Commit [89818b] default  Maximize  Restore  History

Merged revisions 4386,4388-4390,4392,4403,4405,4408,4418,4424-4425,4434,4466,4474,4490,4493,4496,4499-4501,4504,4523-4524,4548-4549,4577-4579,4581-4582,4585-4589,4599-4600,4603-4605,4607-4608,4613,4628,4640,4642-4643,4648,4652-4662,4665,4667,4672,4676,4678-4680,4682,4684,4693-4695,4704-4706,4708-4712,4714-4715,4720,4729,4732-4735,4740-4751,4753-4758,4761-4765,4767-4772,4774,4777-4779,4781-4802,4806-4810,4821-4843,4845-4850,4852-4853,4855-4857,4859,4863-4870,4876-4885,4887-4897,4901-4914,4916-4919,4922,4924-4925,4928,4930,4932-4937,4939-4945,4947,4949-4952,4959,4961,4963,4965,4967-4971,4973-4974,4980,4982-5100,5102-5117,5119-5166,5168-5197 via svnmerge from

https://jython.svn.sourceforge.net/svnroot/jython/branches/asm

........
r4386 | fwierzbicki | 2008-05-08 14:55:44 -0400 (Thu, 08 May 2008) | 2 lines

Recreate asmified old compiler.
........
r4388 | fwierzbicki | 2008-05-10 13:19:27 -0400 (Sat, 10 May 2008) | 3 lines

Hand merged changes in trunk to copiler code. Removed Attribute support --
needs to be replaced by either ASM Attributes or Annotations.
........
r4389 | fwierzbicki | 2008-05-12 13:48:35 -0400 (Mon, 12 May 2008) | 4 lines

Reworked Code.java back towards old API (so methods like code.dup() are back
so it is not necessary to use code.visitInsn(DUP). This makes it easier to
read the code and to compare with the old code base.
........
r4390 | fwierzbicki | 2008-05-12 22:00:02 -0400 (Mon, 12 May 2008) | 2 lines

Converted everything over to antlr from javacc.
........
r4392 | fwierzbicki | 2008-05-13 15:38:04 -0400 (Tue, 13 May 2008) | 3 lines

Moved more code to older style Code.java methods (still based on ASM).
Added Visitor and ParseException.
........
r4403 | fwierzbicki | 2008-05-15 12:39:30 -0400 (Thu, 15 May 2008) | 2 lines

Enough compiler and parser tweaking to keep site.py from exploding.
........
r4405 | fwierzbicki | 2008-05-15 15:18:26 -0400 (Thu, 15 May 2008) | 2 lines

Fix for "from mod import *" and start of setline support.
........
r4408 | fwierzbicki | 2008-05-15 19:29:22 -0400 (Thu, 15 May 2008) | 2 lines

Temporary fix for line number problems causing NPEs.
........
r4418 | fwierzbicki | 2008-05-17 12:20:34 -0400 (Sat, 17 May 2008) | 2 lines

Remove old parser.
........
r4424 | fwierzbicki | 2008-05-20 13:39:32 -0400 (Tue, 20 May 2008) | 4 lines

Added a shell script to run Jython. This was adapted from a script from the JRuby project.
I also included the same suport for The Java Interactive Profiler (JIP) that JRuby includes.
See http://jiprof.sourceforge.net/
........
r4425 | fwierzbicki | 2008-05-20 13:50:01 -0400 (Tue, 20 May 2008) | 2 lines

Alter profile.properties for Jython.
........
r4434 | pjenvey | 2008-05-20 17:11:36 -0400 (Tue, 20 May 2008) | 1 line

missed one part of the PyMethod constructor signature change in r4400, from trunk
........
r4466 | fwierzbicki | 2008-05-27 22:05:18 -0400 (Tue, 27 May 2008) | 12 lines

Now that things are close to working, time to completely
destroy things again. Switched to pulling in 2.5 Python
libraries. Took first baby step by attempting to support

def f():
yield

which is not valid 2.3 syntax. (This is actually what
motivated this -- I was about to "fix" the grammar to
disallow this when I accidentally noticed that it was
valid 2.5 syntax.
........
r4474 | fwierzbicki | 2008-05-29 15:34:22 -0400 (Thu, 29 May 2008) | 2 lines

First stab at decorator support.
........
r4490 | fwierzbicki | 2008-05-30 11:19:42 -0400 (Fri, 30 May 2008) | 15 lines

Begin experimenting with a partial parser -- that is a recognizer
that will be able to tell if a string of code from an interactive
session is a true syntax error, or if it is part of a potentially
valid string of code. For example, in the interactive session:

>>> s = """
... hello

Is the start of a valid string, and should be recognized as a
valid partial parse. On the other hand

>>> print ,

Is still a syntax error.
........
r4493 | nriley | 2008-06-01 15:50:00 -0400 (Sun, 01 Jun 2008) | 1 line

add ImportWarning; update exceptions class doc to 2.5
........
r4496 | fwierzbicki | 2008-06-01 20:11:03 -0400 (Sun, 01 Jun 2008) | 2 lines

Need to see exception print outs.
........
r4499 | fwierzbicki | 2008-06-01 21:03:08 -0400 (Sun, 01 Jun 2008) | 2 lines

Switched to 2.4 CPython lib.
........
r4500 | nriley | 2008-06-01 21:59:37 -0400 (Sun, 01 Jun 2008) | 1 line

startswith/endswith on str/unicode, passes test_str from 2.5
........
r4501 | nriley | 2008-06-01 22:34:18 -0400 (Sun, 01 Jun 2008) | 1 line

TypeError rather than NotImplemented for wrong type on str.__mul__, fixes test_mul
........
r4504 | nriley | 2008-06-01 23:55:14 -0400 (Sun, 01 Jun 2008) | 1 line

r4500 broke http://pylonshq.com/pasties/827, reverting
........
r4523 | fwierzbicki | 2008-06-03 02:39:41 -0400 (Tue, 03 Jun 2008) | 6 lines

Most of this change is generated -- the real code changes
are in Python.g, PythonWalker.g, and asdl_antlr.py.

Moving to antlr 3.1 style of AST production. Changed the
Num processing only so far.
........
r4524 | fwierzbicki | 2008-06-03 10:47:46 -0400 (Tue, 03 Jun 2008) | 6 lines

Moved Str processing from PythonWalker.g to Python.g. Made Unicode and Str
hold PyUnicode and PyString objects instead of java.lang.String objects.
This will help compatibility with CPython AST -- though *may* hurt performance
more investigation will be needed. Certainly the way CodeCompiler is working
will need adjustment if this strategy is the way to go.
........
r4548 | nriley | 2008-06-06 16:48:42 -0400 (Fri, 06 Jun 2008) | 1 line

sync sre opcodes with Python 2.4; implement SRE_OP_GROUPREF_EXISTS
........
r4549 | nriley | 2008-06-06 17:12:40 -0400 (Fri, 06 Jun 2008) | 1 line

fix comment typo
........
r4577 | nriley | 2008-06-08 22:14:13 -0400 (Sun, 08 Jun 2008) | 1 line

Jython-specific test_cmd_line
........
r4578 | fwierzbicki | 2008-06-08 22:34:43 -0400 (Sun, 08 Jun 2008) | 2 lines

Generator Expressions.
........
r4579 | nriley | 2008-06-08 22:39:10 -0400 (Sun, 08 Jun 2008) | 1 line

fix regrtest ant target dependencies so it works from clean
........
r4581 | fwierzbicki | 2008-06-08 22:58:09 -0400 (Sun, 08 Jun 2008) | 3 lines

Better name for comprehension local var (comp instead of lc -- originally
stolen from the list comprehension code).
........
r4582 | nriley | 2008-06-08 23:09:46 -0400 (Sun, 08 Jun 2008) | 1 line

cookielib works now
........
r4585 | fwierzbicki | 2008-06-09 09:06:46 -0400 (Mon, 09 Jun 2008) | 2 lines

Conditional Expression
........
r4586 | fwierzbicki | 2008-06-09 10:10:41 -0400 (Mon, 09 Jun 2008) | 3 lines

switch to _(1_1) style names for the temporaries of gen expressions, so that vars
with unlikely but possible names like __gen1_1 do not get clobbered.
........
r4587 | nriley | 2008-06-09 17:19:49 -0400 (Mon, 09 Jun 2008) | 1 line

provenance of test_cmd_line
........
r4588 | fwierzbicki | 2008-06-09 22:29:18 -0400 (Mon, 09 Jun 2008) | 2 lines

Remove unnecessary vars.
........
r4589 | fwierzbicki | 2008-06-10 07:40:31 -0400 (Tue, 10 Jun 2008) | 2 lines

Remove comment that I no longer believe :)
........
r4599 | fwierzbicki | 2008-06-11 17:26:33 -0400 (Wed, 11 Jun 2008) | 6 lines

Make ParseException the only exception that comes out of the antlr
parsing. Unfortunately I haven't figured out how to remove
"throws RecognitionException" from the generated code, so some
references remain. Also renamed PythonGrammer -> ModuleParser, and
some other related refactorings.
........
r4600 | fwierzbicki | 2008-06-11 17:40:36 -0400 (Wed, 11 Jun 2008) | 2 lines

More descriptive name antlr.java -> ParserFacade.java
........
r4603 | fwierzbicki | 2008-06-11 21:34:06 -0400 (Wed, 11 Jun 2008) | 4 lines

Added a null guard for lexing errors (they have no token) and added an
experimental getOffset() method on ParseException which intends to get
the char position from the start of the file where the error occured.
........
r4604 | fwierzbicki | 2008-06-11 21:38:05 -0400 (Wed, 11 Jun 2008) | 2 lines

Crap, left in debugging. Removed.
........
r4605 | fwierzbicki | 2008-06-11 22:14:26 -0400 (Wed, 11 Jun 2008) | 2 lines

Fix offset for lexer errors.
........
r4607 | fwierzbicki | 2008-06-12 11:27:35 -0400 (Thu, 12 Jun 2008) | 2 lines

oops, wrong directory position.
........
r4608 | fwierzbicki | 2008-06-12 17:08:38 -0400 (Thu, 12 Jun 2008) | 5 lines

pep 263. There is some roughness remaining, especially a strange
problem around utf-8 that I need to look into (I ended up letting
Jython go with its default in the case of a utf-8 encoding
specified for now).
........
r4613 | fwierzbicki | 2008-06-13 13:33:31 -0400 (Fri, 13 Jun 2008) | 4 lines

Add a token source for partial parsing (for now -- can probably become
a subclass of PythonTokenSource once the kinks are out). Unterminated
triple strings now pass.
........
r4628 | zyasoft | 2008-06-14 19:23:52 -0400 (Sat, 14 Jun 2008) | 1 line

Added visitWith to support the with-statement. Changed ClassFile so that it does frame analysis with COMPUTE_FRAMES (used by Java 6, seems OK with Java 5 - and prevents code from being compiled that would fail bytecode verifier. Generally that would be a good thing. Fixed threading.Lock.__enter__ so it returns the lock object, not just do an acquire.
........
r4640 | zyasoft | 2008-06-15 01:13:59 -0400 (Sun, 15 Jun 2008) | 1 line

Changed Code.ldc such that for large string constants (>32767 char), it splits into chunked ldcsand assembles with a StringBuilder
........
r4642 | nriley | 2008-06-15 17:45:22 -0400 (Sun, 15 Jun 2008) | 1 line

jna-posix from http://svn.codehaus.org/jruby-contrib/trunk/jna-posix r179
........
r4643 | nriley | 2008-06-15 18:04:11 -0400 (Sun, 15 Jun 2008) | 1 line

os.stat, chmod, getegid, geteuid, getgid, getlogin, getpgrp, getpid, getppid, getuid, setpgrp, setsid via jna-posix
........
r4648 | nriley | 2008-06-16 01:23:39 -0400 (Mon, 16 Jun 2008) | 1 line

enable test_posix
........
r4652 | fwierzbicki | 2008-06-16 15:23:38 -0400 (Mon, 16 Jun 2008) | 3 lines

Extrated PythonTreeAdaptor from the *Parser classes, some
work to improve offsets.
........
r4653 | zyasoft | 2008-06-16 15:51:12 -0400 (Mon, 16 Jun 2008) | 1 line

Now supports coroutines. At this point, I have only done limited testing, especially with respect to exceptions.
........
r4654 | fwierzbicki | 2008-06-16 16:35:50 -0400 (Mon, 16 Jun 2008) | 2 lines

More offset work for IDE support.
........
r4655 | zyasoft | 2008-06-16 18:54:40 -0400 (Mon, 16 Jun 2008) | 1 line

Reordered public members of PyGenerator so that dir will have the same order as CPython
........
r4656 | fwierzbicki | 2008-06-16 19:11:25 -0400 (Mon, 16 Jun 2008) | 2 lines

Move some imaginary tokens to concrete tokens.
........
r4657 | zyasoft | 2008-06-16 20:21:52 -0400 (Mon, 16 Jun 2008) | 1 line

Added support of __future__.with_statement; this includes aligning PyTableCode with __future__.py as of 2.5. Does not yet support interactive future import of the with_statement, however
........
r4658 | fwierzbicki | 2008-06-16 20:26:20 -0400 (Mon, 16 Jun 2008) | 3 lines

pass statement turned to concrete token. Also preparing to absorb more
of PythonWalker.g into Python.g.
........
r4659 | pjenvey | 2008-06-16 20:42:38 -0400 (Mon, 16 Jun 2008) | 1 line

we want the absolute path here
........
r4660 | fwierzbicki | 2008-06-16 22:34:20 -0400 (Mon, 16 Jun 2008) | 3 lines

parameter passing that I am not using yet. Will be crucial for a
one pass parse.
........
r4661 | fwierzbicki | 2008-06-16 22:47:21 -0400 (Mon, 16 Jun 2008) | 2 lines

break and continue only newed in Python.g
........
r4662 | fwierzbicki | 2008-06-17 11:12:43 -0400 (Tue, 17 Jun 2008) | 2 lines

Slowly adding more needed params to Python.g from PythonWalker.g.
........
r4665 | fwierzbicki | 2008-06-18 10:59:31 -0400 (Wed, 18 Jun 2008) | 3 lines

Introduce an ENDMARK imaginary token to handle unfinished but
partially valid statements.
........
r4667 | fwierzbicki | 2008-06-18 16:21:10 -0400 (Wed, 18 Jun 2008) | 2 lines

Incremental improvement on interactive mode.
........
r4672 | fwierzbicki | 2008-06-18 21:26:41 -0400 (Wed, 18 Jun 2008) | 3 lines

Fix PythonPartialTester. Added some info to toStringTree output of
PythonTree.
........
r4676 | fwierzbicki | 2008-06-19 11:30:05 -0400 (Thu, 19 Jun 2008) | 3 lines

Added an ImaginaryToken type so that the char positions of
INDENT and DEDENT can be easily tracked.
........
r4678 | fwierzbicki | 2008-06-19 13:58:08 -0400 (Thu, 19 Jun 2008) | 2 lines

Replace many imaginary tokens with concrete tokens.
........
r4679 | fwierzbicki | 2008-06-19 18:31:55 -0400 (Thu, 19 Jun 2008) | 2 lines

whitespace
........
r4680 | fwierzbicki | 2008-06-19 19:50:49 -0400 (Thu, 19 Jun 2008) | 2 lines

Update comments and rule positions compared to CPython's Grammar/Grammar
........
r4682 | pjenvey | 2008-06-19 20:07:46 -0400 (Thu, 19 Jun 2008) | 1 line

fix whitespace
........
r4684 | pjenvey | 2008-06-19 20:50:11 -0400 (Thu, 19 Jun 2008) | 4 lines

compile inplace operations to a PyObject._ixx wrapper (which uses _binop_rule,
like _xxx), instead of __ixx__
for #1873148
........
r4693 | pjenvey | 2008-06-20 01:45:14 -0400 (Fri, 20 Jun 2008) | 5 lines

o fix builtin inplace methods to work as MethodType.BINARY
o add missing inplace functions to operator module
o fix CPython specific usage of __iadd__ in lists_tests to operator.__iadd__
for #1873148
........
r4694 | pjenvey | 2008-06-20 02:01:14 -0400 (Fri, 20 Jun 2008) | 3 lines

make deriveds inplace ops work like binary ops
for #1873148
........
r4695 | zyasoft | 2008-06-20 02:21:29 -0400 (Fri, 20 Jun 2008) | 1 line

Updated most builtins to 2.5; intern, object do not yet conform. Float conversion is richer in Java than it is in Python for hex (Java supports floating point hex values, test_float). Derived longs are failing (test_longconversion). cmp might be fixed by Leo Soto. Remaining issues are probably due to partial parse problems. Added contextlib, functools. The latter uses a pure-Python implementation, so it does not look like a native type for some of the tests in test_functools; TBD.
........
r4704 | fwierzbicki | 2008-06-20 13:32:49 -0400 (Fri, 20 Jun 2008) | 3 lines

Cleared up some bad behavior in Python.g (was reporting and recovering
from missing tokens).
........
r4705 | fwierzbicki | 2008-06-20 13:35:59 -0400 (Fri, 20 Jun 2008) | 2 lines

Switch debugging off.
........
r4706 | fwierzbicki | 2008-06-20 14:33:18 -0400 (Fri, 20 Jun 2008) | 2 lines

Repaired partial strings. Partial lists, dicts, etc *sometimes* work.
........
r4708 | fwierzbicki | 2008-06-20 15:58:27 -0400 (Fri, 20 Jun 2008) | 2 lines

Pretty good nested [,(,{ support for interactive mode.
........
r4709 | zyasoft | 2008-06-20 21:16:24 -0400 (Fri, 20 Jun 2008) | 1 line

Added wsgiref from 2.5
........
r4710 | fwierzbicki | 2008-06-21 08:20:13 -0400 (Sat, 21 Jun 2008) | 2 lines

whitespace.
........
r4711 | fwierzbicki | 2008-06-21 08:39:06 -0400 (Sat, 21 Jun 2008) | 2 lines

Now simple tests of the interactive mode are working well.
........
r4712 | pjenvey | 2008-06-21 19:42:05 -0400 (Sat, 21 Jun 2008) | 1 line

always re-generate .class until APIVersion is fixed
........
r4714 | nriley | 2008-06-22 22:38:45 -0400 (Sun, 22 Jun 2008) | 1 line

os.isatty, file.isatty support
........
r4715 | fwierzbicki | 2008-06-23 11:57:31 -0400 (Mon, 23 Jun 2008) | 2 lines

Fix for one line compound statements in interactive mode.
........
r4720 | fwierzbicki | 2008-06-23 15:28:04 -0400 (Mon, 23 Jun 2008) | 2 lines

return None for getfilesystemencoding for now.
........
r4729 | pjenvey | 2008-06-23 18:06:02 -0400 (Mon, 23 Jun 2008) | 7 lines

merge trunk/jython@4725:branches/pep352@4725 to branches/asm
branches/pep352 failures:
test_coercion test_deque test_doctest test_exceptions test_extcall
test_generators test_pickletools
new asm failures:
test_coercion test_deque test_pickletools test_wsgiref
........
r4732 | fwierzbicki | 2008-06-23 19:41:21 -0400 (Mon, 23 Jun 2008) | 2 lines

moving to 2.5 Lib. Also small fix for encoding check.
........
r4733 | pjenvey | 2008-06-23 20:31:42 -0400 (Mon, 23 Jun 2008) | 3 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_exceptions.py@52486
........
r4734 | pjenvey | 2008-06-23 20:45:53 -0400 (Mon, 23 Jun 2008) | 3 lines

o 2.5 EnvironmentError needs a custom __reduce__
o disable failing testInfiniteRecursion
........
r4735 | fwierzbicki | 2008-06-23 21:14:18 -0400 (Mon, 23 Jun 2008) | 2 lines

better order for suite subrule.
........
r4740 | pjenvey | 2008-06-23 22:07:21 -0400 (Mon, 23 Jun 2008) | 3 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_support.py@60337
........
r4741 | fwierzbicki | 2008-06-23 22:10:11 -0400 (Mon, 23 Jun 2008) | 4 lines

Realized that CPython forces an extra newline for *all* compound
statments (including statments like "class Foo:pass"). Greatly
simplified the grammar.
........
r4742 | pjenvey | 2008-06-23 22:14:19 -0400 (Mon, 23 Jun 2008) | 1 line

merge Jython specific changes back into 2.5 test_support
........
r4743 | fwierzbicki | 2008-06-23 22:36:57 -0400 (Mon, 23 Jun 2008) | 2 lines

Fix exploding newlines on interactive prompt.
........
r4744 | pjenvey | 2008-06-23 22:43:05 -0400 (Mon, 23 Jun 2008) | 1 line

re-enable setdict, fixed by pep352
........
r4745 | pjenvey | 2008-06-24 03:01:10 -0400 (Tue, 24 Jun 2008) | 3 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_optparse.py@50791
........
r4746 | pjenvey | 2008-06-24 03:03:33 -0400 (Tue, 24 Jun 2008) | 1 line

allow imports outside of test
........
r4747 | pjenvey | 2008-06-24 03:04:43 -0400 (Tue, 24 Jun 2008) | 1 line

don't assume __builtins__ and skip a CPython specific test
........
r4748 | fwierzbicki | 2008-06-24 09:59:06 -0400 (Tue, 24 Jun 2008) | 2 lines

Finally - the threshold for a version change has been hit. Yay!
........
r4749 | zyasoft | 2008-06-24 11:45:48 -0400 (Tue, 24 Jun 2008) | 1 line

Use CPythonLib's re.py, not our own
........
r4750 | fwierzbicki | 2008-06-24 12:46:35 -0400 (Tue, 24 Jun 2008) | 2 lines

Support longs regardless of radix.
........
r4751 | fwierzbicki | 2008-06-24 13:10:39 -0400 (Tue, 24 Jun 2008) | 3 lines

move the Jython specific test_array out of the way so we can test against
the CPython test_array
........
r4753 | pjenvey | 2008-06-24 17:31:29 -0400 (Tue, 24 Jun 2008) | 1 line

disable test_asynchat for now, re #1064
........
r4754 | pjenvey | 2008-06-24 17:31:42 -0400 (Tue, 24 Jun 2008) | 1 line

small cleanup
........
r4755 | pjenvey | 2008-06-24 19:12:17 -0400 (Tue, 24 Jun 2008) | 3 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_logging.py@54275
........
r4756 | pjenvey | 2008-06-24 19:13:31 -0400 (Tue, 24 Jun 2008) | 1 line

re-integrate r4098:4167
........
r4757 | pjenvey | 2008-06-24 19:57:42 -0400 (Tue, 24 Jun 2008) | 2 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/doctest.py@59174
........
r4758 | pjenvey | 2008-06-24 20:11:28 -0400 (Tue, 24 Jun 2008) | 5 lines

o bring in decimal (test_decimal passes). also hashlib and platform
which don't quite work yet
o Jython changes for doctest, though test_doctest still fails. bring
in trace for test_doctest
........
r4761 | fwierzbicki | 2008-06-25 13:22:09 -0400 (Wed, 25 Jun 2008) | 3 lines

removed special handling of RecognitionException, since errors
are always of type ParseException now anyway.
........
r4762 | fwierzbicki | 2008-06-25 13:51:08 -0400 (Wed, 25 Jun 2008) | 3 lines

Created a flag so that external tools can use Python.g with
Antlr's default error handling.
........
r4763 | fwierzbicki | 2008-06-25 14:10:21 -0400 (Wed, 25 Jun 2008) | 4 lines

antlrErrorHandling needs to be public to be used by external tools.
Looking at the nice error messages that are coming out makes me
wonder if I might want to rethink my error handling mechanism.
........
r4764 | pjenvey | 2008-06-25 15:03:08 -0400 (Wed, 25 Jun 2008) | 1 line

reapply compiled name as <module> from r4321
........
r4765 | pjenvey | 2008-06-25 15:05:09 -0400 (Wed, 25 Jun 2008) | 2 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/inspect.py@56297
........
r4767 | pjenvey | 2008-06-25 15:45:51 -0400 (Wed, 25 Jun 2008) | 1 line

re-integrate r4053
........
r4768 | pjenvey | 2008-06-25 15:49:17 -0400 (Wed, 25 Jun 2008) | 3 lines

o fix string exceptions which I broke in r4754
o kill OverflowWarning and the custom warnings.py that we no longer need
........
r4769 | pjenvey | 2008-06-25 16:12:14 -0400 (Wed, 25 Jun 2008) | 1 line

match messages to CPython 2.5
........
r4770 | pjenvey | 2008-06-25 16:27:31 -0400 (Wed, 25 Jun 2008) | 1 line

small cleanup to unpackSequence
........
r4771 | pjenvey | 2008-06-25 16:46:49 -0400 (Wed, 25 Jun 2008) | 2 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/regrtest.py@54228
........
r4772 | pjenvey | 2008-06-25 16:48:27 -0400 (Wed, 25 Jun 2008) | 2 lines

re-integrate jython specifics: mostly just -e support and java platform
test defines
........
r4774 | fwierzbicki | 2008-06-25 19:08:30 -0400 (Wed, 25 Jun 2008) | 2 lines

Can't live without -m.
........
r4777 | pjenvey | 2008-06-25 20:12:35 -0400 (Wed, 25 Jun 2008) | 4 lines

skip a number of tests we won't pass anytime soon. mark most of the
cjkencodings tests and xml_etree as perm failures for now. re-enable
test_unicode as it's close to passing
........
r4778 | pjenvey | 2008-06-25 20:12:50 -0400 (Wed, 25 Jun 2008) | 1 line

test_uuid passes
........
r4779 | nriley | 2008-06-25 21:16:20 -0400 (Wed, 25 Jun 2008) | 1 line

Return permission denied for files opened for reading.
........
r4781 | nriley | 2008-06-25 21:47:39 -0400 (Wed, 25 Jun 2008) | 1 line

Match CPython behavior, hiding banner/prompt if stdin is not interactive and showing prompt if file argument is a tty; remove related workaround from test_cmd_line.
........
r4782 | fwierzbicki | 2008-06-26 13:11:56 -0400 (Thu, 26 Jun 2008) | 2 lines

hexoct is passing.
........
r4783 | fwierzbicki | 2008-06-26 13:15:03 -0400 (Thu, 26 Jun 2008) | 3 lines

Corner case for fplist fixed. def foo((x)) should be treated the same as
def foo(x).
........
r4784 | fwierzbicki | 2008-06-26 16:11:35 -0400 (Thu, 26 Jun 2008) | 2 lines

genexp should be sole argument if no parens. First half of fix.
........
r4785 | zyasoft | 2008-06-26 18:43:05 -0400 (Thu, 26 Jun 2008) | 1 line

Removed old Jython-specific tests in favor of CPythonLib versions (except that test_urllib.py has a dependency on fileno being int, so changed accordingly). Cleaned up SRE engine code such that it more closely follows _sre.c. Removed optimization of Unicode charset in sre_compile since for UCS-4 for CPython, it doesn't appear to use this optimization anyway.
........
r4786 | zyasoft | 2008-06-26 20:44:30 -0400 (Thu, 26 Jun 2008) | 1 line

re.sub, re.subn now coerce to the right type (unicode/str), based on the spec provided by test_re.test_bug_1140
........
r4787 | zyasoft | 2008-06-26 20:50:25 -0400 (Thu, 26 Jun 2008) | 1 line

Removed obsolete test, superseded by 2.5 coercion rules for emptry strings on re.sub
........
r4788 | pjenvey | 2008-06-26 21:14:25 -0400 (Thu, 26 Jun 2008) | 2 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_urllib2_localnet.py@55806
........
r4789 | pjenvey | 2008-06-26 21:15:47 -0400 (Thu, 26 Jun 2008) | 1 line

explicitly close the server socket instead of relying on GC
........
r4790 | zyasoft | 2008-06-26 21:26:29 -0400 (Thu, 26 Jun 2008) | 1 line

test_re now passes completely. This was done by commenting inapplicable tests re recursion limits, since we are not going to implement a non-recursive variant, as well as adding the unimplemented typecodes for PyArray. Too bad that there's a lot more to do to get array working as well!
........
r4791 | pjenvey | 2008-06-26 21:30:12 -0400 (Thu, 26 Jun 2008) | 2 lines

upgrade socket._fileobject to 2.5's
........
r4792 | pjenvey | 2008-06-26 22:39:20 -0400 (Thu, 26 Jun 2008) | 2 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_hashlib.py@39316
........
r4793 | pjenvey | 2008-06-26 22:52:59 -0400 (Thu, 26 Jun 2008) | 3 lines

add _hashlib with support for everything but sha224, via
java.security.MessageDigest. supersedes our md5 module
........
r4794 | pjenvey | 2008-06-26 23:04:45 -0400 (Thu, 26 Jun 2008) | 2 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_hmac.py@58870
........
r4795 | zyasoft | 2008-06-26 23:14:37 -0400 (Thu, 26 Jun 2008) | 1 line

Added stack_size to thread, threading since it's available since 1.4, but only as a hint to the JVM. test_re was not added in the prev commit. test_threading's test of foreign threads doesn't make sense for Jython, or so I decree
........
r4796 | pjenvey | 2008-06-26 23:25:09 -0400 (Thu, 26 Jun 2008) | 3 lines

optimize _hashlib.HASH to take advantage of MessageDigest.clone which is
supported for all of our hashing types
........
r4797 | pjenvey | 2008-06-26 23:26:55 -0400 (Thu, 26 Jun 2008) | 3 lines

o sha is superceded by hashlib
o special case test_hmac to avoid our lack of sha224 support
........
r4798 | zyasoft | 2008-06-26 23:50:14 -0400 (Thu, 26 Jun 2008) | 1 line

Fixed test_bigmem to reflect Java realities. Didn't get the commit in last time of commenting out test_foreign_thread, not applicable to Jython
........
r4799 | zyasoft | 2008-06-27 00:17:33 -0400 (Fri, 27 Jun 2008) | 1 line

Working too fast here, here's the test_bigmem.py fixes. Also added all methods to threading such that it passes tests in test_contextlib
........
r4800 | pjenvey | 2008-06-27 00:27:32 -0400 (Fri, 27 Jun 2008) | 2 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/fileinput.py@42507
........
r4801 | pjenvey | 2008-06-27 00:28:43 -0400 (Fri, 27 Jun 2008) | 1 line

reapply Jython workaround for lack of fstat
........
r4802 | zyasoft | 2008-06-27 00:30:59 -0400 (Fri, 27 Jun 2008) | 1 line

Enable testing of stack size changing for java platform
........
r4806 | zyasoft | 2008-06-27 00:50:06 -0400 (Fri, 27 Jun 2008) | 1 line

NA'ed test_traceback.test_members since it depends on sys.exc_traceback global var, which is not implemented in Jython (not threadsafe/deprecated since Python 1.5)
........
r4807 | zyasoft | 2008-06-27 01:28:40 -0400 (Fri, 27 Jun 2008) | 1 line

binascii.a2b_qp now escapes '= ' to '= '. a2b_uu takes empty strings
........
r4808 | pjenvey | 2008-06-27 01:45:44 -0400 (Fri, 27 Jun 2008) | 1 line

unused var
........
r4809 | zyasoft | 2008-06-27 03:24:14 -0400 (Fri, 27 Jun 2008) | 1 line

Added to os module: devnull, urandom. Worked around issue in java.io.File#mkdirs when adding directory like x/y/z/.
........
r4810 | fwierzbicki | 2008-06-27 13:47:50 -0400 (Fri, 27 Jun 2008) | 8 lines

Cleaned up Call node creation, and
now gen expressions are checked from both sides so that:

foo(x for x in range(10), 100)
foo(100, x for x in range(10))

properly fail.
........
r4821 | zyasoft | 2008-06-30 03:06:55 -0400 (Mon, 30 Jun 2008) | 1 line

Rewrote cStringIO to use StringBuffer instead of managing its own buffer (also not StringBuilder)
........
r4822 | zyasoft | 2008-06-30 03:41:08 -0400 (Mon, 30 Jun 2008) | 1 line

Fixed cStringIO to support ops for cPickle
........
r4823 | zyasoft | 2008-06-30 04:18:38 -0400 (Mon, 30 Jun 2008) | 1 line

Fixed unicode coercion for str.__mod__ with unicode args
........
r4824 | zyasoft | 2008-06-30 04:59:43 -0400 (Mon, 30 Jun 2008) | 1 line

Fixed unicode coercion logic for PyString.__mod__. Removed 2.5 contextlib.py in favor of direct from CPythonLib
........
r4825 | zyasoft | 2008-06-30 12:14:30 -0400 (Mon, 30 Jun 2008) | 1 line

Change cStringIO to use StringBuilder semantics. Use via thread confinement, or with external locking. Keep it simple, keep it lightweight.
........
r4826 | zyasoft | 2008-06-30 14:08:37 -0400 (Mon, 30 Jun 2008) | 1 line

Restored cStringIO.getString
........
r4827 | zyasoft | 2008-07-01 09:52:38 -0400 (Tue, 01 Jul 2008) | 1 line

Removed modules and tests exactly duplicated from CPythonLib
........
r4828 | zyasoft | 2008-07-01 11:50:43 -0400 (Tue, 01 Jul 2008) | 1 line

Removed modules/tests that are trivially changed with respect to CPythonLib
........
r4829 | zyasoft | 2008-07-01 13:21:43 -0400 (Tue, 01 Jul 2008) | 1 line

Removed modules/tests that should be updated to CPythonLib. Incidentally, test_cmath now tests that cmath.log takes a base
........
r4830 | zyasoft | 2008-07-01 16:38:05 -0400 (Tue, 01 Jul 2008) | 1 line

test_mailbox has been rewritten since 2.2, the version that was committed
........
r4831 | zyasoft | 2008-07-01 17:12:34 -0400 (Tue, 01 Jul 2008) | 1 line

Removed spurious placement of test_bigmem
........
r4832 | leosoto | 2008-07-01 17:30:12 -0400 (Tue, 01 Jul 2008) | 1 line

pyBoolean.__tojava__(Object.class) now returns a Boolean instead of an Integer. Also used valueOf static methods instead of constructors.
........
r4833 | zyasoft | 2008-07-01 18:02:55 -0400 (Tue, 01 Jul 2008) | 1 line

cStringIO#readline should not insert backslash-n (ambigous reading, but CPython implementation says otherwise)
........
r4834 | leosoto | 2008-07-01 18:24:50 -0400 (Tue, 01 Jul 2008) | 1 line

zxJDBC: Added BigInteger support to the PostgreSQLDataHandler
........
r4835 | leosoto | 2008-07-01 19:38:37 -0400 (Tue, 01 Jul 2008) | 1 line

zxJDBC: Avoid preparing a statement when executemany() is called with an empty sequence
........
r4836 | leosoto | 2008-07-01 20:33:20 -0400 (Tue, 01 Jul 2008) | 1 line

Workaround for issue 1047. At least on my system (jdk 1.6.0_06) it makes pulldom usable
........
r4837 | leosoto | 2008-07-01 23:51:43 -0400 (Tue, 01 Jul 2008) | 1 line

Another string interpolation fix, this time when the left operand is an unicode containing non-ascii chars and there are no unicodes on the right argument
........
r4838 | zyasoft | 2008-07-02 01:56:10 -0400 (Wed, 02 Jul 2008) | 1 line

cmath.log now takes a base parameter; all functions now test for ValueError to reraise as a TypeError, to be consistent with CPython
........
r4839 | leosoto | 2008-07-02 16:17:29 -0400 (Wed, 02 Jul 2008) | 1 line

Updated eclipse configuration to use the antlr3.1b1 JAR instead of 3.0.1
........
r4840 | leosoto | 2008-07-02 17:09:14 -0400 (Wed, 02 Jul 2008) | 1 line

Syncing PyBaseExceptionDerived
........
r4841 | leosoto | 2008-07-02 17:43:16 -0400 (Wed, 02 Jul 2008) | 1 line

Fixing Derived#__cmp__ to avoid raise TypeError when __cmp__ has not been overrided and super.__cmp__() returns -2 (see #1031)
........
r4842 | leosoto | 2008-07-02 20:36:50 -0400 (Wed, 02 Jul 2008) | 1 line

Enable utf-8 source encoding support (the test_cookielib issue was fixed upstream: http://bugs.python.org/issue3261). Also recognize 'latin-1' as an alias for 'iso8859-1'
........
r4843 | leosoto | 2008-07-02 22:39:07 -0400 (Wed, 02 Jul 2008) | 1 line

Limiting the buffer size used for discovering the encoding, to avoid crashing when reading big files (such as django/tests/regressiontests/forms/fields.py), as the default size of BufferedReader was too close to the readlimit set by the mark() call
........
r4845 | fwierzbicki | 2008-07-03 21:13:07 -0400 (Thu, 03 Jul 2008) | 3 lines

change token name from ENDMARK to ENDMARKER to better fit with
CPython's Grammar/Grammar file. Alse some tabs->spaces.
........
r4846 | fwierzbicki | 2008-07-03 21:49:41 -0400 (Thu, 03 Jul 2008) | 3 lines

Introduced ENDMARKER to Python.g -- needed for files that
lack a newline.
........
r4847 | fwierzbicki | 2008-07-03 22:01:44 -0400 (Thu, 03 Jul 2008) | 2 lines

import test_support.
........
r4848 | fwierzbicki | 2008-07-04 08:21:30 -0400 (Fri, 04 Jul 2008) | 3 lines

Removed ENDMARKER from Python.g and applied Leo Soto's patch to
add a newline to the end of input that ends w/o a newline.
........
r4849 | fwierzbicki | 2008-07-04 08:33:14 -0400 (Fri, 04 Jul 2008) | 2 lines

Formatting and remove unused variable.
........
r4850 | fwierzbicki | 2008-07-04 17:25:24 -0400 (Fri, 04 Jul 2008) | 3 lines

Moved the actual emit of the generated newline (for input
w/o a newline) to *after* dedents are emitted.
........
r4852 | leosoto | 2008-07-05 20:37:12 -0400 (Sat, 05 Jul 2008) | 20 lines

StringFormatter: '%d' and '%f' support for the __int__ and __float__ protocol
respectively.

The implementation is more convulted than it should be, because we have PyString
implementing __float__ and __int__ at the "java level" but not at the "python
level". For string formatting, only "python level" __float__ and __int__ must be
supported.

Also, considering that __int__ can return a PyLong, this case needs special
care. Basically formatInteger now can call formatLong if a PyLong is found as
the result of calling __int__. Then, as formatLong can also be called from
formatInteger, __hex__, __oct__ and __str__ conversions were moved inside
formatLong.

Finally, test_format_jy was changed to stop checking that we don't support big
floats on '%d' (CPython doesn't, but that seems a limitation of the specific
implementation and I can't imagine a program that could break on Jython because we
*support* it).
........
r4853 | pjenvey | 2008-07-05 21:51:20 -0400 (Sat, 05 Jul 2008) | 1 line

always recompile bytecode from zipimporter too
........
r4855 | fwierzbicki | 2008-07-07 05:08:22 -0400 (Mon, 07 Jul 2008) | 3 lines

Create a jython-complete.jar to make it easier to put out
a 2.5 alpha.
........
r4856 | zyasoft | 2008-07-07 07:22:16 -0400 (Mon, 07 Jul 2008) | 1 line

Removed double-checked locking used to init ThreadStateMap; didn't make sense to do any locking here anyway. Moved trycatch emit in visitWith to be where it ahappens (doesn't change output, but makes it easier to follow code).
........
r4857 | pjenvey | 2008-07-07 15:59:32 -0400 (Mon, 07 Jul 2008) | 1 line

no need to go through __builtin__ for issubclass
........
r4859 | pjenvey | 2008-07-07 19:21:43 -0400 (Mon, 07 Jul 2008) | 1 line

simplify
........
r4863 | pjenvey | 2008-07-07 21:14:55 -0400 (Mon, 07 Jul 2008) | 2 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_enumerate.py@39608
........
r4864 | pjenvey | 2008-07-07 21:15:40 -0400 (Mon, 07 Jul 2008) | 1 line

avoid implementation detail tests
........
r4865 | pjenvey | 2008-07-08 01:20:37 -0400 (Tue, 08 Jul 2008) | 2 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_iter.py@33742
........
r4866 | pjenvey | 2008-07-08 01:24:17 -0400 (Tue, 08 Jul 2008) | 4 lines

o make StopIteration (and the like) a 'sink state' in iter and call-iter, to
fix the test_iter test_sinkstate_* tests
o workaround test_iter test_unpack_iter GC related issues
........
r4867 | pjenvey | 2008-07-08 01:37:05 -0400 (Tue, 08 Jul 2008) | 5 lines

o wrap the fstat usage in is_jython instead of commenting it out
o "sync" with:
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_largefile.py@33923
(no changes; equiv to 2.3's)
........
r4868 | pjenvey | 2008-07-08 02:20:46 -0400 (Tue, 08 Jul 2008) | 4 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_repr.py@58123
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_richcmp.py@43274
........
r4869 | pjenvey | 2008-07-08 03:27:03 -0400 (Tue, 08 Jul 2008) | 4 lines

o workaround jython issues in test_richcmp and test_repr
o our custom test_operations isn't really necessary and we pass stdlib's. also
now pass stdlib's test_long
........
r4870 | fwierzbicki | 2008-07-08 04:59:37 -0400 (Tue, 08 Jul 2008) | 2 lines

some tweaks to help the install task.
........
r4876 | pjenvey | 2008-07-08 15:55:47 -0400 (Tue, 08 Jul 2008) | 3 lines

always re-generating byte code is more trouble than it's worth right now
(breaks the imp module)
........
r4877 | pjenvey | 2008-07-08 16:38:02 -0400 (Tue, 08 Jul 2008) | 5 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_scope.py@50493
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_types.py@51140
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_zlib.py@62236
........
r4878 | pjenvey | 2008-07-08 17:05:48 -0400 (Tue, 08 Jul 2008) | 4 lines

o allow a length argument to zlib decompressobj flush
o workarounds for 2.5 test_scope test_types (which output now matches stdlib's)
and test_zlib
........
r4879 | pjenvey | 2008-07-08 20:41:01 -0400 (Tue, 08 Jul 2008) | 3 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/platform.py@62855
........
r4880 | pjenvey | 2008-07-08 20:42:59 -0400 (Tue, 08 Jul 2008) | 1 line

s/java.os/os/g
........
r4881 | nriley | 2008-07-08 23:50:19 -0400 (Tue, 08 Jul 2008) | 1 line

overriding fillInStackTrace for PyException speeds things up quite a bit
........
r4882 | nriley | 2008-07-10 01:12:05 -0400 (Thu, 10 Jul 2008) | 1 line

need bash to test the script, too
........
r4883 | leosoto | 2008-07-10 15:51:37 -0400 (Thu, 10 Jul 2008) | 1 line

zxJDBC: Map SQLExceptions to zxJDBC.IntegrityError if the sqlState says the exception was caused by a integrity violation
........
r4884 | pjenvey | 2008-07-10 23:36:29 -0400 (Thu, 10 Jul 2008) | 2 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_support.py@60337
........
r4885 | pjenvey | 2008-07-10 23:46:09 -0400 (Thu, 10 Jul 2008) | 3 lines

add a -j option to regrtest for capturing test output to JUnit XML files (one
file per TestCase or module import or doctest) in the specified dir
........
r4887 | fwierzbicki | 2008-07-11 07:52:55 -0400 (Fri, 11 Jul 2008) | 3 lines

move "not needed" checks into separate task because all such tasks *are* needed
during a full build regardless of up to date status of files.
........
r4888 | zyasoft | 2008-07-11 08:16:28 -0400 (Fri, 11 Jul 2008) | 1 line

Begin port of PyPy's _rawffi, which with their pure-Python ctypes wrapper, implements ctypes
........
r4889 | fwierzbicki | 2008-07-11 08:45:41 -0400 (Fri, 11 Jul 2008) | 4 lines

added svnant-jars by default for more convenient full-build.
Fixed jarjar and jar-complete (so jarjar only recompiles on
changes and jar-complete properly depends on jarjar).
........
r4890 | zyasoft | 2008-07-11 10:05:13 -0400 (Fri, 11 Jul 2008) | 1 line

From PyPy's test__rawffi, broke out standard ctypes_test code they use (to avoid using their compilation toolchain). Implemented minimum to get test__rawffi.test_short_addition to pass.
........
r4891 | pjenvey | 2008-07-11 15:37:19 -0400 (Fri, 11 Jul 2008) | 1 line

test_pep263 passes now
........
r4892 | pjenvey | 2008-07-11 16:10:32 -0400 (Fri, 11 Jul 2008) | 3 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/string_tests.py@55878
........
r4893 | pjenvey | 2008-07-11 16:16:53 -0400 (Fri, 11 Jul 2008) | 5 lines

o re-integrate workarounds for string_tests
o add rsplit
fixes #1023
thanks Nimish Telang
........
r4894 | pjenvey | 2008-07-11 20:10:28 -0400 (Fri, 11 Jul 2008) | 2 lines

fix unicode(str_subclass) not using its __unicode__ method
........
r4895 | pjenvey | 2008-07-11 20:13:15 -0400 (Fri, 11 Jul 2008) | 3 lines

fix some odder 2.5 tests of string count/find/rfind. rewrite join to act like
CPython's, mainly so: a.join([b]) is b
........
r4896 | pjenvey | 2008-07-11 20:33:21 -0400 (Fri, 11 Jul 2008) | 1 line

fix java exception names being prefixed with <unknown> in Python tracebacks
........
r4897 | pjenvey | 2008-07-11 20:53:44 -0400 (Fri, 11 Jul 2008) | 1 line

fake reload(sys), like reload(PyJavaClass). fix ClassCastException in reload
........
r4901 | pjenvey | 2008-07-11 21:28:39 -0400 (Fri, 11 Jul 2008) | 1 line

expect test_mhlib to pass on posix and skip elsewhere
........
r4902 | zyasoft | 2008-07-12 05:13:38 -0400 (Sat, 12 Jul 2008) | 1 line

Removed tests not appropriate for Jython, test_peepholer, test_profile; added test_strptime, test_csv
........
r4903 | zyasoft | 2008-07-12 06:08:11 -0400 (Sat, 12 Jul 2008) | 1 line

Replaced Lib/datetime.py with version from PyPy, http://codespeak.net/svn/pypy/dist/pypy/lib/datetime.py@56484
........
r4904 | zyasoft | 2008-07-12 09:50:19 -0400 (Sat, 12 Jul 2008) | 1 line

Fixed str|string|unicode.replace for PyString or PyUnicode in the basic plane
........
r4905 | zyasoft | 2008-07-12 09:51:06 -0400 (Sat, 12 Jul 2008) | 1 line

Removed tests not relevant to Jython around CPython code compilation
........
r4906 | zyasoft | 2008-07-12 10:09:20 -0400 (Sat, 12 Jul 2008) | 1 line

Operations on unicode objs in basic plane can only result in basic plane objs (btw, not true of astral)
........
r4907 | zyasoft | 2008-07-12 11:42:09 -0400 (Sat, 12 Jul 2008) | 1 line

Added datetime.strptime
........
r4908 | zyasoft | 2008-07-12 11:55:44 -0400 (Sat, 12 Jul 2008) | 1 line

datetime.fromtimestamp now rounds seconds,microseconds within 1 microsecond of an integer second to that second
........
r4909 | zyasoft | 2008-07-12 12:50:22 -0400 (Sat, 12 Jul 2008) | 1 line

ascii codec will always be in basic plane, so at least optimize setting of that when constructing unicode
........
r4910 | leosoto | 2008-07-12 15:50:43 -0400 (Sat, 12 Jul 2008) | 1 line

Now test_strptime pass. This is done by improving the current SimpleDateFormat-backed implementation a bit, and delegating to _strptime.strptime for formats not supported by SimpleDateFormat.
........
r4911 | leosoto | 2008-07-12 15:52:59 -0400 (Sat, 12 Jul 2008) | 1 line

Removing a duplicate line on CPythonLib.includes I introduced on r4910
........
r4912 | pjenvey | 2008-07-12 15:57:48 -0400 (Sat, 12 Jul 2008) | 4 lines

o add popen2.cleanup for CPython compat. and to fix test_popen2
o make threading.currentThread() is threading.currentThread() (with the help of
a WeakKeyDictionary) to allow _threading_local to work
........
r4913 | pjenvey | 2008-07-12 20:47:08 -0400 (Sat, 12 Jul 2008) | 4 lines

max_read_chunk of 10mb kills test_gzip.test_many_append's performance due to
repeated 10mb reads. CPython doesn't suffer from this -- maybe its allocator
recycles the same 10mb. lowering to 256K speeds test time from 15 to 1 second
........
r4914 | pjenvey | 2008-07-12 22:07:51 -0400 (Sat, 12 Jul 2008) | 1 line

add C_EXTENSION because pkgutil references it
........
r4916 | pjenvey | 2008-07-12 23:36:31 -0400 (Sat, 12 Jul 2008) | 2 lines

fix to make the 'g' formatter closer to CPython's. fixes test_tokenize
........
r4917 | zyasoft | 2008-07-14 06:47:01 -0400 (Mon, 14 Jul 2008) | 1 line

split and rsplit do not need to be UTF-16 aware
........
r4918 | zyasoft | 2008-07-14 06:52:18 -0400 (Mon, 14 Jul 2008) | 3 lines

From
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_str@64946
........
r4919 | zyasoft | 2008-07-14 07:03:54 -0400 (Mon, 14 Jul 2008) | 1 line

Removed inapplicable test due to different algorithm in expandtabs which cannot overflow in Jython
........
r4922 | zyasoft | 2008-07-14 07:44:28 -0400 (Mon, 14 Jul 2008) | 1 line

time.struct_time can now take any obj that can be turned into a list and converted into 9 integers
........
r4924 | zyasoft | 2008-07-14 08:11:26 -0400 (Mon, 14 Jul 2008) | 3 lines

From
http://svn.python.org/projects/python/branches/release25-maint/Lib/uu.py@64946
........
r4925 | zyasoft | 2008-07-14 08:19:10 -0400 (Mon, 14 Jul 2008) | 2 lines

Fixed uu.encode, decode so that it does not rely on refcounting to close files
........
r4928 | zyasoft | 2008-07-14 11:19:01 -0400 (Mon, 14 Jul 2008) | 3 lines

ReferenceQueue objs, by their nature, only make sense if transient.
So make them so, this enables interoperation with Terracotta.
........
r4930 | pjenvey | 2008-07-14 15:24:49 -0400 (Mon, 14 Jul 2008) | 1 line

couple small array fixes
........
r4932 | fwierzbicki | 2008-07-14 15:25:59 -0400 (Mon, 14 Jul 2008) | 2 lines

unbreak --memo again :)
........
r4933 | pjenvey | 2008-07-14 15:26:58 -0400 (Mon, 14 Jul 2008) | 3 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/distutils@64955
........
r4934 | fwierzbicki | 2008-07-14 15:27:40 -0400 (Mon, 14 Jul 2008) | 3 lines

generate newlines before the dedents again. Although the reverse
was solving some problems, it was creating more.
........
r4935 | pjenvey | 2008-07-14 15:31:38 -0400 (Mon, 14 Jul 2008) | 3 lines

o re-integrate Jython distutils changes
o check for $py.class on Jython instead of pyc in test_build_py
........
r4936 | leosoto | 2008-07-14 16:12:27 -0400 (Mon, 14 Jul 2008) | 1 line

Allow derived types written in Jython to implement a __tojava__ method. This will allow better integration with java for python types such as decimal.Decimal, datetime.date, etc.
........
r4937 | fwierzbicki | 2008-07-15 11:23:32 -0400 (Tue, 15 Jul 2008) | 4 lines

Fixes in build.xml to make full-build work again. Some cleanup
like the removal of old parser and old website tasks. Changed
README to describe 2.5 alpha release.
........
r4939 | leosoto | 2008-07-15 12:13:50 -0400 (Tue, 15 Jul 2008) | 1 line

decimal.py from http://svn.python.org/projects/python/branches/release25-maint/Lib/decimal.py@64967
........
r4940 | leosoto | 2008-07-15 12:24:18 -0400 (Tue, 15 Jul 2008) | 1 line

Implemented decimal.Decimal.__tojava__ to convert decimal instances to java BigDecimal when python -> java conversion is requested
........
r4941 | leosoto | 2008-07-15 13:16:16 -0400 (Tue, 15 Jul 2008) | 1 line

Implemented __tojava__ on datetime.time, datetime.date and datetime.datetime. On the current implementation, they can be converted to java.sql.Time, java.sql.Date, java.sql.Timestamp, respectively. All of them can also be converted to java.util.Calendar. Only one caveat: the tzinfo attribute of times and datetimes is currently ignored
........
r4942 | leosoto | 2008-07-15 14:00:57 -0400 (Tue, 15 Jul 2008) | 1 line

Removing decimal.py from CPythonLib.includes. I forgot to do that on r4939
........
r4943 | fwierzbicki | 2008-07-15 14:33:40 -0400 (Tue, 15 Jul 2008) | 2 lines

build tweaks + workaround for an NT problem in site.py
........
r4944 | nriley | 2008-07-15 14:42:08 -0400 (Tue, 15 Jul 2008) | 1 line

fix for startup problem on Windows - AttributeError was accidentally trapping posixpath abspath failure
........
r4945 | fwierzbicki | 2008-07-15 15:27:53 -0400 (Tue, 15 Jul 2008) | 3 lines

already incrementing to 2.5a1 -- sourceforge is caching the old one too hard
and I really need to get Windows fixed.
........
r4947 | leosoto | 2008-07-15 17:21:30 -0400 (Tue, 15 Jul 2008) | 18 lines

Improvements to the zxJDBC default DataHandler:

- Map DATE, TIME and TIMESTAMP columns to datetime.date, datetime.time and
datetime.datetime respectively. Factory methods to create instances of
datetime.* were added to Py.java.

- Map VARCHAR columns to unicode instead of str.

- Don't try to guess the precision of NUMERIC columns .

- Use Py.True and Py.False instead of Py.One and Py.Zero.

Some of this changes may cause backwards-incompatibility, so I've copied the old
DataHandler logic into Jython22DataHandler. Then, old code should simply setup
this datahandler on its connections if it is not going to be adapted to the new
DataHandler default behaviour.
........
r4949 | amak | 2008-07-15 17:33:35 -0400 (Tue, 15 Jul 2008) | 1 line

Merging support for IPv6 and UDP <broadcast>
........
r4950 | nriley | 2008-07-15 18:33:04 -0400 (Tue, 15 Jul 2008) | 1 line

(Partial) Windows port of Jython launcher script; should work on Windows 2000 and later.
........
r4951 | nriley | 2008-07-15 18:37:38 -0400 (Tue, 15 Jul 2008) | 1 line

Oops, left some debugging code in.
........
r4952 | amak | 2008-07-15 18:51:51 -0400 (Tue, 15 Jul 2008) | 1 line

Backing out what I thought were simple changes, which appear to have broken the build.
........
r4959 | pjenvey | 2008-07-16 21:23:59 -0400 (Wed, 16 Jul 2008) | 1 line

add symlink on posix
........
r4961 | pjenvey | 2008-07-16 23:21:35 -0400 (Wed, 16 Jul 2008) | 5 lines

fix the symlink related tests by adding readlink, using _posix.lstat and
utilizing _posix symlink stuff in posixpath.
also utilize real stat results in posixpath, and re-enable ismount in ntpath
since we can support it in posixpath
........
r4963 | pjenvey | 2008-07-17 00:59:53 -0400 (Thu, 17 Jul 2008) | 9 lines

o fix PythonPOSIXHandler getEnv/getCurrentWorkingDirectory return signatures
o fix posixpath.__all__
o only enable os.symlink/readlink and posixpath friends if the posix factory
appears native. pure java support for symlinks is still sketchy -- in
particular with dead links -- switching these on enables tests in test_glob,
test_posixpath, test_tarfile that utilize dead links and are doomed to failure
o allow partial filling of stat results from jna-posix, padding the rest with
0s
........
r4965 | fwierzbicki | 2008-07-17 17:19:56 -0400 (Thu, 17 Jul 2008) | 2 lines

added custom toStringTree to each node.
........
r4967 | pjenvey | 2008-07-17 20:28:49 -0400 (Thu, 17 Jul 2008) | 3 lines

boost the mark limit even higher for the lenghty lines in
pygments.lexers._vimbuiltins
........
r4968 | nriley | 2008-07-18 15:20:39 -0400 (Fri, 18 Jul 2008) | 1 line

Work with jython-complete.jar; handle spaces in $JAVA_HOME.
........
r4969 | nriley | 2008-07-18 15:21:37 -0400 (Fri, 18 Jul 2008) | 1 line

Actually use JYTHON_OPTS; correct error message and properly delete _JYTHON_HOME.
........
r4970 | nriley | 2008-07-18 16:40:17 -0400 (Fri, 18 Jul 2008) | 1 line

Make ant regrtest work on Windows again.
........
r4971 | nriley | 2008-07-18 16:51:00 -0400 (Fri, 18 Jul 2008) | 1 line

Conditionalize launcher test; no need for Cygwin test.
........
r4973 | nriley | 2008-07-19 01:54:20 -0400 (Sat, 19 Jul 2008) | 1 line

Clean up install targets; allow install to be run without full-build; normalize source directory structure between checked-out and installed versions.
........
r4974 | fwierzbicki | 2008-07-19 18:48:21 -0400 (Sat, 19 Jul 2008) | 4 lines

Make PythonTree's toStringTree more easily comparable with ast/astview.py
output. This will help with the walkerless Python.g, as it doesn't quite
parse ast/astview.py yet.
........
r4980 | pjenvey | 2008-07-21 01:00:49 -0400 (Mon, 21 Jul 2008) | 1 line

integrate the sink state conditions
........
r4982 | pjenvey | 2008-07-21 15:11:48 -0400 (Mon, 21 Jul 2008) | 1 line

fix exposed __getitem__ possibly returning null
........
r4983 | pjenvey | 2008-07-21 15:45:12 -0400 (Mon, 21 Jul 2008) | 3 lines

o fix filter to pass test_builtin.test_filter and test_filter_subclasses
o remove the unnecessary __builtin__.tuple
........
r4984 | pjenvey | 2008-07-21 15:48:17 -0400 (Mon, 21 Jul 2008) | 1 line

make the failure verbose because test_site_flag occasionally fails on the buildbot mysteriously
........
r4985 | pjenvey | 2008-07-22 21:18:53 -0400 (Tue, 22 Jul 2008) | 4 lines

o make Py.makeCharacter consisitent in never returning unicode unless it was asked for
o move calculateIndex into PySequence for PyArray
o allow PyLong.getLong access for PyArray
........
r4986 | pjenvey | 2008-07-22 21:19:34 -0400 (Tue, 22 Jul 2008) | 2 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_array.py@60839
........
r4987 | pjenvey | 2008-07-22 21:21:32 -0400 (Tue, 22 Jul 2008) | 1 line

many various PyArray fixes for 2.5 test_array
........
r4988 | fwierzbicki | 2008-07-22 21:58:01 -0400 (Tue, 22 Jul 2008) | 4 lines

Made Str offsets better match CPython.
PythonTree now extends BaseTree (too much of CommonTree was getting changed anyway).
astdump.py repaired so that it is a useful testing ground for ast comparisons.
........
r4989 | fwierzbicki | 2008-07-23 14:39:09 -0400 (Wed, 23 Jul 2008) | 2 lines

Offset fixes for Try/Catch and Call.
........
r4990 | fwierzbicki | 2008-07-23 15:13:04 -0400 (Wed, 23 Jul 2008) | 6 lines

Made except blocks optional in PythonPartial.g since:

try:

Is a valid partial parse.
........
r4991 | pjenvey | 2008-07-23 16:36:42 -0400 (Wed, 23 Jul 2008) | 2 lines

o fix passing chars > 256 to Py.makeCharacter
o make struct.error a class and raise struct.errors instead of OverflowErrors
........
r4992 | pjenvey | 2008-07-23 16:55:03 -0400 (Wed, 23 Jul 2008) | 3 lines

support the new float.__get/setformat__ classmethods -- they just claim our
formats are 'unknown'
........
r4993 | pjenvey | 2008-07-24 03:22:38 -0400 (Thu, 24 Jul 2008) | 2 lines

fix a regression from the bug 1140 fix
........
r4994 | pjenvey | 2008-07-24 03:29:54 -0400 (Thu, 24 Jul 2008) | 3 lines

add codecs.charmap_build which builds optimized encoding maps now used by the
charmap encoder
........
r4995 | fwierzbicki | 2008-07-24 09:51:47 -0400 (Thu, 24 Jul 2008) | 2 lines

Make stopCharIndex consistent.
........
r4996 | fwierzbicki | 2008-07-24 12:58:24 -0400 (Thu, 24 Jul 2008) | 2 lines

Fix for wrong line numbers in func.co_firstlineno.
........
r4997 | pjenvey | 2008-07-24 19:42:09 -0400 (Thu, 24 Jul 2008) | 1 line

cleanup unused var
........
r4998 | nriley | 2008-07-25 02:07:24 -0400 (Fri, 25 Jul 2008) | 1 line

Suppress exception truncating /dev/null when opening it for write on Linux and Solaris; should fix test_os.
........
r4999 | fwierzbicki | 2008-07-25 21:27:25 -0400 (Fri, 25 Jul 2008) | 3 lines

Re-enabled APIVersion checking in imp.java. Used ASM to read and write APIVersion
as an annotation. Incremented APIVersion from 12 to 13.
........
r5000 | fwierzbicki | 2008-07-25 21:34:48 -0400 (Fri, 25 Jul 2008) | 2 lines

Small cleanups in imp.
........
r5001 | fwierzbicki | 2008-07-25 21:51:48 -0400 (Fri, 25 Jul 2008) | 2 lines

Small cleanup in Py.java -- killing stray System.outs.
........
r5002 | fwierzbicki | 2008-07-25 21:55:04 -0400 (Fri, 25 Jul 2008) | 2 lines

Removed stray System.outs.
........
r5003 | fwierzbicki | 2008-07-26 10:41:46 -0400 (Sat, 26 Jul 2008) | 2 lines

Changing comment convention back to proper style (start with Returns).
........
r5004 | fwierzbicki | 2008-07-27 09:21:12 -0400 (Sun, 27 Jul 2008) | 2 lines

comment out DEBUG.class in the repo.
........
r5005 | fwierzbicki | 2008-07-27 22:41:38 -0400 (Sun, 27 Jul 2008) | 2 lines

Better error messages in the face of parser errors (line numbers + char column).
........
r5006 | pjenvey | 2008-07-28 00:56:03 -0400 (Mon, 28 Jul 2008) | 3 lines

o add os.link (fixes test_shutil)
o fix test_re's test output -- it shouldn't have any now
........
r5007 | pjenvey | 2008-07-28 02:38:44 -0400 (Mon, 28 Jul 2008) | 7 lines

PEP 357 (__index__), adds:
o PyObject.isIndex - true if int, long or implements __index__
o PyObject.asIndex - like asInt for indexes
o also fixed a dispatching to __get/set/delitem__ vs __get/set/delslice__ bug
and __get/set/delitem__ indices weren't utilizing __len__ when available (fixes
test_userstring)
........
r5008 | pjenvey | 2008-07-28 02:40:09 -0400 (Mon, 28 Jul 2008) | 2 lines

deriveds changes for __index__
........
r5009 | zyasoft | 2008-07-28 07:51:51 -0400 (Mon, 28 Jul 2008) | 3 lines

test__rawffi will now skip if it cannot load ctypes_test, a test
dynamic library that we do not yet have an ant task to create.
........
r5010 | zyasoft | 2008-07-28 08:13:52 -0400 (Mon, 28 Jul 2008) | 3 lines

Replaced test_inspect.py with rewritten version (using unittest!) from 2.5:
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_inspect.py@65262
........
r5011 | zyasoft | 2008-07-28 09:33:50 -0400 (Mon, 28 Jul 2008) | 5 lines

Excluded test cases that depend on tuple unpacking in the arg
specification, since this depends on introspecing bytecode. (Perhaps
we should provide a Jython-specific version where appropriate metadata
is emitted in the compiled code object?)
........
r5012 | pjenvey | 2008-07-28 15:46:39 -0400 (Mon, 28 Jul 2008) | 5 lines

fix stock UserList/test_userlist:
o handle __get/set/delslice__ start/stop as null or None in PySlice.indices2
o fix list slice assignments of the same underlying list (always copy first
as CPython does)
........
r5013 | pjenvey | 2008-07-28 15:50:43 -0400 (Mon, 28 Jul 2008) | 1 line

include UserList
........
r5014 | nriley | 2008-07-28 18:55:35 -0400 (Mon, 28 Jul 2008) | 1 line

jna-posix from http://svn.codehaus.org/jruby-contrib/trunk/jna-posix r210
........
r5015 | nriley | 2008-07-28 18:56:55 -0400 (Mon, 28 Jul 2008) | 1 line

pwd, grp; test_posix no longer skipped
........
r5016 | pjenvey | 2008-07-28 19:13:21 -0400 (Mon, 28 Jul 2008) | 1 line

fix chdir('/') on windows
........
r5017 | fwierzbicki | 2008-07-28 21:40:32 -0400 (Mon, 28 Jul 2008) | 2 lines

IParserHost is no longer used.
........
r5018 | fwierzbicki | 2008-07-29 08:31:46 -0400 (Tue, 29 Jul 2008) | 2 lines

Move testing classes out of main code.
........
r5019 | pjenvey | 2008-07-29 16:29:08 -0400 (Tue, 29 Jul 2008) | 4 lines

add an InputStream tie-in to TextIOBases, for parsing in universal newlines
mode
fixes #1082
........
r5020 | zyasoft | 2008-07-29 18:33:56 -0400 (Tue, 29 Jul 2008) | 5 lines

cStringIO.tell needs to return an int for it to be used by __len__
(perhaps that should be relaxed instead?). Found by testing against
PyAMF which uses cStringIO as a backing store for its
BufferedByteStream objs.
........
r5021 | pjenvey | 2008-07-29 18:40:05 -0400 (Tue, 29 Jul 2008) | 2 lines

fix readChunk reading more than CHUNK_SIZE, which'll blow up packReadahead
........
r5022 | zyasoft | 2008-07-30 11:46:31 -0400 (Wed, 30 Jul 2008) | 4 lines

Support inf, -inf, nan literals (and any
uppercasing). struct.pack("d") ducktypes like CPython and calls
__float__. This fixes a number of issues for PyAMF.
........
r5023 | zyasoft | 2008-07-30 15:57:16 -0400 (Wed, 30 Jul 2008) | 3 lines

Changed default format for __getformat__/__setformat__ to "IEEE,
big-endian", and supported switching back and forth with "unknown".
........
r5024 | pjenvey | 2008-07-30 18:39:24 -0400 (Wed, 30 Jul 2008) | 2 lines

add a builtin _functools.partial type
........
r5025 | pjenvey | 2008-07-30 19:00:32 -0400 (Wed, 30 Jul 2008) | 1 line

include functools
........
r5026 | fwierzbicki | 2008-07-31 09:10:57 -0400 (Thu, 31 Jul 2008) | 2 lines

Switch to Antlr 3.1b2.
........
r5027 | nriley | 2008-07-31 12:03:59 -0400 (Thu, 31 Jul 2008) | 1 line

Update Eclipse classpath to Antlr 3.1b2.
........
r5028 | fwierzbicki | 2008-07-31 12:14:55 -0400 (Thu, 31 Jul 2008) | 14 lines

Adding a Jython specific version of CommonErrorNode to use when there are errors
in the AST but we want to continue parsing (for now this is only of use by external
tools, but I'm considering making it an option since the information can be useful.

Thanks to Allan Davis for the original version of CommonErrorNode. I've refactored his
version so that PythonErrorNode does not repeat much of CommonErrorNode and put the
creation logic into PythonTreeAdaptor.

I have also added an interface ErrorHandler that allows the error handling in antlr to
be customized. FailFastHandler implements the default behavior of dying on the first
parse error. I also made the parse error output more useful, and factored a BaseParser
out of the org/python/antlr/*Parser classes.
........
r5029 | pjenvey | 2008-07-31 14:47:34 -0400 (Thu, 31 Jul 2008) | 3 lines

rearrange imports to fix test_threaded_import -- a bogus fix, but also what
CPython did for this problem
........
r5030 | pjenvey | 2008-07-31 15:50:45 -0400 (Thu, 31 Jul 2008) | 4 lines

o fix test_softspace and reset softspace (and append a newline) in various
places (equiv. of CPython's ceval.c::Py_FlushLine)
o fix traceback source lines without trailing newlines (fixes #1087)
........
r5031 | fwierzbicki | 2008-07-31 16:13:12 -0400 (Thu, 31 Jul 2008) | 2 lines

Fixed generator args check (antlr3.1b2 broke this, fixed it again).
........
r5032 | fwierzbicki | 2008-07-31 16:50:59 -0400 (Thu, 31 Jul 2008) | 3 lines

Tightened up signature of ErrorHandler and added a ListErrorHandler that
mimics Antlr default behavior for error handling.
........
r5033 | pjenvey | 2008-07-31 18:07:13 -0400 (Thu, 31 Jul 2008) | 1 line

deprecate complex divmod(), // and %
........
r5034 | pjenvey | 2008-07-31 19:16:37 -0400 (Thu, 31 Jul 2008) | 1 line

move __index__ out of the generated code section
........
r5035 | pjenvey | 2008-07-31 19:18:33 -0400 (Thu, 31 Jul 2008) | 1 line

sync with a couple of manual changes PyInstance has had
........
r5036 | pjenvey | 2008-07-31 19:32:03 -0400 (Thu, 31 Jul 2008) | 2 lines

support threading._RLock, as twisted uses it
fixes #1079
........
r5037 | pjenvey | 2008-07-31 19:49:50 -0400 (Thu, 31 Jul 2008) | 5 lines

fix a couple test_coercion cases:
o fix broken PyComplex.__rfloordiv__
o prevent stack overflows when old style classes return __coerce__ args in
reversed order
........
r5038 | pjenvey | 2008-07-31 19:57:11 -0400 (Thu, 31 Jul 2008) | 2 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_operator.py@54178
........
r5039 | pjenvey | 2008-07-31 19:59:25 -0400 (Thu, 31 Jul 2008) | 2 lines

o filter operator slice args through __index__
o workaround Jython's lack of a real repeat/concat binop in test_operator
........
r5040 | fwierzbicki | 2008-07-31 20:50:28 -0400 (Thu, 31 Jul 2008) | 2 lines

remove unneeded member.
........
r5041 | nriley | 2008-08-01 16:06:44 -0400 (Fri, 01 Aug 2008) | 1 line

Make the ihooks module importable.
........
r5042 | nriley | 2008-08-01 16:07:47 -0400 (Fri, 01 Aug 2008) | 1 line

Support one-argument builtin apply.
........
r5043 | nriley | 2008-08-01 16:22:22 -0400 (Fri, 01 Aug 2008) | 1 line

Replace python.prepath with python.path.
........
r5044 | fwierzbicki | 2008-08-01 16:37:46 -0400 (Fri, 01 Aug 2008) | 3 lines

Better handling of bad dedent. Not perfect since it throws the
wrong exception.
........
r5045 | pjenvey | 2008-08-01 17:47:30 -0400 (Fri, 01 Aug 2008) | 2 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_class.py@52663
........
r5046 | pjenvey | 2008-08-01 17:49:29 -0400 (Fri, 01 Aug 2008) | 1 line

fix instancemethod cmp to make test_class pass
........
r5047 | nriley | 2008-08-01 18:28:33 -0400 (Fri, 01 Aug 2008) | 1 line

Read python.executable from whole registry, not just top level.
........
r5048 | pjenvey | 2008-08-01 18:28:52 -0400 (Fri, 01 Aug 2008) | 2 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_descrtut.py@38295
........
r5049 | pjenvey | 2008-08-01 18:38:32 -0400 (Fri, 01 Aug 2008) | 3 lines

list.__nonzero__ needn't be exposed, couple test_descrtut fixes (though it
still fails due to parser problems)
........
r5050 | nriley | 2008-08-01 19:02:47 -0400 (Fri, 01 Aug 2008) | 1 line

JYTHONPATH support.
........
r5051 | nriley | 2008-08-02 04:48:11 -0400 (Sat, 02 Aug 2008) | 1 line

Handle scripts on stdin.
........
r5052 | nriley | 2008-08-02 04:49:57 -0400 (Sat, 02 Aug 2008) | 1 line

jna-posix from http://svn.codehaus.org/jruby-contrib/trunk/jna-posix r213 + patch in http://jira.codehaus.org/browse/JRUBY-2885 for isatty on Windows
........
r5053 | pjenvey | 2008-08-02 15:51:15 -0400 (Sat, 02 Aug 2008) | 5 lines

switch some new style unary ops to throw TypeErrors instead of AttributeErrors
when not implemented. the old styles still raise AttributeErrors (both styles
now match CPython). we assume __len/int/long/float/complex__ raise
AttributeErrors in many places so leaving them alone for now
........
r5054 | zyasoft | 2008-08-02 17:35:23 -0400 (Sat, 02 Aug 2008) | 3 lines

Raise a ValueError if attempt to set thread.stack_size > 0 but <
32768. Fixes test_thread.
........
r5055 | zyasoft | 2008-08-02 17:40:59 -0400 (Sat, 02 Aug 2008) | 4 lines

threading does not maintain a dict of limbo threads, since this
doesn't make sense in Jython. Remove code testing cleanup of
threading._limbo in test_support.threading_setup/thread_cleanup
........
r5056 | zyasoft | 2008-08-02 17:43:22 -0400 (Sat, 02 Aug 2008) | 5 lines

Fixed roundtripping so that eval(str(a)) == a where a uses a typecode
of a Java object, along with other similar minor issues. Rewrote
test_array_jy to use unittest, and removed testing now being done by
standard 2.5 test_array
........
r5057 | pjenvey | 2008-08-02 22:13:36 -0400 (Sat, 02 Aug 2008) | 3 lines

o pre and xreadlines were removed in CPython 2.4
o include some other modules I hope actually work
........
r5058 | zyasoft | 2008-08-02 23:09:49 -0400 (Sat, 02 Aug 2008) | 4 lines

Unicode arrays -- array.array('u') -- now work as expected, by special
casing the fact that we are storing them as an array of codepoints
(ints). Because of this, byteswap doesn't make sense.
........
r5059 | nriley | 2008-08-03 00:33:38 -0400 (Sun, 03 Aug 2008) | 1 line

Remove workaround for issue 1059 (was fixed in r4713).
........
r5060 | nriley | 2008-08-03 00:45:30 -0400 (Sun, 03 Aug 2008) | 1 line

jna-posix from http://svn.codehaus.org/jruby-contrib/trunk/jna-posix r213 + patch in http://jira.codehaus.org/browse/JRUBY-2885 for isatty on Windows + patch in http://jira.codehaus.org/browse/JRUBY-2887 for atime setting
........
r5061 | zyasoft | 2008-08-03 00:45:39 -0400 (Sun, 03 Aug 2008) | 5 lines

Allow typecodes of B, H, I to fully use their storage. Don't run
test_byteswap for u (Unicode) because it doesn't make sense given
UTF-16 decoding.
Fixes test_array.
........
r5062 | nriley | 2008-08-03 00:47:37 -0400 (Sun, 03 Aug 2008) | 1 line

Support setting atime from os.utime.
........
r5063 | nriley | 2008-08-03 00:49:27 -0400 (Sun, 03 Aug 2008) | 1 line

Don't create file opened with r+; better validate file modes (fixes test_mailbox).
........
r5064 | zyasoft | 2008-08-03 01:11:47 -0400 (Sun, 03 Aug 2008) | 2 lines

Basic cleanup so that test_socket is runnable under current testing.
........
r5065 | nriley | 2008-08-03 03:42:24 -0400 (Sun, 03 Aug 2008) | 1 line

Fix mistaken assumption that 'r+' will create a file.
........
r5066 | nriley | 2008-08-03 05:50:01 -0400 (Sun, 03 Aug 2008) | 1 line

jna-posix from http://svn.codehaus.org/jruby-contrib/trunk/jna-posix r213 + patch in http://jira.codehaus.org/browse/JRUBY-2885 for isatty on Windows + patch in http://jira.codehaus.org/browse/JRUBY-2887 for atime setting (now with Windows support)
........
r5067 | fwierzbicki | 2008-08-03 10:42:44 -0400 (Sun, 03 Aug 2008) | 7 lines

Passed filename along to parsers and PythonTokenSource. created
a PyIndentationError to allow me derive IndentationErrors from
ParseException. Added a "type" to ParseException to recognize
when an IndentationError is needed. Changed test_traceback slightly
because I believe Antlr comes up with a better placement for the
carot vs CPython (at the indent and not end of line for indent errors.
........
r5068 | fwierzbicki | 2008-08-03 12:30:08 -0400 (Sun, 03 Aug 2008) | 5 lines

Small change for Jython -- it will be too hard in the short term to use
CPython's "Store" where Jython uses "AugStore" for AugAssigns. Eventually I do
intend to fix this, but it will take a little more work in the compiler than I
want to do just now.
........
r5069 | pjenvey | 2008-08-03 14:14:53 -0400 (Sun, 03 Aug 2008) | 1 line

expand tabs
........
r5070 | pjenvey | 2008-08-03 14:18:34 -0400 (Sun, 03 Aug 2008) | 2 lines

o FileIO whitespace, small refactor
o re-enable test_ast
........
r5071 | fwierzbicki | 2008-08-03 16:17:40 -0400 (Sun, 03 Aug 2008) | 2 lines

Partial fix for #1092 (just a comment no newline).
........
r5072 | fwierzbicki | 2008-08-03 17:45:49 -0400 (Sun, 03 Aug 2008) | 2 lines

whitespace
........
r5073 | fwierzbicki | 2008-08-03 18:54:14 -0400 (Sun, 03 Aug 2008) | 3 lines

CommonToken now appears to cover the cases that ClassicToken was left around
to cover. Removed all ClassicToken.
........
r5074 | fwierzbicki | 2008-08-03 21:15:48 -0400 (Sun, 03 Aug 2008) | 2 lines

Fixes to handle parsing with files that have no newline at the end.
........
r5075 | fwierzbicki | 2008-08-03 22:36:46 -0400 (Sun, 03 Aug 2008) | 2 lines

Some checks for assignment to None.
........
r5076 | nriley | 2008-08-03 23:23:57 -0400 (Sun, 03 Aug 2008) | 1 line

Truncate files opened with the 'w+' mode; some FileIO cleanup. Re-fixes test_fileno.
........
r5077 | fwierzbicki | 2008-08-04 13:57:34 -0400 (Mon, 04 Aug 2008) | 15 lines

In test_decorators.py, test_eval_order cleverly instruments a decorated call so
the order of evaluation and calls can be determined. It also checks to see if
the decoration works exactly the same when translated to nested function calls.
Jython was evaluating the arguments before the function name, the opposite of
CPython. This patch makes the order correct for both decorators and normal
function calls. Also, PyObject.invoke states in its comments that it is for use
from Java and shows the equivalent

o.__getattr__(name).__call__(args, keywords)

which is how I am constructing a call in the bytecode. I also thought that
"Invoke" with a capitalized name was non-ideal, so I changed the name to
invokeNoKeywords, which also allowed me to remove a comment that conveyed the
same meaning.
........
r5078 | pjenvey | 2008-08-04 17:34:24 -0400 (Mon, 04 Aug 2008) | 1 line

gather more info about the flakey test_hash OutOfMemoryError
........
r5079 | fwierzbicki | 2008-08-05 12:55:18 -0400 (Tue, 05 Aug 2008) | 4 lines

Reworked lexing of newlines and EOF. Broke up PythonTokenSource into more
useful methods. Removed inSingle from Python.g and added it to
PythonTokenSource.
........
r5080 | fwierzbicki | 2008-08-05 13:16:19 -0400 (Tue, 05 Aug 2008) | 2 lines

Guard against null text in PySyntaxError.
........
r5081 | fwierzbicki | 2008-08-05 16:21:24 -0400 (Tue, 05 Aug 2008) | 3 lines

Commenting out the strange eval parsing errors - I'm working elsewhere at the
moment and having test_jy_compile as a coal mine canary will be very helpful.
........
r5082 | fwierzbicki | 2008-08-05 18:39:04 -0400 (Tue, 05 Aug 2008) | 3 lines

Fix test_pkgimport for Jython (was deleting .pyc and .pyo files, now
deleted $py.class files on Jython).
........
r5083 | fwierzbicki | 2008-08-05 18:40:19 -0400 (Tue, 05 Aug 2008) | 2 lines

Forgot our convention.
........
r5084 | fwierzbicki | 2008-08-05 18:45:16 -0400 (Tue, 05 Aug 2008) | 3 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_pkgimport.py@51333
........
r5085 | fwierzbicki | 2008-08-05 18:51:03 -0400 (Tue, 05 Aug 2008) | 3 lines

Fix test_pkgimport for Jython. Now deleting $py.class files on Jython, and
using __builtin__ instead of the CPython implementatin specific __buitlins__.
........
r5086 | fwierzbicki | 2008-08-05 19:44:49 -0400 (Tue, 05 Aug 2008) | 2 lines

2 new passing tests.
........
r5087 | fwierzbicki | 2008-08-05 19:46:30 -0400 (Tue, 05 Aug 2008) | 2 lines

Finish out "IN" as a token.
........
r5088 | fwierzbicki | 2008-08-05 22:19:54 -0400 (Tue, 05 Aug 2008) | 2 lines

fix for Java integration ("in" and "class" can now be used as non-keywords).
........
r5089 | fwierzbicki | 2008-08-05 22:27:45 -0400 (Tue, 05 Aug 2008) | 2 lines

Extra comment as a reminder about keywords.
........
r5090 | pjenvey | 2008-08-05 22:48:46 -0400 (Tue, 05 Aug 2008) | 4 lines

PEP 305 (csv module)
patches from Paul Drummond, Dave Kuhlman (thanks!)
fixes #1650802
........
r5091 | fwierzbicki | 2008-08-06 07:59:35 -0400 (Wed, 06 Aug 2008) | 4 lines

call clean if grammars change. Really only needed when *tokens* change,
but token changes break things so subtly, and so few work on the grammars
anyway, that this is the way to go I think.
........
r5092 | fwierzbicki | 2008-08-06 08:08:16 -0400 (Wed, 06 Aug 2008) | 2 lines

trigger jarjar if grammar changes (since clean is triggered).
........
r5093 | fwierzbicki | 2008-08-06 09:05:15 -0400 (Wed, 06 Aug 2008) | 8 lines

Support "in" as an attribute for Java compatibility, but not as a
NAME since this is overkill and causes errors (for example:

for x, in ((1,),(2,)):

parses "x, in" as a tuple (x,in) if "in" is treated like a name. bad.
........
r5094 | fwierzbicki | 2008-08-06 13:54:07 -0400 (Wed, 06 Aug 2008) | 10 lines

Changed grammars to allow most keywords to be used as attributes for Java
compatibility. Also added some tests. Removed a couple of tests that use
keywords as function names as this is no longer permitted. Allowing keywords to
be used as arbitrary NAMEs causes very weird grammar issues. For example

for x, in ((1,),(2,)):

matched "x, in" as a tuple. Now the permitted use is limited to things like
foo.in()
........
r5095 | fwierzbicki | 2008-08-06 14:15:17 -0400 (Wed, 06 Aug 2008) | 4 lines

improved jarjar conditional -- but made build.xml slightly more impenetrable.
Sigh.
........
r5096 | pjenvey | 2008-08-06 14:45:50 -0400 (Wed, 06 Aug 2008) | 3 lines

fix zipimport not handling mismatched APIVersion, and imp.load_module not
handling it either for packages
........
r5097 | fwierzbicki | 2008-08-06 15:05:52 -0400 (Wed, 06 Aug 2008) | 2 lines

More thorough keyword testing.
........
r5098 | fwierzbicki | 2008-08-06 15:35:41 -0400 (Wed, 06 Aug 2008) | 10 lines

Line and column position for

from __future__ import braces

Also fixed

from __future__ import GIL

Good thing there are unit tests for this :)
........
r5099 | fwierzbicki | 2008-08-06 19:51:03 -0400 (Wed, 06 Aug 2008) | 4 lines

Putting in fake SystemRandom that throws NotImplementedError on
all accesses so the rest of test_random.py can run (and fail for
better reasons).
........
r5100 | leosoto | 2008-08-06 22:27:08 -0400 (Wed, 06 Aug 2008) | 1 line

Experimental interpreter restarting support, aimed at webapp dev-server 'reloading', without having to resort to process forking. There is more info here: <http://codereview.appspot.com/2810>, where I posted the first attempt of this patch
........
r5102 | nriley | 2008-08-07 13:11:16 -0400 (Thu, 07 Aug 2008) | 1 line

Don't attempt to test AWT if we have no display.
........
r5103 | zyasoft | 2008-08-07 17:07:09 -0400 (Thu, 07 Aug 2008) | 3 lines

From
http://svn.python.org/projects/python/branches/release25-maint/Lib/pkgutil.py@65582
........
r5104 | zyasoft | 2008-08-07 17:18:44 -0400 (Thu, 07 Aug 2008) | 9 lines

Replaces pkgutil.read_code with read_jython_code (diff args); this
depends on the actual bytecode storage, so we have to have something
specific here. get_loader special cases for the sys module, since
Jython doesn't actually implement this as a module (and it appears to
be a non-trivial decoupling to make it so with the current
PySystemState).

This fixes test_runpy.
........
r5105 | zyasoft | 2008-08-08 11:37:43 -0400 (Fri, 08 Aug 2008) | 7 lines

With statements now keep their own stack of exit functions in PyFrame
so that they don't get stepped on by generators save/restore of local
temporary registers. Fixes most problems in test_with, including JVM
verification errors. (Remaining issues seem to be with generator exit
issues, to be resolved separately.)
........
r5106 | fwierzbicki | 2008-08-08 13:42:06 -0400 (Fri, 08 Aug 2008) | 3 lines

make jarjar run all of the time until I figure out why the conditionals
are not working.
........
r5107 | fwierzbicki | 2008-08-08 13:56:53 -0400 (Fri, 08 Aug 2008) | 3 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_codeop.py@51333
........
r5108 | fwierzbicki | 2008-08-08 14:24:44 -0400 (Fri, 08 Aug 2008) | 10 lines

Various fixes to improve Jython compatibility with CPython, especially when
passing PyCF_DONT_IMPLY_DEDENT. Added PyCF_DONT_IMPLY_DEDENT to codeop.py,
PyTableCode.java, and CompilerFlags.java to help test_codeop.py run. Began to
integrate tests from test_jy_compile.py, which appears to have started as a copy
of test_codeop.py, into test_codeop.py. Once test_jy_compile is fully
integrated into test_codeop.py I should be able to delete test_jy_compile. Also
added eval support to PythonPartial.g, since test_codeop exercises that
functionality.
........
r5109 | fwierzbicki | 2008-08-08 15:24:38 -0400 (Fri, 08 Aug 2008) | 2 lines

Remove debug string.
........
r5110 | fwierzbicki | 2008-08-08 15:30:38 -0400 (Fri, 08 Aug 2008) | 5 lines

Extracted the actions from the grammars into a single GrammarActions
file. This will help if I ever get far enough on integrating the two
grammars into one. Also started an ErrorStmt node to handle bogus
statements in the walker.
........
r5111 | pjenvey | 2008-08-08 18:07:24 -0400 (Fri, 08 Aug 2008) | 5 lines

o another fix to make 'currentThread() is currentThread()' from
thread.start_new_thread threads. use a WeakHashMap because weakrefs of
javainstances ref the wrapper, not the actual java object
o custom toString for FunctionThread
........
r5112 | fwierzbicki | 2008-08-08 19:54:06 -0400 (Fri, 08 Aug 2008) | 2 lines

New error nodes for tool ASTs. Added errorhandling methods to ErrorHandler.
........
r5113 | pjenvey | 2008-08-08 20:37:11 -0400 (Fri, 08 Aug 2008) | 1 line

allow subclasses in getPyObjectByType
........
r5114 | zyasoft | 2008-08-09 01:36:17 -0400 (Sat, 09 Aug 2008) | 3 lines

Reordered code to test for raising exceptions thrown into a generator,
fixing test cases in test_contextlib, test_generators, and test_with.
........
r5115 | pjenvey | 2008-08-09 02:12:17 -0400 (Sat, 09 Aug 2008) | 2 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/list_tests.py@52138
........
r5116 | pjenvey | 2008-08-09 02:17:43 -0400 (Sat, 09 Aug 2008) | 4 lines

o re-integrate list_tests workaround and use stock 2.5 seq_tests
o fix list.index handling of long start values and slice assignment of strings
o raise MemoryErrors as per seq_tests.test_bigrepeat
........
r5117 | pjenvey | 2008-08-09 02:18:29 -0400 (Sat, 09 Aug 2008) | 1 line

fix test_weakref, re-enable 2.5isms
........
r5119 | fwierzbicki | 2008-08-09 09:52:14 -0400 (Sat, 09 Aug 2008) | 3 lines

debug() is not really all that necessary now that these have been extracted from
the .g files.
........
r5120 | fwierzbicki | 2008-08-09 10:00:30 -0400 (Sat, 09 Aug 2008) | 2 lines

Surely a typo?
........
r5121 | fwierzbicki | 2008-08-09 10:25:25 -0400 (Sat, 09 Aug 2008) | 3 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/test_genexps.py@51333
........
r5122 | fwierzbicki | 2008-08-09 12:27:42 -0400 (Sat, 09 Aug 2008) | 4 lines

added #doctest: +IGNORE_EXCEPTION_DETAIL for SyntaxErrors that
just have "invalid syntax" as the detail, since this is the
default message and Antlr actually gives more information.
........
r5123 | fwierzbicki | 2008-08-09 13:07:19 -0400 (Sat, 09 Aug 2008) | 2 lines

Don't allow generator expressions on left side of assignment.
........
r5124 | fwierzbicki | 2008-08-09 13:35:06 -0400 (Sat, 09 Aug 2008) | 2 lines

Call __iter__ on the iter part of gen expressions.
........
r5125 | pjenvey | 2008-08-09 17:53:44 -0400 (Sat, 09 Aug 2008) | 3 lines

fix remove() not fully zeroing out the end of the array, causing a small memory
leak. fixes test_deque.test_weakref
........
r5126 | pjenvey | 2008-08-09 18:02:39 -0400 (Sat, 09 Aug 2008) | 2 lines

reprStack isn't exposed anywhere, simplify it as a Stack
........
r5127 | pjenvey | 2008-08-09 18:05:00 -0400 (Sat, 09 Aug 2008) | 2 lines

typo/thinko
pointed out by lsoto
........
r5128 | pjenvey | 2008-08-09 19:02:50 -0400 (Sat, 09 Aug 2008) | 1 line

oops, Stack won't work because we need to compare via identity. revert r5126
........
r5129 | nriley | 2008-08-09 19:40:43 -0400 (Sat, 09 Aug 2008) | 1 line

Don't try tests expecting MemoryError which can pass with large Java heaps.
........
r5130 | fwierzbicki | 2008-08-10 11:03:21 -0400 (Sun, 10 Aug 2008) | 3 lines

Ignoring other SyntaxError details for now -- though I probably want to
re-enable these two later.
........
r5131 | fwierzbicki | 2008-08-10 11:31:52 -0400 (Sun, 10 Aug 2008) | 4 lines

Choose a name for the bound parameter to generator expressions that can't
collide with a real NAME. Arbitrarily picked _(x) just because it looks a bit
like the names of the gen expressions themselves.
........
r5132 | zyasoft | 2008-08-10 15:52:51 -0400 (Sun, 10 Aug 2008) | 5 lines

After a more careful reading of PEP 342: need to emit bytecode for
getGeneratorInput and possible raising exception before any yields;
and after each yield point.
This one-line change fixes test_contextlib
........
r5133 | leosoto | 2008-08-10 16:50:55 -0400 (Sun, 10 Aug 2008) | 3 lines

Added coercion to the cmp() algorithm, which is also used for comparison operators in the absence of rich cmp methods (See #1031). In the process, __coerce_ex__ was added to deriveds, and a slight refactor was made on types exposing __coerce__ to avoid using the __coerce_ex__ virtual method on the exposed methods code.

Fixes test_coercion.
........
r5134 | leosoto | 2008-08-10 17:09:42 -0400 (Sun, 10 Aug 2008) | 1 line

Added test_cmp_jy, with should have tests to ensure CPython compatibility on cmp(). Currently, it only contains test cases for the problems exposed on http://bugs.jython.org/issue1889394
........
r5135 | leosoto | 2008-08-10 17:14:36 -0400 (Sun, 10 Aug 2008) | 1 line

Added cmp test for http://bugs.jython.org/issue1869347
........
r5136 | leosoto | 2008-08-10 17:19:00 -0400 (Sun, 10 Aug 2008) | 1 line

Added test for http://bugs.jython.org/issue1804011 (derived_dict() must be equal to {})
........
r5137 | pjenvey | 2008-08-10 17:30:15 -0400 (Sun, 10 Aug 2008) | 3 lines

o __float__ should always return a PyFloat, no need to cast in complex new
o prefer PyObject.isCallable to __builtin__'s
........
r5138 | zyasoft | 2008-08-10 18:57:07 -0400 (Sun, 10 Aug 2008) | 3 lines

Properly closes a generator upon close/throw including throwing
StopIteration upon a send to a closed generator.
........
r5139 | zyasoft | 2008-08-10 19:14:13 -0400 (Sun, 10 Aug 2008) | 2 lines

Import test_support from test
........
r5140 | fwierzbicki | 2008-08-10 20:32:24 -0400 (Sun, 10 Aug 2008) | 3 lines

Can't assign to Num and better placement for checkAssign, so
lambdef and fplist also get checks for args.
........
r5141 | zyasoft | 2008-08-10 22:56:53 -0400 (Sun, 10 Aug 2008) | 3 lines

from X import * now raises the appropriate syntax errors (for nested
scopes) or warning (otherwise).
........
r5142 | zyasoft | 2008-08-10 23:17:44 -0400 (Sun, 10 Aug 2008) | 5 lines

Commented out inappropriate doctest in test_genexps that tested for
the reuse of memory objects. This obviously doesn't apply to Jython
and other nonrefcounted implementations.
Fixes test_genexps
........
r5143 | zyasoft | 2008-08-11 00:09:35 -0400 (Mon, 11 Aug 2008) | 3 lines

From
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_generators.py@65635
........
r5144 | leosoto | 2008-08-11 01:35:18 -0400 (Mon, 11 Aug 2008) | 1 line

Really fix coercion on derived this time
........
r5145 | fwierzbicki | 2008-08-11 01:37:15 -0400 (Mon, 11 Aug 2008) | 2 lines

Properly handle tuples in subscripts (last test_grammar problem!)
........
r5146 | zyasoft | 2008-08-11 01:52:43 -0400 (Mon, 11 Aug 2008) | 4 lines

Another corner case for generators: .throw on a just-started generator
should immediately raise an exception in the current context. Also
fixed buggy tests in test_generators.
........
r5147 | pjenvey | 2008-08-11 02:59:28 -0400 (Mon, 11 Aug 2008) | 4 lines

o change PySlice.indices to work like CPython's PySlice_GetIndicesEx (include a
slicelength value)
o reject PySequence slice assignments with a step and mismatching slicelengths
........
r5148 | pjenvey | 2008-08-11 03:03:25 -0400 (Mon, 11 Aug 2008) | 1 line

cleanup/coding standards
........
r5149 | fwierzbicki | 2008-08-11 08:44:54 -0400 (Mon, 11 Aug 2008) | 2 lines

Switch to for-each syntax in walker.
........
r5150 | fwierzbicki | 2008-08-11 12:43:44 -0400 (Mon, 11 Aug 2008) | 2 lines

Added 3 arg and 4 arg __call__
........
r5151 | fwierzbicki | 2008-08-11 14:31:41 -0400 (Mon, 11 Aug 2008) | 3 lines

remove LeadingSpaceSkippingStream and replace with LEADING_WS in Python.g.
No idea what I was thinking with LeadingSpaceSkippingStream...
........
r5152 | fwierzbicki | 2008-08-11 15:25:46 -0400 (Mon, 11 Aug 2008) | 2 lines

Tightened up eval and uncommented tests in test_jy_compile.
........
r5153 | fwierzbicki | 2008-08-11 15:48:31 -0400 (Mon, 11 Aug 2008) | 2 lines

Moved buffered stream creation from many places to one place.
........
r5154 | pjenvey | 2008-08-11 18:24:20 -0400 (Mon, 11 Aug 2008) | 3 lines

make the stepless slice ops more consistent with the compiler by passing null
for step instead of Py.One, fixes test_operator.test_setslice
........
r5155 | leosoto | 2008-08-11 19:27:27 -0400 (Mon, 11 Aug 2008) | 1 line

__findattr__ refactor. The overridable version is not called __findattr_ex__ (in the abscense of a better name), and can return null or raise AttributeError, whatever is more efficient for the implementing code. This makes *Derived getattr() more correct and fixes #1041 and #1095.
........
r5156 | fwierzbicki | 2008-08-12 13:17:12 -0400 (Tue, 12 Aug 2008) | 4 lines

Implement BOM handling as much as CPython 2.5 does. Altered
bad_coding2.py because Java, unlike CPython accepts "utf8" as
a synonymn for "utf-8"
........
r5157 | fwierzbicki | 2008-08-12 14:11:38 -0400 (Tue, 12 Aug 2008) | 2 lines

Disallow encodings from unicode input. Fixes one test in test_compile.py.
........
r5158 | fwierzbicki | 2008-08-12 14:17:00 -0400 (Tue, 12 Aug 2008) | 3 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_compile.py@65007
........
r5159 | fwierzbicki | 2008-08-12 14:24:24 -0400 (Tue, 12 Aug 2008) | 2 lines

Don't try really long args on Jython for now.
........
r5160 | fwierzbicki | 2008-08-12 14:40:38 -0400 (Tue, 12 Aug 2008) | 2 lines

Trailing commas not allowed outside of parens.
........
r5161 | fwierzbicki | 2008-08-12 14:57:22 -0400 (Tue, 12 Aug 2008) | 3 lines

Since the comment for test_leading_newlines suggests that this
may be CPython implementation specific, skip.
........
r5162 | fwierzbicki | 2008-08-12 15:47:24 -0400 (Tue, 12 Aug 2008) | 2 lines

fixes, workarounds, and a couple of punts for test_compile.py.
........
r5163 | leosoto | 2008-08-12 18:58:47 -0400 (Tue, 12 Aug 2008) | 1 line

subprocess.Popen now "inherits" os.environ if env is not explictely set. Fixes #1104
........
r5164 | fwierzbicki | 2008-08-12 20:30:38 -0400 (Tue, 12 Aug 2008) | 4 lines

Before this check-in there was an AST inconsistency with CPython. To pick one
example, CPython's AST for --1 is (USub(-1)) and Jython is (Usub(Usub(1)). This
fixes that inconsistency.
........
r5165 | fwierzbicki | 2008-08-12 20:53:09 -0400 (Tue, 12 Aug 2008) | 4 lines

Avoiding a test for constant folding -- I think it is an implementation detail
though I think it wouldn't be too hard to fix in the grammar. It's just not a
top priority.
........
r5166 | fwierzbicki | 2008-08-12 23:48:00 -0400 (Tue, 12 Aug 2008) | 2 lines

Remove direct exception throwing for tooling.
........
r5168 | fwierzbicki | 2008-08-13 10:52:12 -0400 (Wed, 13 Aug 2008) | 2 lines

Jython can handle atof with hex floats and large unicode values.
........
r5169 | fwierzbicki | 2008-08-13 12:21:18 -0400 (Wed, 13 Aug 2008) | 2 lines

remove comment that is no longer True, thanks to Leo Soto for pointing this out.
........
r5170 | fwierzbicki | 2008-08-13 15:25:13 -0400 (Wed, 13 Aug 2008) | 3 lines

from:
http://svn.python.org/projects/python/branches/release25-maint/Lib/test/test_syntax.py@54954
........
r5171 | fwierzbicki | 2008-08-13 15:48:08 -0400 (Wed, 13 Aug 2008) | 2 lines

Make errtext optional in test_syntax and fixed illegal delete of function call.
........
r5172 | pjenvey | 2008-08-13 19:35:12 -0400 (Wed, 13 Aug 2008) | 3 lines

fix alignment of little and big endian long longs. caused extra padding which
broke test_zipfile zip64 tests
........
r5173 | fwierzbicki | 2008-08-13 20:07:31 -0400 (Wed, 13 Aug 2008) | 2 lines

Add EOF check for file_input.
........
r5174 | fwierzbicki | 2008-08-13 20:35:01 -0400 (Wed, 13 Aug 2008) | 2 lines

antlr 3.1 final.
........
r5175 | fwierzbicki | 2008-08-13 22:00:01 -0400 (Wed, 13 Aug 2008) | 2 lines

Add EOF to single_input.
........
r5176 | fwierzbicki | 2008-08-14 13:51:34 -0400 (Thu, 14 Aug 2008) | 2 lines

Updated comment.
........
r5177 | fwierzbicki | 2008-08-14 14:48:37 -0400 (Thu, 14 Aug 2008) | 4 lines

Ignoring exception message details. This actually allows more of this test to
run (the doctests don't run if any of the regular tests fail). This reveals
more legitimate failures than before.
........
r5178 | fwierzbicki | 2008-08-15 07:43:53 -0400 (Fri, 15 Aug 2008) | 2 lines

Use same lexer and token source for parsing and partial parsing.
........
r5179 | fwierzbicki | 2008-08-15 13:55:20 -0400 (Fri, 15 Aug 2008) | 2 lines

small name change.
........
r5180 | fwierzbicki | 2008-08-15 14:37:14 -0400 (Fri, 15 Aug 2008) | 3 lines

don't allow assignment to yields. Fixes 2 doctests in test_generator as long as we
ignore the exact error message.
........
r5181 | fwierzbicki | 2008-08-15 15:44:57 -0400 (Fri, 15 Aug 2008) | 2 lines

Refactor: put repeated code into one method.
........
r5182 | leosoto | 2008-08-15 16:15:50 -0400 (Fri, 15 Aug 2008) | 1 line

A start for unicodedata. Only implements normalize()
........
r5183 | leosoto | 2008-08-15 19:03:17 -0400 (Fri, 15 Aug 2008) | 1 line

Reverted r5182, as the implementation of unicodedata.normalize required Java 1.6. Sorry!
........
r5184 | zyasoft | 2008-08-16 01:35:48 -0400 (Sat, 16 Aug 2008) | 18 lines

Fixes for and in test_builtin:

* test_builtin.test_cmp because we let Jython compare objects with cycles
in them, unlike CPython

* test_builtin.test_general_eval - Jython can use arbitrary maps for
globals; this was discussed for CPython. Not in current 2.6,
perhaps because it touched a lot of code or issues with performance
(see http://bugs.python.org/issue1402289)

* __builtin__#eval/compile - Now validate that locals or globals are
in fact mapping objects, as well as ensure that locals().keys() is
a PyList (to avoid throwing a CheckCastException)

* Py#compile_flags - if passed in a String to compile, ensure it does
not contain a null byte
........
r5185 | nriley | 2008-08-16 02:09:33 -0400 (Sat, 16 Aug 2008) | 1 line

Permit removal/modification of attributes on PySystemState; add __name__.
........
r5186 | nriley | 2008-08-16 02:18:30 -0400 (Sat, 16 Aug 2008) | 1 line

Update Eclipse classpath to Antlr 3.1 final.
........
r5187 | zyasoft | 2008-08-16 02:22:40 -0400 (Sat, 16 Aug 2008) | 5 lines

input and raw_input now appropriately raise a RuntimeError if
sys.stdin/stdout are deleted (via the facade of
PySystemState.PyAttributeDeleted)
........
r5188 | fwierzbicki | 2008-08-16 09:56:28 -0400 (Sat, 16 Aug 2008) | 2 lines

Method not used.
........
r5189 | fwierzbicki | 2008-08-16 15:31:55 -0400 (Sat, 16 Aug 2008) | 2 lines

Shouldn't be discarding hidden tokens.
........
r5190 | fwierzbicki | 2008-08-16 18:46:04 -0400 (Sat, 16 Aug 2008) | 2 lines

Fix comment/newline parsing problems in interactive (single) parsing.
........
r5191 | fwierzbicki | 2008-08-16 20:02:42 -0400 (Sat, 16 Aug 2008) | 3 lines

check for Py.None in runCode as well as null for locals
and globals. Fixes test_decorators.py
........
r5192 | nriley | 2008-08-16 22:23:19 -0400 (Sat, 16 Aug 2008) | 1 line

Option to enable bytecode verifier.
........
r5193 | nriley | 2008-08-17 11:24:17 -0400 (Sun, 17 Aug 2008) | 1 line

Better comment for PySystemState method hack: depends on "functions" in __dict__ appearing as methods.
........
r5194 | leosoto | 2008-08-17 22:12:34 -0400 (Sun, 17 Aug 2008) | 1 line

PythonInterpreter#cflags initialization moved to PythonInterpreter itself. Fixes a crash on modjy
........
r5195 | leosoto | 2008-08-17 22:39:57 -0400 (Sun, 17 Aug 2008) | 1 line

PyArray#array_fromunicode: Worked around the issue described on #1105 (wrong exposed bytecode when receiving a PyUnicode argument on a exposed method), by receiving a PyObject and doing the type-check manually
........
r5196 | zyasoft | 2008-08-18 03:17:46 -0400 (Mon, 18 Aug 2008) | 7 lines

Strings are UTF-8 encoded in the runtime constant pool, meaning that
naively (and that is a reasonable strategy here), we need up to 4
bytes per codepoint. So allocate string constants up to 16000 bytes
and combine as necessary. This resolves one Pygments problem (and the
one we can directly do anything about, the other is isolated
surrogates in unistring.Cs).
........
r5197 | nriley | 2008-08-18 10:46:22 -0400 (Mon, 18 Aug 2008) | 1 line

Run Jython regrtest (on unix) with bytecode verification.
........

fwierzbicki@79b11f02-f210-0410-99aa-fb189b3d1ac4 fwierzbicki@79b11f02-f210-0410-99aa-fb189b3d1ac4 2008-08-19

1 2 3 .. 21 > >> (Page 1 of 21)
changed .hgsvnexternals
changed jython/.classpath
changed jython/CPythonLib.includes
changed jython/CoreExposed.includes
changed jython/Lib/codeop.py
changed jython/Lib/datetime.py
changed jython/Lib/distutils/ccompiler.py
changed jython/Lib/distutils/command/bdist.py
changed jython/Lib/distutils/command/bdist_dumb.py
changed jython/Lib/distutils/command/install.py
changed jython/Lib/distutils/command/sdist.py
changed jython/Lib/distutils/file_util.py
changed jython/Lib/distutils/spawn.py
changed jython/Lib/distutils/sysconfig.py
changed jython/Lib/distutils/util.py
changed jython/Lib/doctest.py
changed jython/Lib/fileinput.py
changed jython/Lib/gzip.py
changed jython/Lib/inspect.py
changed jython/Lib/ntpath.py
changed jython/Lib/os.py
changed jython/Lib/popen2.py
changed jython/Lib/posixpath.py
changed jython/Lib/site.py
changed jython/Lib/socket.py
.hgsvnexternals Diff Switch to side-by-side view
Loading...
jython/.classpath Diff Switch to side-by-side view
Loading...
jython/CPythonLib.includes Diff Switch to side-by-side view
Loading...
jython/CoreExposed.includes Diff Switch to side-by-side view
Loading...
jython/Lib/codeop.py Diff Switch to side-by-side view
Loading...
jython/Lib/datetime.py Diff Switch to side-by-side view
Loading...
jython/Lib/distutils/ccompiler.py Diff Switch to side-by-side view
Loading...
jython/Lib/distutils/command/bdist.py Diff Switch to side-by-side view
Loading...
jython/Lib/distutils/command/bdist_dumb.py Diff Switch to side-by-side view
Loading...
jython/Lib/distutils/command/install.py Diff Switch to side-by-side view
Loading...
jython/Lib/distutils/command/sdist.py Diff Switch to side-by-side view
Loading...
jython/Lib/distutils/file_util.py Diff Switch to side-by-side view
Loading...
jython/Lib/distutils/spawn.py Diff Switch to side-by-side view
Loading...
jython/Lib/distutils/sysconfig.py Diff Switch to side-by-side view
Loading...
jython/Lib/distutils/util.py Diff Switch to side-by-side view
Loading...
jython/Lib/doctest.py Diff Switch to side-by-side view
Loading...
jython/Lib/fileinput.py Diff Switch to side-by-side view
Loading...
jython/Lib/gzip.py Diff Switch to side-by-side view
Loading...
jython/Lib/inspect.py Diff Switch to side-by-side view
Loading...
jython/Lib/ntpath.py Diff Switch to side-by-side view
Loading...
jython/Lib/os.py Diff Switch to side-by-side view
Loading...
jython/Lib/popen2.py Diff Switch to side-by-side view
Loading...
jython/Lib/posixpath.py Diff Switch to side-by-side view
Loading...
jython/Lib/site.py Diff Switch to side-by-side view
Loading...
jython/Lib/socket.py Diff Switch to side-by-side view
Loading...
1 2 3 .. 21 > >> (Page 1 of 21)

Get latest updates about Open Source Projects, Conferences and News.

Sign up for the SourceForge newsletter:





No, thanks