Fix some logic in PythonCmd() -- if the AsString() call failed, we'd
be calling LEAVE_PYTHON twice (once in PythonCmd_Error() and once in
PythonCmd() itself).
This fix doesn't fix the IDLE problems, but at least makes them more
debuggable by getting rid of the segfault.
Fix two bugs in socket_getaddr():
(a) 'single' shouldn't be DECREF'ed in the cleanup code;
(b) the fallback case in makesockaddr() should use y# instead of s#
in the format, since the data is bytes, not UTF-8 text.
Oops. The PyObject_Print() function was totally broken; the original code
was relying on PyString.tp_print but that no longer works.
Fortunately it's rarely called; only the gdb 'pyo' command seems affected.
Tests for dict comprehensions I forgot to 'svn add' when committing the
actual feature.
I was _sure_ I had checked in these new tests already, but I guess the
checkin failed and I didn't notice.
................
r56847 | guido.van.rossum | 2007-08-09 07:03:34 -0700 (Thu, 09 Aug 2007) | 30 lines
Merged revisions 56760-56846 via svnmerge from
svn+ssh://pythondev@svn.python.org/python/trunk
SF patch# 1770008 by Christian Heimes (plus some extras).
Completely get rid of StringIO.py and cStringIO.c.
I had to fix a few tests and modules beyond what Christian did, and
invent a few conventions. E.g. in elementtree, I chose to
write/return Unicode strings whe no encoding is given, but bytes when
an explicit encoding is given. Also mimetools was made to always
assume binary files.
Thomas Heller [Wed, 8 Aug 2007 18:47:32 +0000 (18:47 +0000)]
Fix the ctypes tests. Patch from Victor Stinner. He writes:
The problem is that ctypes c_char (and c_char_p) creates unicode string
instead of byte string. I attached a proposition (patch) to change this
behaviour (use bytes for c_char).
So in next example, it will display 'bytes' and not 'str':
from ctypes import c_buffer, c_char
buf = c_buffer("abcdef")
print (type(buf[0]))
Other behaviour changes:
- repr(c_char) adds a "b"
eg. repr(c_char('x')) is "c_char(b'x')" instead of "c_char('x')"
- bytes is mutable whereas str is not:
this may break some modules based on ctypes
Jeremy Hylton [Wed, 8 Aug 2007 17:36:33 +0000 (17:36 +0000)]
Fix several failing tests in test_urllib2net.
The HTTPResponse object is being passed to BufferedReader, but it
wasn't designed to be used that way. These changes extend the hacks
that have already been made in urllib2 to get the tests to pass.
The hacks need to be removed and proper support for use with the io
library. That's a project for another day.
Fix the last remaining test_csv failure.
We were using T_CHAR for a UNICODE character.
(This happened to work on x86 most of the time due to endianness;
but not on PPC.)
Remove the coding line, and convert umlauted characters to \x escapes.
All code in the stdlib should be ASCII only, unless (a) it's test
input for testing the *parsing* of unicode, or (b) it's an author's
name in a comment.
Fix the test_minidom failure.
We just need to force the encoding when no encoding is passed to toxml()
or toprettyxml(), rather than relying on the default encoding (which is
unreliable).
Change the repr() of frozenset instances (and set subclasses)
from name([e1, e2, ...]) to name({e1, e2, ...}).
This makes more sense now we have the set notation.
Selectively enable tests for asyncore.readwrite based on the presence
of poll support in the select module (since this is the only case in
which readwrite can be called). [GSoC - Alan McIntyre]
........
r56488 | nick.coghlan | 2007-07-22 03:18:07 -0700 (Sun, 22 Jul 2007) | 1 line
Correctly cleanup sys.modules after executing runpy relative import
tests
Restore Python 2.4 ImportError when attempting to execute a package
(as imports cannot be guaranteed to work properly if you try it)
........
r56519 | nick.coghlan | 2007-07-24 06:07:38 -0700 (Tue, 24 Jul 2007) | 1 line
Tweak runpy test to do a better job of confirming that sys has been manipulated correctly
........
r56520 | nick.coghlan | 2007-07-24 06:58:28 -0700 (Tue, 24 Jul 2007) | 1 line
Fix an incompatibility between the -i and -m command line switches as reported on python-dev by PJE - runpy.run_module now leaves any changes it makes to the sys module intact after the function terminates
........
r56523 | nick.coghlan | 2007-07-24 07:39:23 -0700 (Tue, 24 Jul 2007) | 1 line
Try to get rid of spurious failure in test_resource on the Debian buildbots by changing the file size limit before attempting to close the file
........
r56533 | facundo.batista | 2007-07-24 14:20:42 -0700 (Tue, 24 Jul 2007) | 7 lines
New tests for basic behavior of smtplib.SMTP and
smtpd.DebuggingServer. Change to use global host & port number
variables. Modified the 'server' to take a string to send back in
order to vary test server responses. Added a test for the reaction of
smtplib.SMTP to a non-200 HELO response. [GSoC - Alan McIntyre]
........
r56538 | nick.coghlan | 2007-07-25 05:57:48 -0700 (Wed, 25 Jul 2007) | 1 line
More buildbot cleanup - let the OS assign the port for test_urllib2_localnet
........
r56539 | nick.coghlan | 2007-07-25 06:18:58 -0700 (Wed, 25 Jul 2007) | 1 line
Add a temporary diagnostic message before a strange failure on the alpha Debian buildbot
........
r56543 | martin.v.loewis | 2007-07-25 09:24:23 -0700 (Wed, 25 Jul 2007) | 2 lines
Change location of the package index to pypi.python.org/pypi
........
r56551 | georg.brandl | 2007-07-26 02:36:25 -0700 (Thu, 26 Jul 2007) | 2 lines
tabs, newlines and crs are valid XML characters.
........
r56553 | nick.coghlan | 2007-07-26 07:03:00 -0700 (Thu, 26 Jul 2007) | 1 line
Add explicit test for a misbehaving math.floor
........
r56561 | mark.hammond | 2007-07-26 21:52:32 -0700 (Thu, 26 Jul 2007) | 3 lines
In consultation with Kristjan Jonsson, only define WINVER and _WINNT_WIN32
if (a) we are building Python itself and (b) no one previously defined them
........
r56562 | mark.hammond | 2007-07-26 22:08:54 -0700 (Thu, 26 Jul 2007) | 2 lines
Make test_math error messages more meaningful for small discrepancies in results
........
r56588 | martin.v.loewis | 2007-07-27 11:28:22 -0700 (Fri, 27 Jul 2007) | 2 lines
Bug #978833: Close https sockets by releasing the _ssl object.
........
r56601 | martin.v.loewis | 2007-07-28 00:03:05 -0700 (Sat, 28 Jul 2007) | 3 lines
Bug #1704793: Return UTF-16 pair if unicodedata.lookup cannot
represent the result in a single character.
........
r56604 | facundo.batista | 2007-07-28 07:21:22 -0700 (Sat, 28 Jul 2007) | 9 lines
Moved all of the capture_server socket setup code into the try block
so that the event gets set if a failure occurs during server setup
(otherwise the test will block forever). Changed to let the OS assign
the server port number, and client side of test waits for port number
assignment before proceeding. The test data in DispatcherWithSendTests
is also sent in multiple send() calls instead of one to make sure this
works properly. [GSoC - Alan McIntyre]
........
r56611 | georg.brandl | 2007-07-29 01:26:10 -0700 (Sun, 29 Jul 2007) | 2 lines
try-except-finally is new in 2.5.
........
r56617 | facundo.batista | 2007-07-29 07:23:08 -0700 (Sun, 29 Jul 2007) | 9 lines
Added tests for asynchat classes simple_producer & fifo, and the
find_prefix_at_end function. Check behavior of a string given as a
producer. Added tests for behavior of asynchat.async_chat when given
int, long, and None terminator arguments. Added usepoll attribute to
TestAsynchat to allow running the asynchat tests with poll support
chosen whether it's available or not (improves coverage of asyncore
code). [GSoC - Alan McIntyre]
........
r56620 | georg.brandl | 2007-07-29 10:38:35 -0700 (Sun, 29 Jul 2007) | 2 lines
Correct use of Py_BUILD_CORE - now make sure it is defined before it is
referenced, and also fix definition of _WIN32_WINNT.
Resolves patch 1761803.
........
r56632 | facundo.batista | 2007-07-30 20:03:34 -0700 (Mon, 30 Jul 2007) | 8 lines
When running asynchat tests on OS X (darwin), the test client now
overrides asyncore.dispatcher.handle_expt to do nothing, since
select.poll gives a POLLHUP error at the completion of these tests.
Added timeout & count arguments to several asyncore.loop calls to
avoid the possibility of a test hanging up a build. [GSoC - Alan
McIntyre]
........
r56633 | nick.coghlan | 2007-07-31 06:38:01 -0700 (Tue, 31 Jul 2007) | 1 line
Define _BSD_SOURCE, to get access to POSIX extensions on OpenBSD 4.1+.
........
r56653 | facundo.batista | 2007-08-01 16:18:36 -0700 (Wed, 01 Aug 2007) | 9 lines
Allow the OS to select a free port for each test server. For
DebuggingServerTests, construct SMTP objects with a localhost argument
to avoid abysmally long FQDN lookups (not relevant to items under
test) on some machines that would cause the test to fail. Moved server
setup code in the server function inside the try block to avoid the
possibility of setup failure hanging the test. Minor edits to conform
to PEP 8. [GSoC - Alan McIntyre]
........
r56681 | matthias.klose | 2007-08-02 14:33:13 -0700 (Thu, 02 Aug 2007) | 2 lines
- Allow Emacs 22 for building the documentation in info format.
........
r56689 | neal.norwitz | 2007-08-02 23:46:29 -0700 (Thu, 02 Aug 2007) | 1 line
Py_ssize_t is defined regardless of HAVE_LONG_LONG. Will backport
........
r56727 | hyeshik.chang | 2007-08-03 21:10:18 -0700 (Fri, 03 Aug 2007) | 3 lines
Fix gb18030 codec's bug that doesn't map two-byte characters on
GB18030 extension in encoding. (bug reported by Bjorn Stabell)
........
r56751 | neal.norwitz | 2007-08-04 20:23:31 -0700 (Sat, 04 Aug 2007) | 7 lines
Handle errors when generating a warning.
The value is always written to the returned pointer if getting it was
successful, even if a warning causes an error. (This probably doesn't matter
as the caller will probably discard the value.)
In the case of syntax errors, in py3k format_exception_only()
was including line number and position in the final line of the
exception notification, duplicating info in previous lines.
........
r56501 | kurt.kaiser | 2007-07-22 19:35:50 -0700 (Sun, 22 Jul 2007) | 2 lines
Hum, needed a newline in the last change.
........
r56536 | kurt.kaiser | 2007-07-24 19:06:48 -0700 (Tue, 24 Jul 2007) | 5 lines
Not all instantiations of SyntaxError set the args attribute.
e.g. symtable.c
Modify format_exception_only() to get SyntaxError attributes directly
instead of unpacking 'args'.
........
r56537 | kurt.kaiser | 2007-07-24 19:13:03 -0700 (Tue, 24 Jul 2007) | 3 lines
Update doctest strings: traceback.py no longer prints redundant location
information in the last line of the exception display.
........
r56627 | kurt.kaiser | 2007-07-29 21:06:57 -0700 (Sun, 29 Jul 2007) | 2 lines
Eliminate extra lines before and after tracebacks.
........
r56638 | kurt.kaiser | 2007-07-31 19:36:45 -0700 (Tue, 31 Jul 2007) | 3 lines
Refactor syntax error display in shell and edit windows; move
colorize_syntax_error() to EditorWindow; update to py3k.
........
r56685 | neal.norwitz | 2007-08-02 22:20:23 -0700 (Thu, 02 Aug 2007) | 10 lines
Remove several h/w and o/s specific modules that are undocumented, obsolete,
and/or not widely used:
linuxaudiodev.c, sunaudiodev.c Lib/plat-sunos5/SUNAUDIODEV.py
Lib/audiodev.py Tools/audiopy/audiopy
Move Lib/toaiff.py to Demo.
See PEP 3108 for most of the details.
........
r56686 | neal.norwitz | 2007-08-02 22:21:48 -0700 (Thu, 02 Aug 2007) | 4 lines
Missed one module that should have been removed since it relied
on audiodev which was removed.
........
r56748 | neal.norwitz | 2007-08-04 19:19:04 -0700 (Sat, 04 Aug 2007) | 1 line
Make from X import * outside module scope an error.
........
r56750 | neal.norwitz | 2007-08-04 19:35:01 -0700 (Sat, 04 Aug 2007) | 1 line
Jeremy Hylton [Sat, 4 Aug 2007 19:22:00 +0000 (19:22 +0000)]
Change urllib to use HTTPConnection rather than old HTTP class.
The HTTP class is a backwards compatibility layer for the Python 1.5
API. (The only remaining use in the std library is xmlrpclib.)
The current change makes urllib issue HTTP/1.0 requests with
HTTPConnection, because is accesses HTTPResponse.fp directly instead
of using the read() method. Using fp directly interacts poorly with
persistent connections. There are probably better solutions than the
current one, but this is a start.
Make test_tokenize pass again:
Add code to test_roundtrip() that figures out the encoding
from the first two lines of the file.
(We need to refactor this again to make it available to
all places that need this, e.g. linecache.py.)
Fix an obvious bug caused by a switch to Unicode.
However, this will need to be fixed further to allow non-ASCII letters in
names; leaving this to MvL.
Jeremy Hylton [Sat, 4 Aug 2007 03:25:17 +0000 (03:25 +0000)]
Fix several more paths from the SSL code.
In particular, watch out for comparing b"" to "". They're not equal,
but you can start at the code asking whether buf == "" for a long time
before realizing that it will never be True.
Jeremy Hylton [Sat, 4 Aug 2007 02:34:24 +0000 (02:34 +0000)]
HTTPResponse should not inherit from io.IOBase.
I'm not sure why I thought it should originally, but it introduces an
__del__() method on the response which cause the close() to be called
too soon using the HTTP compat class.
Also, remove some stale comments. The HTTPResponse calls makefile()
immediately, so there is no risk of it closing the socket.
Jeremy Hylton [Fri, 3 Aug 2007 20:40:09 +0000 (20:40 +0000)]
Make sure socket.close() doesn't interfere with socket.makefile().
If a makefile()-generated object is open and its parent socket is
closed, the parent socket should remain open until the child is
closed, too. The code to support this is moderately complex and
requires one extra slots in the socket object.
This change fixes httplib so that several urllib2net test cases pass
again.
Add SocketCloser class to socket.py, which encapsulates the
refcounting logic for sockets after makefile() has been called.
Move SocketIO class from io module to socket module. It's only use is
to implement the raw I/O methods on top of a socket to support
makefile().
Add unittests to test_socket to cover various patterns of close and
makefile.
Jeremy Hylton [Fri, 3 Aug 2007 20:31:38 +0000 (20:31 +0000)]
Hack: Fix some test_urllib2.net tests by getting them access to the
raw socket so they can check the timeout value. Should change the
code under test to expose the timeout in a more direct way.
SF patch# 1762940 by Joe Gregorio.
Fix test_cookielib and test_urllib2.
(The changes to urllib make urllib.quote() work correctly for Unicode
strings; but they don't fix test_urllib.)
Add a default __prepare__() method to 'type', so it can be called
using super(). (See recent conversation on python-3000 with Talin
and Phillip Eby about PEP 3115 chaining rules.)
Changes to long and float by Jeffrey Jasskin to conform to PEP 3141.
In particular, add trivial implementations of .real, .imag and .conjugate()
to both, and add .numerator and .denominator to long.
Also some small optimizations (e.g. remove long_pos in favor of long_long).
Tests for @abstractproperty by Jeffrey Yasskin.
(The previous changes to abc.py were also by him).
Put back a comment about using super() for properties
(I didn't realize this worked).
Walter Dörwald [Mon, 30 Jul 2007 13:31:40 +0000 (13:31 +0000)]
Bytes (which are the input for decoding) are mutable now. If a decoding
error callback changes the bytes object in the exception the decoder might
use memory that's no longer in use. Change unicode_decode_call_errorhandler()
so that it fetches the adresses of the bytes array (start and end) from the
exception object and passes them back to the caller.
Add a missing "rf.depth = 0;" to marshal_load().
Reported by Joe Smith. This makes the CYGWIN tests pass;
it's a miracle it didn't fail on other platforms.
Seems like it was accidentally dropped (maybe a merge artifact?).
Fix the minidom test.
In order to do this, I added an optional encoding argument to io.StringIO.
The toprettyxml() function returns bytes when you specify an encoding now.
projects / python / log