alan Kennedy's amazing modjy project http://modjy.xhaus.com/ into core of Jython
Cut the branch comments down to the interesting ones. Thanks Alan!
........
r6068 | amak | 2009-03-05 10:43:01 -0500 (Thu, 05 Mar 2009) | 1 line

Modification to the ModjyJServlet to reflect that it is now part of the jython distribution, rather than a standalone product.
........
r6069 | amak | 2009-03-05 10:54:43 -0500 (Thu, 05 Mar 2009) | 3 lines

Adding the original modjy demo application to the Demo subdirectory.
This application gives the users a complete running application that they can use as a template for aplication development.
In the future, we might develop a more sophisticated application.
........
r6070 | amak | 2009-03-05 17:07:35 -0500 (Thu, 05 Mar 2009) | 1 line

Committing the unit tests. There is a readme.txt file in the directory which explains how to run the tests, which depend on mockrunner.
........

2009-03-06 18:27:15 Tree
[04db96] (modjy) by fwierzbicki@79b11f02-f210-0410-99aa-fb189b3d1ac4

Merged revisions 5752,5756-5758,5761-5763,5765-5775,5778-5816,5818-5840,5842-5847,5854,5857-5859,5863-5865,5869,5875,5877,5884-5885,5889-5893,5900-5901,5903-5906,5908,5912-5918,5920,5923,5925,5927,5930-5932,5934-5947,5950-5951,5953-5957,5959,5961,5965-5966,5971-5976,5978-5979,5981-5983,5992,6001,6003,6006-6014,6023,6031-6037,6039-6040,6045-6047,6049-6052,6061-6065,6071 via svnmerge from
https://jython.svn.sourceforge.net/svnroot/jython/trunk/jython

2009-03-06 16:15:19 Tree
[9a312e] by fwierzbicki@79b11f02-f210-0410-99aa-fb189b3d1ac4

Added mtime to the compiled Jython classes to match CPython behavior.

Fixes:
http://bugs.jython.org/issue1567212 and http://bugs.jython.org/issue1024

2009-03-06 04:15:54 Tree
[42eb8a] by amak@79b11f02-f210-0410-99aa-fb189b3d1ac4

Committing the unit tests. There is a readme.txt file in the directory which explains how to run the tests, which depend on mockrunner.

2009-03-05 22:07:35 Tree
[0a8db1] by amak@79b11f02-f210-0410-99aa-fb189b3d1ac4

Adding the original modjy demo application to the Demo subdirectory.
This application gives the users a complete running application that they can use as a template for aplication development.
In the future, we might develop a more sophisticated application.

2009-03-05 15:54:43 Tree
[832e91] by amak@79b11f02-f210-0410-99aa-fb189b3d1ac4

Modification to the ModjyJServlet to reflect that it is now part of the jython distribution, rather than a standalone product.

2009-03-05 15:43:01 Tree
[d79144] by amak@79b11f02-f210-0410-99aa-fb189b3d1ac4

Copying over latest version of modjy, version 0.25.3.

2009-03-05 13:18:55 Tree
[2ac9c0] by amak@79b11f02-f210-0410-99aa-fb189b3d1ac4

Removing old version of modjy, to get a fresh start on copying in the latest version.

2009-03-05 13:15:50 Tree
[b40553] by otmarhumbel@79b11f02-f210-0410-99aa-fb189b3d1ac4

use the correct name of the new antlr 3.1.2 runtime .jar in /extlibs

2009-03-03 22:54:05 Tree
[dabf37] by fwierzbicki@79b11f02-f210-0410-99aa-fb189b3d1ac4

Test and fix for http://bugs.jython.org/issue1222877. Duplicate keyword args
where erroneously permitted in Jython. Thanks to Khalid Zuberi for taking a
crack at a patch, I modified the patch to avoid an extra loop.

2009-03-03 15:54:04 Tree
[6adedc] by otmarhumbel@79b11f02-f210-0410-99aa-fb189b3d1ac4

only antlr runtime jar is needed

2009-03-03 07:09:37 Tree
[8d035d] by otmarhumbel@79b11f02-f210-0410-99aa-fb189b3d1ac4

Update to antlr 3.1.2. for eclipse

2009-03-03 07:03:09 Tree
[6acb67] by fwierzbicki@79b11f02-f210-0410-99aa-fb189b3d1ac4

Merged in tests of mapping in exec and eval written by Mehendran in
Issue1815116. Thanks Mehendran!

2009-03-03 01:24:04 Tree
[42c4c4] by fwierzbicki@79b11f02-f210-0410-99aa-fb189b3d1ac4

Intital 2.7 tweaks: parser updates, with statement always on, add abc.py.

2009-03-02 18:12:22 Tree
[2e5cbf] by thobes@79b11f02-f210-0410-99aa-fb189b3d1ac4

Moving the advanced branch. All useful things imported.
At this point there are two patches to jython for integrating the advanced compiler:
* One patch to make the flags that are passed around in compilation type safe.
* One patch to decouple the compiler implementation and provide an extension
point where a new compiler can be inserted.

2009-03-02 15:59:20 Tree
[6085a4] by fwierzbicki@79b11f02-f210-0410-99aa-fb189b3d1ac4

Merged revisions 5934-5947,5950-5951,5953-5957,5959,5961,5965-5966,5971-5976,5978-5979,5981-5983,5992,6001,6003,6006-6014,6023,6031-6037,6039-6040,6045-6047,6049-6052 via svnmerge from
https://jython.svn.sourceforge.net/svnroot/jython/trunk/jython

................
r5934 | fwierzbicki | 2009-01-14 22:36:09 -0500 (Wed, 14 Jan 2009) | 3 lines

Change negate so that -0 and -0.0 behave as CPython behaves (no USub node, just
negate zero).
................
r5935 | cgroves | 2009-01-16 03:11:48 -0500 (Fri, 16 Jan 2009) | 6 lines

underlying_class is the class of instances of this type, so PyJavaType shouldn't store the class
it's wrapping in it. Instead, it should use the general javaProxy field on PyObject and leave
underlying_class up to its bases. Without this, Python classes can't subclass a Java interface and
a Python class subclassing a Java class or interface.
................
r5936 | cgroves | 2009-01-16 04:17:19 -0500 (Fri, 16 Jan 2009) | 4 lines

Check for parent beans of the same name to fill in missing getters or setters when adding beans for
a type. Fixes bug #1132. Thanks to garyh for tracking it down and supplying a test case.
................
r5937 | zyasoft | 2009-01-17 13:12:57 -0500 (Sat, 17 Jan 2009) | 6 lines

Fixes #1227 where for cStringIO.StringIO, orderings of method calls
like write-seek-write did not properly work. This also addresses a
problem for PyAMF support.

Thanks go to Nick Joyce for the patch!
................
r5938 | zyasoft | 2009-01-17 13:31:06 -0500 (Sat, 17 Jan 2009) | 6 lines

Fixes #1217 where we report the underlying Java error
(java.util.zip.DataFormatException) instead of the standard Python one
(zlib.error). This also helps support PyAMF on Jython.

Thanks Nick Joyce for the patch.
................
r5939 | cgroves | 2009-01-17 13:42:14 -0500 (Sat, 17 Jan 2009) | 1 line

Shoulda gone out with r5607.
................
r5940 | cgroves | 2009-01-17 15:35:09 -0500 (Sat, 17 Jan 2009) | 3 lines

Add some tests for overriding attribute lookup and adding mapping methods to wrapped Java types.
................
r5941 | cgroves | 2009-01-17 15:56:28 -0500 (Sat, 17 Jan 2009) | 5 lines

Implement Iterable in PyIterator. This means generators are usable in foreach loops. The
implementation could be moved up to PyObject, but I feel like implementing the interface says that
the object will be iterable, and not all PyObjects provide an __iter__.
................
r5942 | cgroves | 2009-01-17 17:36:27 -0500 (Sat, 17 Jan 2009) | 3 lines

Resurrect some PyJavaClass code to fix bug #1235.
................
r5943 | cgroves | 2009-01-17 18:57:54 -0500 (Sat, 17 Jan 2009) | 1 line

Match the statuses from CPython
................
r5944 | cgroves | 2009-01-17 19:53:10 -0500 (Sat, 17 Jan 2009) | 1 line

ecj likes this, but javac disputes that it's compilable, so go back to the old way
................
r5945 | cgroves | 2009-01-18 05:04:29 -0500 (Sun, 18 Jan 2009) | 7 lines

Add an Ant task to compile .py files to $py.class. It runs on Lib as part of developer-build to
exercise it a little bit.

Also, only rename the xerces parts of org.apache. Otherwise our ant imports get renamed when
jarjaring.
................
r5946 | otmarhumbel | 2009-01-18 08:46:12 -0500 (Sun, 18 Jan 2009) | 9 lines

enable running jython from a directory with an exclamation mark in it's name

due to jdk limitations this is only possible for
+ jdk 1.6 and higher
+ directory names not ending with an exclamation mark

due to the special treatment of exclamation marks in enabledelayedexpansion, this is currently not possible for jython.bat on windows

installer autotests now use an exclamation mark if possible
................
r5947 | otmarhumbel | 2009-01-19 16:43:36 -0500 (Mon, 19 Jan 2009) | 2 lines

add xercesImpl as zipfileset to jarjar again:
without this, there are no org.python.apache.* packages in jython.jar
................
r5950 | zyasoft | 2009-01-20 13:28:03 -0500 (Tue, 20 Jan 2009) | 4 lines

Added os.getcwdu. This fixes #1219 (necessary for Django 1.0.2
support), but not the larger question of better Unicode handling in
the os module, specifically around os.path.
................
r5951 | leosoto | 2009-01-20 14:29:17 -0500 (Tue, 20 Jan 2009) | 1 line

Use the new getcwdu on POSIXHandler, to avoid unnecesary conversions from unicode to bytestring. Also added getcwdu to os.__all__
................
r5953 | cgroves | 2009-01-21 15:47:19 -0500 (Wed, 21 Jan 2009) | 1 line

Break out the compile method for use in subclasses
................
r5954 | cgroves | 2009-01-21 16:26:50 -0500 (Wed, 21 Jan 2009) | 1 line

Turn includeJavaStackInExceptions on by default as it was always done in previous versions.
................
r5955 | cgroves | 2009-01-21 18:32:55 -0500 (Wed, 21 Jan 2009) | 8 lines

Add a jython initializer service.

If META-INF/services/org.python.core.JythonInitializer is on the classpath, the class named in that
file will be instantiated and used in Jython's initialization. This is useful when Jython is
initialized by a library outside of your control, but some customization still needs to be done to
Jython's environment. I promise not to add a JythonFactory or a JythonFactoryFactory.
................
r5956 | fwierzbicki | 2009-01-21 21:21:08 -0500 (Wed, 21 Jan 2009) | 4 lines

Fix for http://bugs.jython.org/issue1239. When jars where added to sys.path,
they did not always trigger package caching, leading to unexpected import
failures. This is Charlie Grove's fix, which was much better than mine.
................
r5957 | fwierzbicki | 2009-01-21 23:16:47 -0500 (Wed, 21 Jan 2009) | 7 lines

Thanks to Marc Downie for this patch that fixes

http://bugs.jython.org/issue1230 importing with '*' from java packages does not
work with 2.5b1.

Also added a test to test_java_integration.py.
................
r5959 | cgroves | 2009-01-22 14:50:18 -0500 (Thu, 22 Jan 2009) | 1 line

Expose rich comparision methods on Java classes implementing Comparable
................
r5961 | fwierzbicki | 2009-01-22 20:03:39 -0500 (Thu, 22 Jan 2009) | 3 lines

Over-ride pyset in PyTuple so our tuples go back to being immutable. Fixes
http://bugs.jython.org/issue1242.
................
r5965 | pjenvey | 2009-01-23 18:40:46 -0500 (Fri, 23 Jan 2009) | 2 lines

more java.lang.Strings to regular str instead of unicode
................
r5966 | pjenvey | 2009-01-23 19:17:40 -0500 (Fri, 23 Jan 2009) | 3 lines

merge from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/platform.py -c 66214
................
r5971 | cgroves | 2009-01-24 15:10:59 -0500 (Sat, 24 Jan 2009) | 1 line

Run the java import * test from a sub-interpreter to allow regrtest to be run with python.cachedir.skip=true
................
r5972 | cgroves | 2009-01-24 15:11:54 -0500 (Sat, 24 Jan 2009) | 1 line

Try the current classloader for places where the context classloader isn't set
................
r5973 | cgroves | 2009-01-24 15:17:39 -0500 (Sat, 24 Jan 2009) | 1 line

Exclude the xerces service declaration as our jarjar'd version doesn't match the names in the declaration
................
r5974 | cgroves | 2009-01-25 07:15:25 -0500 (Sun, 25 Jan 2009) | 1 line

Bring up to the current state of affairs
................
r5975 | cgroves | 2009-01-25 07:19:27 -0500 (Sun, 25 Jan 2009) | 3 lines

Use Python's mro to determine an mro for Java classes and interfaces. If the Java class has an mro conflict, resolve it by arbitrarily choosing an interface or class. If an interface or class involved in an mro conflict has had its dict modified, or has its dict modified, raise a TypeError.
................
r5976 | cgroves | 2009-01-25 07:19:58 -0500 (Sun, 25 Jan 2009) | 1 line

Test support classes for the last commit
................
r5978 | nriley | 2009-01-25 16:06:04 -0500 (Sun, 25 Jan 2009) | 1 line

Typo fix.
................
r5979 | cgroves | 2009-01-25 19:33:21 -0500 (Sun, 25 Jan 2009) | 4 lines

Include Python Java subclasses as bases as well as the Java proxy type, as it may have additional methods in its Python.

Return the proxyClass directly from getProxyType as we're setting that on Python subclasses now.
................
r5981 | cgroves | 2009-01-25 21:21:05 -0500 (Sun, 25 Jan 2009) | 1 line

Defer resolving inner classes till the containing class is fully resolved. Fixes issue #1234.
................
r5982 | cgroves | 2009-01-25 22:10:36 -0500 (Sun, 25 Jan 2009) | 1 line

Try to assign to a static field for a given name in __setattr__ on a Java type before filling something in on the dict. Fixes bug #1241
................
r5983 | cgroves | 2009-01-25 22:36:25 -0500 (Sun, 25 Jan 2009) | 1 line

Fix javatests
................
r5992 | amak | 2009-01-28 11:05:59 -0500 (Wed, 28 Jan 2009) | 3 lines

Fix and unit test for bug 1244.
Problem letting system choose the port for binding UDP socket
http://bugs.jython.org/issue1244
................
r6001 | amak | 2009-01-29 12:59:04 -0500 (Thu, 29 Jan 2009) | 2 lines

Re-arranging the socket shutdown methods. TCP client sockets can shutdown their input streams, but this is not appropriate for either TCP server sockets or UDP sockets, which don't have input and output streams. For these latter two types, the shutdown method should have the same effect as the close method, and thus the shutdown method is a no-op for these types. I have documented this difference between cpython and jython on the wiki. This should finally resolve bug 1121: listening socket shutdown expects the wrong kind of socket
http://bugs.jython.org/issue1121
................
r6003 | pjenvey | 2009-01-30 21:20:48 -0500 (Fri, 30 Jan 2009) | 1 line

make --version resemble CPython
................
r6006 | amak | 2009-02-02 10:50:25 -0500 (Mon, 02 Feb 2009) | 4 lines

Fix for bug 1258.
select() semantics differ from CPython, causing pydoc HTTPd to fail
http://bugs.jython.org/issue1258
Thanks to Ryan Blair for reporting the bug and providing a patch.
................
r6007 | amak | 2009-02-03 07:01:46 -0500 (Tue, 03 Feb 2009) | 2 lines

Committing the latest version of modjy.
See the release notes for details.
................
r6008 | pjenvey | 2009-02-03 15:19:57 -0500 (Tue, 03 Feb 2009) | 1 line

fix test_version per r6003
................
r6009 | pjenvey | 2009-02-03 16:25:23 -0500 (Tue, 03 Feb 2009) | 2 lines

include PATH for platforms lacking bash in /bin
................
r6010 | leosoto | 2009-02-03 19:51:47 -0500 (Tue, 03 Feb 2009) | 1 line

PyObject#reduce_2: Use invoke("iteritems") instead of this.iteritems() to avoid problems with dict subclasses which override iteritems on python code. Fixes #1257
................
r6011 | amak | 2009-02-04 10:20:56 -0500 (Wed, 04 Feb 2009) | 4 lines

Our asyncore is completely broken; see bug 1237.
Asnycore does not seem to work in Jython 2.5 (b0 or b1)
http://bugs.jython.org/issue1237
Starting again from scratch; Step 1: remove existing module.
................
r6012 | amak | 2009-02-04 10:22:21 -0500 (Wed, 04 Feb 2009) | 1 line

Fresh start with asyncore; step 2: copy over asyncore from cpython 2.5.2.
................
r6013 | amak | 2009-02-04 10:27:59 -0500 (Wed, 04 Feb 2009) | 3 lines

Fresh start with asycnore; step 3: make necessary changes to cpython 2.5 module to work on jython 2.5.
The code in this bug report now works identically to cpython.
http://bugs.jython.org/issue1237
................
r6014 | otmarhumbel | 2009-02-05 13:08:10 -0500 (Thu, 05 Feb 2009) | 1 line

added dummy_thread.py and dummy_threading.py
................
r6023 | fwierzbicki | 2009-02-08 20:40:53 -0500 (Sun, 08 Feb 2009) | 2 lines

Fix for http://bugs.jython.org/issue1243. Thanks RJ Ryan for suggesting the fix.
................
r6031 | fwierzbicki | 2009-02-15 09:23:28 -0500 (Sun, 15 Feb 2009) | 3 lines

Fix potential infinite loop (and wrong slice behavior) -- found with FindBugs
................
r6032 | fwierzbicki | 2009-02-15 13:22:18 -0500 (Sun, 15 Feb 2009) | 2 lines

Fixed two infinite recursions found with FindBugs.
................
r6033 | fwierzbicki | 2009-02-15 13:31:27 -0500 (Sun, 15 Feb 2009) | 2 lines

Fixed NPE found with FindBugs.
................
r6034 | amak | 2009-02-15 13:35:41 -0500 (Sun, 15 Feb 2009) | 2 lines

A new version of modjy, containing some bug fixes kindly contributed by Philip Jenvey; see contained release notes for more information.
http://github.com/pjenvey/modjy-pjenvey
................
r6035 | pjenvey | 2009-02-15 18:27:44 -0500 (Sun, 15 Feb 2009) | 3 lines

double antlr's conversiontimeout for slower environments, in particular the
hudson buildot
................
r6036 | fwierzbicki | 2009-02-16 10:41:50 -0500 (Mon, 16 Feb 2009) | 2 lines

Useless double assignment.
................
r6037 | fwierzbicki | 2009-02-16 21:38:09 -0500 (Mon, 16 Feb 2009) | 2 lines

Make the public adapters in AstAdapters final.
................
r6039 | amak | 2009-02-18 08:21:13 -0500 (Wed, 18 Feb 2009) | 1 line

Updating the zip file for modjy 0.25.3; previously checked in version had the wrong version of modjy.jar in WEB-INF/lib.
................
r6040 | fwierzbicki | 2009-02-19 15:16:18 -0500 (Thu, 19 Feb 2009) | 4 lines

Our handling of writes after a seek beyond the end of cStringIO was not
consistent with CPython (or even our own use of StringIO). Found while trying
to get Mercurial to respond to "hg log".
................
r6045 | zyasoft | 2009-02-24 00:04:05 -0500 (Tue, 24 Feb 2009) | 3 lines

Initialized merge tracking via "svnmerge" with revisions "1-5948" from
https://jython.svn.sourceforge.net/svnroot/jython/branches/pbcvm
................
r6046 | zyasoft | 2009-02-24 00:45:38 -0500 (Tue, 24 Feb 2009) | 206 lines

Merged revisions 5949,5952,5962,5964,5967-5970,5980,5984,5987-5989,5998-5999,6004-6005,6015-6022,6024-6030,6038,6041-6042 via svnmerge from
https://jython.svn.sourceforge.net/svnroot/jython/branches/pbcvm

........
r5949 | zyasoft | 2009-01-20 06:56:46 -0700 (Tue, 20 Jan 2009) | 2 lines

Initial commit.
........
r5952 | zyasoft | 2009-01-20 22:02:29 -0700 (Tue, 20 Jan 2009) | 3 lines

Fixed binary and inplace bytecodes to use the proper methods.
Added block support to support iteration.
........
r5962 | zyasoft | 2009-01-22 21:40:26 -0700 (Thu, 22 Jan 2009) | 3 lines

Added support for try/except and try/finally blocks (and their
combination).
........
r5964 | zyasoft | 2009-01-23 08:30:09 -0700 (Fri, 23 Jan 2009) | 2 lines

Basic generators are almost supported.
........
r5967 | zyasoft | 2009-01-23 19:49:52 -0700 (Fri, 23 Jan 2009) | 2 lines

Now supports generators, via the standard PyGenerator protocol.
........
r5968 | zyasoft | 2009-01-23 21:36:06 -0700 (Fri, 23 Jan 2009) | 2 lines

Added support for keyword args, *arg, and **kwarg calling conventions.
........
r5969 | zyasoft | 2009-01-24 09:47:59 -0700 (Sat, 24 Jan 2009) | 3 lines

Added support for basic coroutines (no support for throw) and
generator expressions.
........
r5970 | zyasoft | 2009-01-24 11:02:37 -0700 (Sat, 24 Jan 2009) | 2 lines

Supports throwing exceptions into generators.
........
r5980 | zyasoft | 2009-01-25 18:19:12 -0700 (Sun, 25 Jan 2009) | 4 lines

Properly handles *args, **kwargs for the receiver function.
func_code can now be set to a PyBytecode (or other extender of PyBaseCode).
Turned off debugging by default for PyBytecode objects.
........
r5984 | zyasoft | 2009-01-25 22:02:21 -0700 (Sun, 25 Jan 2009) | 2 lines

Added extract tool (to be run by CPython) for creating test modules.
........
r5987 | zyasoft | 2009-01-26 18:45:48 -0700 (Mon, 26 Jan 2009) | 2 lines

Fixed raising of exceptions and PyStack.rot
........
r5988 | zyasoft | 2009-01-26 19:16:49 -0700 (Mon, 26 Jan 2009) | 2 lines

Fixed opcode support for LOAD/STORE/DELETE_NAME and BUILD_CLASS.
........
r5989 | zyasoft | 2009-01-26 19:26:20 -0700 (Mon, 26 Jan 2009) | 2 lines

Fixed RAISE_VARARGS opcode to raise exceptions properly.
........
r5998 | zyasoft | 2009-01-28 20:44:58 -0700 (Wed, 28 Jan 2009) | 3 lines

extract.py now creates modules where codeobjects respect their
dependency relationships (via a tsort).
........
r5999 | zyasoft | 2009-01-29 07:24:33 -0700 (Thu, 29 Jan 2009) | 6 lines

Extract script now processes most/all of the test.test_* modules with
a test_main and importable by CPython. May want to consider
generalizing support for instance methods and closure setup with
respect to new.function (= PyFunction), but this seems purely a
testing nicety for top-level defs.
........
r6004 | zyasoft | 2009-02-01 01:12:13 -0700 (Sun, 01 Feb 2009) | 3 lines

From
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_marshal@27825
........
r6005 | zyasoft | 2009-02-01 01:24:06 -0700 (Sun, 01 Feb 2009) | 13 lines

Added _marshal, a Java implementation of the Python module, and made
current. test_marshal now passes (with minor modifications around GC
and floating point accuracy on to/from string representation), except
for support of code objects. That's next.

Pulled IOFile (now PyIOFile) from cPickle to share its duck typing
with _marshal

A historical note: from what I can discern in the log, the original
marshal.py was written by Guido van Rossum specifically for "JPython"
in 1997, one of the few bits he apparently did. Now it survives only
as a shell, to import in _marshal.
........
r6015 | zyasoft | 2009-02-05 20:50:10 -0700 (Thu, 05 Feb 2009) | 5 lines

Changed read/write binary_float so that it doesn't go through struct
but instead j.l.Double's support for converting to/from bits. This
also resolves endian ordering issues so that it follows the native
endian ordering.
........
r6016 | zyasoft | 2009-02-06 00:07:45 -0700 (Fri, 06 Feb 2009) | 7 lines

Fixed some cellvar/freevar nested scope issues, with work to continue.

Added a slightly modified version of pycimport.py from Tobias' pyasm
implementation in the sandbox. This, in conjunction with compileall
(run from CPython; compileall.compile_dir against the unit tests) is
allowing a significant fraction of the regrtest to run against PBC-VM.
........
r6017 | zyasoft | 2009-02-06 00:42:15 -0700 (Fri, 06 Feb 2009) | 2 lines

Fixed slice, exec support.
........
r6018 | zyasoft | 2009-02-06 21:04:54 -0700 (Fri, 06 Feb 2009) | 2 lines

Fixed stack ops (DUP_TOPX, ROT_THREE, ROT_FOUR), INPLACE_POWER.
........
r6019 | zyasoft | 2009-02-06 23:29:37 -0700 (Fri, 06 Feb 2009) | 5 lines

Fixed IMPORT_STAR for nested modules. sys.exc_info now returns Py.None
for the traceback (this fixes an issue in PBC where null is being used
as both a sentinel value by __getitem__ AND an actual value for
exception handling).
........
r6020 | zyasoft | 2009-02-07 10:47:19 -0700 (Sat, 07 Feb 2009) | 2 lines

Fixes nested scope treatment of free variables.
........
r6021 | zyasoft | 2009-02-07 11:14:33 -0700 (Sat, 07 Feb 2009) | 2 lines

Further refinement of nested scopes.
........
r6022 | zyasoft | 2009-02-07 14:26:31 -0700 (Sat, 07 Feb 2009) | 3 lines

Fixed control flow by pushing the wrapped Why onto the stack (instead
of its string representation from the earliest version of this code).
........
r6024 | zyasoft | 2009-02-10 20:11:10 -0700 (Tue, 10 Feb 2009) | 7 lines

Support tostring/fromstring of unsigned types so that they can be
stored in the same size as their signed representation.

This fixes an issues with peak.util.assembler where it expects that
unsigned bytes (typecode 'B') take one byte to store and therefore are
directly interoperable with co_code.
........
r6025 | zyasoft | 2009-02-11 21:38:04 -0700 (Wed, 11 Feb 2009) | 2 lines

Beginning of PBC compilation support. Pretty much a shell for now.
........
r6026 | zyasoft | 2009-02-14 02:23:58 -0700 (Sat, 14 Feb 2009) | 7 lines

Now interns names in co_names, etc., when unmarshalling, which fixes
equality tests on special case like __class__, __dict__, etc.

Fix instanceof test in eval and exec so that it looks at PyBaseCode,
not PyTableCode. Also, exec cannot be called with Py.None instead of
null. Fixes test_scope.
........
r6027 | zyasoft | 2009-02-14 03:44:17 -0700 (Sat, 14 Feb 2009) | 3 lines

Set the traceback when intercepting an exception.
Set __doc__ strings for functions from the constant pool (if available).
........
r6028 | zyasoft | 2009-02-14 04:17:17 -0700 (Sat, 14 Feb 2009) | 3 lines

LOAD_DEREF needs to do the same lookup as LOAD_CLOSURE.
Populate doc strings for MAKE_CLOSURE.
........
r6029 | zyasoft | 2009-02-14 04:58:48 -0700 (Sat, 14 Feb 2009) | 3 lines

Fixed WITH_CLEANUP so that it properly calls __exit__ with the
exception components off the stack.
........
r6030 | zyasoft | 2009-02-14 12:17:04 -0700 (Sat, 14 Feb 2009) | 3 lines

Fixed UNPACK_SEQUENCE so that ValueError is raised if too few/too many
values to unpack.
........
r6038 | zyasoft | 2009-02-17 17:06:55 -0700 (Tue, 17 Feb 2009) | 8 lines

Added support for code.getlineno (using co_lnotab), co_code,
co_consts, co_lnotab. co_code and co_lnotab now use byte[] for a more
compact representation and then manage accessing as if it were an
unsigned representation.

This enables a good chunk of the peak.util.assembler unit tests to
pass; the remainder seem to have a dictionary traversal dependency.
........
r6041 | zyasoft | 2009-02-22 07:40:49 -0700 (Sun, 22 Feb 2009) | 4 lines

Inlined PyTableCode#interpret into #call(PyFrame,PyClosure) (which
restores this method to what it was before the PBC-VM work). This
change restores the performance as tested by test/pystone.py.
........
r6042 | zyasoft | 2009-02-23 21:25:13 -0700 (Mon, 23 Feb 2009) | 8 lines

LineCache implementation to support efficient setline edge detection,
as necessary for tracing.

No longer looks up the ThreadState twice, in the caller of
PyBytecode#interpret and interpret itself, so this should provide for
some optimization.
........
................
r6047 | fwierzbicki | 2009-02-24 15:42:25 -0500 (Tue, 24 Feb 2009) | 6 lines

Added a target and infrastructure for compiling the ctypes_test.c needed for
test__rawffi.py. So far only tested on OS X.

Used cpptasks from http://ant-contrib.sourceforge.net/cpptasks/index.html
................
r6049 | fwierzbicki | 2009-03-01 15:42:46 -0500 (Sun, 01 Mar 2009) | 3 lines

Update to antlr 3.1.2. This required deleting mismatch() override in Python.g
and PythonPartial.g.
................
r6050 | fwierzbicki | 2009-03-01 19:59:05 -0500 (Sun, 01 Mar 2009) | 2 lines

Bump CPython version to 70085.
................
r6051 | fwierzbicki | 2009-03-01 20:00:34 -0500 (Sun, 01 Mar 2009) | 3 lines

from
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_new@52060
................
r6052 | fwierzbicki | 2009-03-01 20:04:20 -0500 (Sun, 01 Mar 2009) | 2 lines

maybe 'never' was too strong for new.code in Jython. Disable test for now.
................

2009-03-02 03:29:30 Tree
[32469a] by fwierzbicki@79b11f02-f210-0410-99aa-fb189b3d1ac4

maybe 'never' was too strong for new.code in Jython. Disable test for now.

2009-03-02 01:04:20 Tree
[76c3ac] by fwierzbicki@79b11f02-f210-0410-99aa-fb189b3d1ac4

from
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_new@52060

2009-03-02 01:00:34 Tree
Older >